freeipa/ipaserver/plugins/ldap2.py

1791 lines
67 KiB
Python
Raw Normal View History

2009-03-31 13:11:44 -05:00
# Authors:
# Pavel Zuna <pzuna@redhat.com>
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
# John Dennis <jdennis@redhat.com>
2009-03-31 13:11:44 -05:00
#
# Copyright (C) 2009 Red Hat
# see file 'COPYING' for use and warranty information
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
2009-03-31 13:11:44 -05:00
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
2009-03-31 13:11:44 -05:00
"""
Backend plugin for LDAP.
"""
# Entries are represented as (dn, entry_attrs), where entry_attrs is a dict
# mapping attribute names to values. Values can be a single value or list/tuple
# of virtually any type. Each method passing these values to the python-ldap
# binding encodes them into the appropriate representation. This applies to
# everything except the CrudBackend methods, where dn is part of the entry dict.
import copy
import os
import socket
import string
import shutil
import tempfile
import time
import re
import pwd
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
import sys
from decimal import Decimal
2009-03-31 13:11:44 -05:00
import krbV
from ipapython.ipa_log_manager import *
2009-03-31 13:11:44 -05:00
import ldap as _ldap
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
from ldap.ldapobject import SimpleLDAPObject
2009-03-31 13:11:44 -05:00
import ldap.filter as _ldap_filter
import ldap.sasl as _ldap_sasl
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
from ipapython.dn import DN, RDN
from ipapython.ipautil import CIDict
from collections import namedtuple
from ipalib.errors import NetworkError, DatabaseError
try:
from ldap.controls.simple import GetEffectiveRightsControl #pylint: disable=F0401,E0611
except ImportError:
"""
python-ldap 2.4.x introduced a new API for effective rights control, which
needs to be used or otherwise bind dn is not passed correctly. The following
class is created for backward compatibility with python-ldap 2.3.x.
Relevant BZ: https://bugzilla.redhat.com/show_bug.cgi?id=802675
"""
from ldap.controls import LDAPControl
class GetEffectiveRightsControl(LDAPControl):
def __init__(self, criticality, authzId=None):
LDAPControl.__init__(self, '1.3.6.1.4.1.42.2.27.9.5.2', criticality, authzId)
# for backward compatibility
from ipalib import _
2009-03-31 13:11:44 -05:00
from ipalib import api, errors
2009-03-31 13:11:44 -05:00
from ipalib.crud import CrudBackend
from ipalib.request import context
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
_debug_log_ldap = False
# Make python-ldap tuple style result compatible with Entry and Entity
# objects by allowing access to the dn (tuple index 0) via the 'dn'
# attribute name and the attr dict (tuple index 1) via the 'data'
# attribute name. Thus:
# r = result[0]
# r[0] == r.dn
# r[1] == r.data
LDAPEntry = namedtuple('LDAPEntry', ['dn', 'data'])
# Group Member types
MEMBERS_ALL = 0
MEMBERS_DIRECT = 1
MEMBERS_INDIRECT = 2
# SASL authentication mechanism
SASL_AUTH = _ldap_sasl.sasl({}, 'GSSAPI')
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
DN_SYNTAX_OID = '1.3.6.1.4.1.1466.115.121.1.12'
def unicode_from_utf8(val):
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
'''
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
val is a UTF-8 encoded string, return a unicode object.
'''
return val.decode('utf-8')
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def value_to_utf8(val):
'''
Coerce the val parameter to a UTF-8 encoded string representation
of the val.
'''
# If val is not a string we need to convert it to a string
# (specifically a unicode string). Naively we might think we need to
# call str(val) to convert to a string. This is incorrect because if
# val is already a unicode object then str() will call
# encode(default_encoding) returning a str object encoded with
# default_encoding. But we don't want to apply the default_encoding!
# Rather we want to guarantee the val object has been converted to a
# unicode string because from a unicode string we want to explicitly
# encode to a str using our desired encoding (utf-8 in this case).
#
# Note: calling unicode on a unicode object simply returns the exact
# same object (with it's ref count incremented). This means calling
# unicode on a unicode object is effectively a no-op, thus it's not
# inefficient.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
return unicode(val).encode('utf-8')
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
class _ServerSchema(object):
'''
Properties of a schema retrieved from an LDAP server.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
'''
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def __init__(self, server, schema):
self.server = server
self.schema = schema
self.retrieve_timestamp = time.time()
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
class SchemaCache(object):
'''
Cache the schema's from individual LDAP servers.
'''
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def __init__(self):
log_mgr.get_logger(self, True)
self.servers = {}
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def get_schema(self, url, conn=None, force_update=False):
'''
Return schema belonging to a specific LDAP server.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
For performance reasons the schema is retrieved once and
cached unless force_update is True. force_update flushes the
existing schema for the server from the cache and reacquires
it.
'''
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if force_update:
self.flush(url)
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
server_schema = self.servers.get(url)
if server_schema is None:
schema = self._retrieve_schema_from_server(url, conn)
server_schema = _ServerSchema(url, schema)
self.servers[url] = server_schema
return server_schema.schema
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def flush(self, url):
self.debug('flushing %s from SchemaCache', url)
try:
del self.servers[url]
except KeyError:
pass
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def _retrieve_schema_from_server(self, url, conn=None):
"""
Retrieve the LDAP schema from the provided url and determine if
User-Private Groups (upg) are configured.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
Bind using kerberos credentials. If in the context of the
in-tree "lite" server then use the current ccache. If in the context of
Apache then create a new ccache and bind using the Apache HTTP service
principal.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
If a connection is provided then it the credentials bound to it are
used. The connection is not closed when the request is done.
"""
tmpdir = None
has_conn = conn is not None
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.debug('retrieving schema for SchemaCache url=%s conn=%s', url, conn)
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
try:
if api.env.context == 'server' and conn is None:
# FIXME: is this really what we want to do?
# This seems like this logic is in the wrong place and may conflict with other state.
try:
# Create a new credentials cache for this Apache process
tmpdir = tempfile.mkdtemp(prefix = "tmp-")
ccache_file = 'FILE:%s/ccache' % tmpdir
krbcontext = krbV.default_context()
principal = str('HTTP/%s@%s' % (api.env.host, api.env.realm))
keytab = krbV.Keytab(name='/etc/httpd/conf/ipa.keytab', context=krbcontext)
principal = krbV.Principal(name=principal, context=krbcontext)
prev_ccache = os.environ.get('KRB5CCNAME')
os.environ['KRB5CCNAME'] = ccache_file
ccache = krbV.CCache(name=ccache_file, context=krbcontext, primary_principal=principal)
ccache.init(principal)
ccache.init_creds_keytab(keytab=keytab, principal=principal)
except krbV.Krb5Error, e:
raise StandardError('Unable to retrieve LDAP schema. Error initializing principal %s in %s: %s' % (principal.name, '/etc/httpd/conf/ipa.keytab', str(e)))
finally:
if prev_ccache is not None:
os.environ['KRB5CCNAME'] = prev_ccache
if conn is None:
conn = IPASimpleLDAPObject(url)
if url.startswith('ldapi://'):
conn.set_option(_ldap.OPT_HOST_NAME, api.env.host)
conn.sasl_interactive_bind_s(None, SASL_AUTH)
try:
schema_entry = conn.search_s('cn=schema', _ldap.SCOPE_BASE,
attrlist=['attributetypes', 'objectclasses'])[0]
except _ldap.NO_SUCH_OBJECT:
# try different location for schema
# openldap has schema located in cn=subschema
self.debug('cn=schema not found, fallback to cn=subschema')
schema_entry = conn.search_s('cn=subschema', _ldap.SCOPE_BASE,
attrlist=['attributetypes', 'objectclasses'])[0]
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if not has_conn:
conn.unbind_s()
except _ldap.SERVER_DOWN:
raise NetworkError(uri=url,
error=u'LDAP Server Down, unable to retrieve LDAP schema')
except _ldap.LDAPError, e:
desc = e.args[0]['desc'].strip()
info = e.args[0].get('info', '').strip()
raise DatabaseError(desc = u'uri=%s' % url,
info = u'Unable to retrieve LDAP schema: %s: %s' % (desc, info))
except IndexError:
# no 'cn=schema' entry in LDAP? some servers use 'cn=subschema'
# TODO: DS uses 'cn=schema', support for other server?
# raise a more appropriate exception
raise
finally:
if tmpdir:
shutil.rmtree(tmpdir)
return _ldap.schema.SubSchema(schema_entry[1])
schema_cache = SchemaCache()
class IPASimpleLDAPObject(object):
'''
The purpose of this class is to provide a boundary between IPA and
python-ldap. In IPA we use IPA defined types because they are
richer and are designed to meet our needs. We also require that we
consistently use those types without exception. On the other hand
python-ldap uses different types. The goal is to be able to have
IPA code call python-ldap methods using the types native to
IPA. This class accomplishes that goal by exposing python-ldap
methods which take IPA types, convert them to python-ldap types,
call python-ldap, and then convert the results returned by
python-ldap into IPA types.
IPA code should never call python-ldap directly, it should only
call python-ldap methods in this class.
'''
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
# Note: the oid for dn syntax is: 1.3.6.1.4.1.1466.115.121.1.12
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
_SYNTAX_MAPPING = {
'1.3.6.1.4.1.1466.115.121.1.1' : str, # ACI item
'1.3.6.1.4.1.1466.115.121.1.4' : str, # Audio
'1.3.6.1.4.1.1466.115.121.1.5' : str, # Binary
'1.3.6.1.4.1.1466.115.121.1.8' : str, # Certificate
'1.3.6.1.4.1.1466.115.121.1.9' : str, # Certificate List
'1.3.6.1.4.1.1466.115.121.1.10' : str, # Certificate Pair
'1.3.6.1.4.1.1466.115.121.1.23' : str, # Fax
'1.3.6.1.4.1.1466.115.121.1.28' : str, # JPEG
'1.3.6.1.4.1.1466.115.121.1.40' : str, # OctetString (same as Binary)
'1.3.6.1.4.1.1466.115.121.1.49' : str, # Supported Algorithm
'1.3.6.1.4.1.1466.115.121.1.51' : str, # Teletext Terminal Identifier
DN_SYNTAX_OID : DN, # DN, member, memberof
'2.16.840.1.113730.3.8.3.3' : DN, # enrolledBy
'2.16.840.1.113730.3.8.3.18' : DN, # managedBy
'2.16.840.1.113730.3.8.3.5' : DN, # memberUser
'2.16.840.1.113730.3.8.3.7' : DN, # memberHost
'2.16.840.1.113730.3.8.3.20' : DN, # memberService
'2.16.840.1.113730.3.8.11.4' : DN, # ipaNTFallbackPrimaryGroup
'2.16.840.1.113730.3.8.11.21' : DN, # ipaAllowToImpersonate
'2.16.840.1.113730.3.8.11.22' : DN, # ipaAllowedTarget
'2.16.840.1.113730.3.8.7.1' : DN, # memberAllowCmd
'2.16.840.1.113730.3.8.7.2' : DN, # memberDenyCmd
'2.16.840.1.113719.1.301.4.14.1' : DN, # krbRealmReferences
'2.16.840.1.113719.1.301.4.17.1' : DN, # krbKdcServers
'2.16.840.1.113719.1.301.4.18.1' : DN, # krbPwdServers
'2.16.840.1.113719.1.301.4.26.1' : DN, # krbPrincipalReferences
'2.16.840.1.113719.1.301.4.29.1' : DN, # krbAdmServers
'2.16.840.1.113719.1.301.4.36.1' : DN, # krbPwdPolicyReference
'2.16.840.1.113719.1.301.4.40.1' : DN, # krbTicketPolicyReference
'2.16.840.1.113719.1.301.4.41.1' : DN, # krbSubTrees
'2.16.840.1.113719.1.301.4.52.1' : DN, # krbObjectReferences
'2.16.840.1.113719.1.301.4.53.1' : DN, # krbPrincContainerRef
}
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
# In most cases we lookup the syntax from the schema returned by
# the server. However, sometimes attributes may not be defined in
# the schema (e.g. extensibleObject which permits undefined
# attributes), or the schema was incorrectly defined (i.e. giving
# an attribute the syntax DirectoryString when in fact it's really
# a DN). This (hopefully sparse) table allows us to trap these
# anomalies and force them to be the syntax we know to be in use.
#
# FWIW, many entries under cn=config are undefined :-(
_SCHEMA_OVERRIDE = CIDict({
'managedtemplate': DN_SYNTAX_OID, # DN
'managedbase': DN_SYNTAX_OID, # DN
'originscope': DN_SYNTAX_OID, # DN
})
def __init__(self, uri):
log_mgr.get_logger(self, True)
self.uri = uri
self.conn = SimpleLDAPObject(uri)
self._schema = None
def _get_schema(self):
if self._schema is None:
# The schema may be updated during install or during
# updates, make sure we have a current version of the
# schema, not an out of date cached version.
force_update = api.env.context in ('installer', 'updates')
self._schema = schema_cache.get_schema(self.uri, self.conn, force_update=force_update)
return self._schema
schema = property(_get_schema, None, None, 'schema associated with this LDAP server')
def flush_cached_schema(self):
'''
Force this instance to forget it's cached schema and reacquire
it from the schema cache.
'''
# Currently this is called during bind operations to assure
# we're working with valid schema for a specific
# connection. This causes self._get_schema() to query the
# schema cache for the server's schema passing along a flag
# indicating if we're in a context that requires freshly
# loading the schema vs. returning the last cached version of
# the schema. If we're in a mode that permits use of
# previously cached schema the flush and reacquire is a very
# low cost operation.
#
# The schema is reacquired whenever this object is
# instantiated or when binding occurs. The schema is not
# reacquired for operations during a bound connection, it is
# presumed schema cannot change during this interval. This
# provides for maximum efficiency in contexts which do need
# schema refreshing by only peforming the refresh inbetween
# logical operations that have the potential to cause a schema
# change.
self._schema = None
def get_syntax(self, attr):
# Is this a special case attribute?
syntax = self._SCHEMA_OVERRIDE.get(attr)
if syntax is not None:
return syntax
# Try to lookup the syntax in the schema returned by the server
obj = self.schema.get_obj(_ldap.schema.AttributeType, attr)
if obj is not None:
return obj.syntax
else:
return None
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def has_dn_syntax(self, attr):
"""
Check the schema to see if the attribute uses DN syntax.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
Returns True/False
"""
syntax = self.get_syntax(attr)
return syntax == DN_SYNTAX_OID
def encode(self, val):
'''
'''
# Booleans are both an instance of bool and int, therefore
# test for bool before int otherwise the int clause will be
# entered for a boolean value instead of the boolean clause.
if isinstance(val, bool):
if val:
return 'TRUE'
else:
return 'FALSE'
elif isinstance(val, (unicode, float, int, long, Decimal, DN)):
return value_to_utf8(val)
elif isinstance(val, str):
return val
elif isinstance(val, list):
return [self.encode(m) for m in val]
elif isinstance(val, tuple):
return tuple(self.encode(m) for m in val)
elif isinstance(val, dict):
dct = dict((self.encode(k), self.encode(v)) for k, v in val.iteritems())
return dct
elif val is None:
return None
else:
raise TypeError("attempt to pass unsupported type to ldap, value=%s type=%s" %(val, type(val)))
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def convert_value_list(self, attr, target_type, values):
'''
'''
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
ipa_values = []
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
for original_value in values:
if isinstance(target_type, type) and isinstance(original_value, target_type):
ipa_value = original_value
else:
try:
ipa_value = target_type(original_value)
except Exception, e:
msg = 'unable to convert the attribute "%s" value "%s" to type %s' % (attr, original_value, target_type)
self.error(msg)
raise ValueError(msg)
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
ipa_values.append(ipa_value)
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
return ipa_values
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def convert_result(self, result):
'''
result is a python-ldap result tuple of the form (dn, attrs),
where dn is a string containing the dn (distinguished name) of
the entry, and attrs is a dictionary containing the attributes
associated with the entry. The keys of attrs are strings, and
the associated values are lists of strings.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
We convert the dn to a DN object.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
We convert every value associated with an attribute according
to it's syntax into the desired Python type.
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
returns a IPA result tuple of the same form as a python-ldap
result tuple except everything inside of the result tuple has
been converted to it's preferred IPA python type.
'''
ticket #1870 - subclass SimpleLDAPObject We use convenience types (classes) in IPA which make working with LDAP easier and more robust. It would be really nice if the basic python-ldap library understood our utility types and could accept them as parameters to the basic ldap functions and/or the basic ldap functions returned our utility types. Normally such a requirement would trivially be handled in an object- oriented language (which Python is) by subclassing to extend and modify the functionality. For some reason we didn't do this with the python-ldap classes. python-ldap objects are primarily used in two different places in our code, ipaserver.ipaldap.py for the IPAdmin class and in ipaserver/plugins/ldap2.py for the ldap2 class's .conn member. In IPAdmin we use a IPA utility class called Entry to make it easier to use the results returned by LDAP. The IPAdmin class is derived from python-ldap.SimpleLDAPObject. But for some reason when we added the support for the use of the Entry class in SimpleLDAPObject we didn't subclass SimpleLDAPObject and extend it for use with the Entry class as would be the normal expected methodology in an object-oriented language, rather we used an obscure feature of the Python language to override all methods of the SimpleLDAPObject class by wrapping those class methods in another function call. The reason why this isn't a good approach is: * It violates object-oriented methodology. * Other classes cannot be derived and inherit the customization (because the method wrapping occurs in a class instance, not within the class type). * It's non-obvious and obscure * It's inefficient. Here is a summary of what the code was doing: It iterated over every member of the SimpleLDAPObject class and if it was callable it wrapped the method. The wrapper function tested the name of the method being wrapped, if it was one of a handful of methods we wanted to customize we modified a parameter and called the original method. If the method wasn't of interest to use we still wrapped the method. It was inefficient because every non-customized method (the majority) executed a function call for the wrapper, the wrapper during run-time used logic to determine if the method was being overridden and then called the original method. So every call to ldap was doing extra function calls and logic processing which for the majority of cases produced nothing useful (and was non-obvious from brief code reading some methods were being overridden). Object-orientated languages have support built in for calling the right method for a given class object that do not involve extra function call overhead to realize customized class behaviour. Also when programmers look for customized class behaviour they look for derived classes. They might also want to utilize the customized class as the base class for their use. Also the wrapper logic was fragile, it did things like: if the method name begins with "add" I'll unconditionally modify the first and second argument. It would be some much cleaner if the "add", "add_s", etc. methods were overridden in a subclass where the logic could be seen and where it would apply to only the explicit functions and parameters being overridden. Also we would really benefit if there were classes which could be used as a base class which had specific ldap customization. At the moment our ldap customization needs are: 1) Support DN objects being passed to ldap operations 2) Support Entry & Entity objects being passed into and returned from ldap operations. We want to subclass the ldap SimpleLDAPObject class, that is the base ldap class with all the ldap methods we're using. IPASimpleLDAPObject class would subclass SimpleLDAPObject class which knows about DN objects (and possilby other IPA specific types that are universally used in IPA). Then IPAEntrySimpleLDAPObject would subclass IPASimpleLDAPObject which knows about Entry objects. The reason for the suggested class hierarchy is because DN objects will be used whenever we talk to LDAP (in the future we may want to add other IPA specific classes which will always be used). We don't add Entry support to the the IPASimpleLDAPObject class because Entry objects are (currently) only used in IPAdmin. What this patch does is: * Introduce IPASimpleLDAPObject derived from SimpleLDAPObject. IPASimpleLDAPObject is DN object aware. * Introduce IPAEntryLDAPObject derived from IPASimpleLDAPObject. IPAEntryLDAPObject is Entry object aware. * Derive IPAdmin from IPAEntryLDAPObject and remove the funky method wrapping from IPAdmin. * Code which called add_s() with an Entry or Entity object now calls addEntry(). addEntry() always existed, it just wasn't always used. add_s() had been modified to accept Entry or Entity object (why didn't we just call addEntry()?). The add*() ldap routine in IPAEntryLDAPObject have been subclassed to accept Entry and Entity objects, but that should proably be removed in the future and just use addEntry(). * Replace the call to ldap.initialize() in ldap2.create_connection() with a class constructor for IPASimpleLDAPObject. The ldap.initialize() is a convenience function in python-ldap, but it always returns a SimpleLDAPObject created via the SimpleLDAPObject constructor, thus ldap.initialize() did not allow subclassing, yet has no particular ease-of-use advantage thus we better off using the obvious class constructor mechanism. * Fix the use of _handle_errors(), it's not necessary to construct an empty dict to pass to it. If we follow the standard class derivation pattern for ldap we can make us of our own ldap utilities in a far easier, cleaner and more efficient manner.
2011-09-26 16:33:32 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
ipa_result = []
for dn_tuple in result:
original_dn = dn_tuple[0]
original_attrs = dn_tuple[1]
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
ipa_dn = DN(original_dn)
ipa_attrs = dict()
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
for attr, original_values in original_attrs.items():
target_type = self._SYNTAX_MAPPING.get(self.get_syntax(attr), unicode_from_utf8)
ipa_attrs[attr.lower()] = self.convert_value_list(attr, target_type, original_values)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
ipa_result.append(LDAPEntry(ipa_dn, ipa_attrs))
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if _debug_log_ldap:
self.debug('ldap.result: %s', ipa_result)
return ipa_result
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
#---------- python-ldap emulations ----------
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def add(self, dn, modlist):
assert isinstance(dn, DN)
dn = str(dn)
modlist = self.encode(modlist)
return self.conn.add(dn, modlist)
def add_ext(self, dn, modlist, serverctrls=None, clientctrls=None):
assert isinstance(dn, DN)
dn = str(dn)
modlist = self.encode(modlist)
return self.conn.add_ext(dn, modlist, serverctrls, clientctrls)
def add_ext_s(self, dn, modlist, serverctrls=None, clientctrls=None):
assert isinstance(dn, DN)
dn = str(dn)
modlist = self.encode(modlist)
return self.conn.add_ext_s(dn, modlist, serverctrls, clientctrls)
def add_s(self, dn, modlist):
assert isinstance(dn, DN)
dn = str(dn)
modlist = self.encode(modlist)
return self.conn.add_s(dn, modlist)
def bind(self, who, cred, method=_ldap.AUTH_SIMPLE):
self.flush_cached_schema()
if who is None:
who = DN()
assert isinstance(who, DN)
who = str(who)
cred = self.encode(cred)
return self.conn.bind(who, cred, method)
def delete(self, dn):
assert isinstance(dn, DN)
dn = str(dn)
return self.conn.delete(dn)
def delete_s(self, dn):
assert isinstance(dn, DN)
dn = str(dn)
return self.conn.delete_s(dn)
def get_option(self, option):
return self.conn.get_option(option)
def modify_s(self, dn, modlist):
assert isinstance(dn, DN)
dn = str(dn)
modlist = [(x[0], self.encode(x[1]), self.encode(x[2])) for x in modlist]
return self.conn.modify_s(dn, modlist)
def modrdn_s(self, dn, newrdn, delold=1):
assert isinstance(dn, DN)
dn = str(dn)
assert isinstance(newrdn, (DN, RDN))
newrdn = str(newrdn)
return self.conn.modrdn_s(dn, newrdn, delold)
def passwd_s(self, dn, oldpw, newpw, serverctrls=None, clientctrls=None):
assert isinstance(dn, DN)
dn = str(dn)
oldpw = self.encode(oldpw)
newpw = self.encode(newpw)
return self.conn.passwd_s(dn, oldpw, newpw, serverctrls, clientctrls)
def rename_s(self, dn, newrdn, newsuperior=None, delold=1):
# NOTICE: python-ldap of version 2.3.10 and lower does not support
# serverctrls and clientctrls for rename_s operation. Thus, these
# options are ommited from this command until really needed
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
dn = str(dn)
assert isinstance(newrdn, (DN, RDN))
newrdn = str(newrdn)
return self.conn.rename_s(dn, newrdn, newsuperior, delold)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def result(self, msgid=_ldap.RES_ANY, all=1, timeout=None):
resp_type, resp_data = self.conn.result(msgid, all, timeout)
resp_data = self.convert_result(resp_data)
return resp_type, resp_data
def sasl_interactive_bind_s(self, who, auth, serverctrls=None, clientctrls=None, sasl_flags=_ldap.SASL_QUIET):
self.flush_cached_schema()
if who is None:
who = DN()
assert isinstance(who, DN)
who = str(who)
return self.conn.sasl_interactive_bind_s(who, auth, serverctrls, clientctrls, sasl_flags)
def search(self, base, scope, filterstr='(objectClass=*)', attrlist=None, attrsonly=0):
assert isinstance(base, DN)
base = str(base)
filterstr = self.encode(filterstr)
attrlist = self.encode(attrlist)
return self.conn.search(base, scope, filterstr, attrlist, attrsonly)
def search_ext(self, base, scope, filterstr='(objectClass=*)', attrlist=None, attrsonly=0, serverctrls=None, clientctrls=None, timeout=-1, sizelimit=0):
assert isinstance(base, DN)
base = str(base)
filterstr = self.encode(filterstr)
attrlist = self.encode(attrlist)
if _debug_log_ldap:
self.debug("ldap.search_ext: dn: %s\nfilter: %s\nattrs_list: %s", base, filterstr, attrlist)
return self.conn.search_ext(base, scope, filterstr, attrlist, attrsonly, serverctrls, clientctrls, timeout, sizelimit)
def search_ext_s(self, base, scope, filterstr='(objectClass=*)', attrlist=None, attrsonly=0, serverctrls=None, clientctrls=None, timeout=-1, sizelimit=0):
assert isinstance(base, DN)
base = str(base)
filterstr = self.encode(filterstr)
attrlist = self.encode(attrlist)
ldap_result = self.conn.search_ext_s(base, scope, filterstr, attrlist, attrsonly, serverctrls, clientctrls, timeout, sizelimit)
ipa_result = self.convert_result(ldap_result)
return ipa_result
def search_s(self, base, scope, filterstr='(objectClass=*)', attrlist=None, attrsonly=0):
assert isinstance(base, DN)
base = str(base)
filterstr = self.encode(filterstr)
attrlist = self.encode(attrlist)
ldap_result = self.conn.search_s(base, scope, filterstr, attrlist, attrsonly)
ipa_result = self.convert_result(ldap_result)
return ipa_result
def search_st(self, base, scope, filterstr='(objectClass=*)', attrlist=None, attrsonly=0, timeout=-1):
assert isinstance(base, DN)
base = str(base)
filterstr = self.encode(filterstr)
attrlist = self.encode(attrlist)
ldap_result = self.conn.search_st(base, scope, filterstr, attrlist, attrsonly, timeout)
ipa_result = self.convert_result(ldap_result)
return ipa_result
def set_option(self, option, invalue):
return self.conn.set_option(option, invalue)
def simple_bind_s(self, who=None, cred='', serverctrls=None, clientctrls=None):
self.flush_cached_schema()
if who is None:
who = DN()
assert isinstance(who, DN)
who = str(who)
cred = self.encode(cred)
return self.conn.simple_bind_s(who, cred, serverctrls, clientctrls)
def start_tls_s(self):
return self.conn.start_tls_s()
def unbind(self):
self.flush_cached_schema()
return self.conn.unbind()
def unbind_s(self):
self.flush_cached_schema()
return self.conn.unbind_s()
class ldap2(CrudBackend):
"""
LDAP Backend Take 2.
"""
# attributes in this list cannot be deleted by update_entry
# only MOD_REPLACE operations are generated for them
_FORCE_REPLACE_ON_UPDATE_ATTRS = []
2009-03-31 13:11:44 -05:00
# rules for generating filters from entries
MATCH_ANY = '|' # (|(filter1)(filter2))
MATCH_ALL = '&' # (&(filter1)(filter2))
MATCH_NONE = '!' # (!(filter1)(filter2))
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
# search scope for find_entries()
SCOPE_BASE = _ldap.SCOPE_BASE
SCOPE_ONELEVEL = _ldap.SCOPE_ONELEVEL
SCOPE_SUBTREE = _ldap.SCOPE_SUBTREE
def __init__(self, shared_instance=True, ldap_uri=None, base_dn=None,
schema=None):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
log_mgr.get_logger(self, True)
CrudBackend.__init__(self, shared_instance=shared_instance)
try:
self.ldap_uri = ldap_uri or api.env.ldap_uri
except AttributeError:
self.ldap_uri = 'ldap://example.com'
try:
if base_dn is not None:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.base_dn = DN(base_dn)
else:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.base_dn = DN(api.env.basedn)
except AttributeError:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.base_dn = DN()
2009-03-31 13:11:44 -05:00
def __del__(self):
if self.isconnected():
self.disconnect()
2009-03-31 13:11:44 -05:00
def __str__(self):
return self.ldap_uri
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def _get_schema(self):
return self.conn.schema
schema = property(_get_schema, None, None, 'schema associated with this LDAP server')
# universal LDAPError handler
def handle_errors(self, e):
"""
Centralize error handling in one place.
e is the error to be raised
"""
if not isinstance(e, _ldap.TIMEOUT):
desc = e.args[0]['desc'].strip()
info = e.args[0].get('info', '').strip()
else:
desc = ''
info = ''
try:
# re-raise the error so we can handle it
raise e
except _ldap.NO_SUCH_OBJECT:
raise errors.NotFound(reason='no such entry')
except _ldap.ALREADY_EXISTS:
raise errors.DuplicateEntry()
except _ldap.CONSTRAINT_VIOLATION:
# This error gets thrown by the uniqueness plugin
if info.startswith('Another entry with the same attribute value already exists'):
raise errors.DuplicateEntry()
else:
raise errors.DatabaseError(desc=desc, info=info)
except _ldap.INSUFFICIENT_ACCESS:
raise errors.ACIError(info=info)
except _ldap.INVALID_CREDENTIALS:
raise errors.ACIError(info="%s %s" % (info, desc))
except _ldap.NO_SUCH_ATTRIBUTE:
# this is raised when a 'delete' attribute isn't found.
# it indicates the previous attribute was removed by another
# update, making the oldentry stale.
raise errors.MidairCollision()
except _ldap.INVALID_SYNTAX:
raise errors.InvalidSyntax(attr=info)
except _ldap.OBJECT_CLASS_VIOLATION:
raise errors.ObjectclassViolation(info=info)
except _ldap.ADMINLIMIT_EXCEEDED:
raise errors.LimitsExceeded()
except _ldap.SIZELIMIT_EXCEEDED:
raise errors.LimitsExceeded()
except _ldap.TIMELIMIT_EXCEEDED:
raise errors.LimitsExceeded()
except _ldap.NOT_ALLOWED_ON_RDN:
raise errors.NotAllowedOnRDN(attr=info)
except _ldap.FILTER_ERROR:
raise errors.BadSearchFilter(info=info)
except _ldap.NOT_ALLOWED_ON_NONLEAF:
raise errors.NotAllowedOnNonLeaf()
except _ldap.SERVER_DOWN:
raise NetworkError(uri=self.ldap_uri,
error=u'LDAP Server Down')
except _ldap.LOCAL_ERROR:
raise errors.ACIError(info=info)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
except _ldap.SUCCESS:
pass
except _ldap.LDAPError, e:
if 'NOT_ALLOWED_TO_DELEGATE' in info:
raise errors.ACIError(info="KDC returned NOT_ALLOWED_TO_DELEGATE")
self.info('Unhandled LDAPError: %s' % str(e))
raise errors.DatabaseError(desc=desc, info=info)
def get_syntax(self, attr, value):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if self.schema is None:
return None
obj = self.schema.get_obj(_ldap.schema.AttributeType, attr)
if obj is not None:
return obj.syntax
else:
return None
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def has_dn_syntax(self, attr):
return self.conn.has_dn_syntax(attr)
def get_allowed_attributes(self, objectclasses, raise_on_unknown=False):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if self.schema is None:
return None
allowed_attributes = []
for oc in objectclasses:
obj = self.schema.get_obj(_ldap.schema.ObjectClass, oc)
if obj is not None:
allowed_attributes += obj.must + obj.may
elif raise_on_unknown:
raise errors.NotFound(reason=_('objectclass %s not found') % oc)
return [unicode(a).lower() for a in list(set(allowed_attributes))]
def get_single_value(self, attr):
"""
Check the schema to see if the attribute is single-valued.
If the attribute is in the schema then returns True/False
If there is a problem loading the schema or the attribute is
not in the schema return None
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if self.schema is None:
return None
obj = self.schema.get_obj(_ldap.schema.AttributeType, attr)
return obj and obj.single_value
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def create_connection(self, ccache=None, bind_dn=None, bind_pw='',
tls_cacertfile=None, tls_certfile=None, tls_keyfile=None,
debug_level=0, autobind=False):
2009-03-31 13:11:44 -05:00
"""
Connect to LDAP server.
Keyword arguments:
ldapuri -- the LDAP server to connect to
2009-03-31 13:11:44 -05:00
ccache -- Kerberos V5 ccache name
bind_dn -- dn used to bind to the server
bind_pw -- password used to bind to the server
debug_level -- LDAP debug level option
tls_cacertfile -- TLS CA certificate filename
tls_certfile -- TLS certificate filename
tls_keyfile - TLS bind key filename
autobind - autobind as the current user
2009-03-31 13:11:44 -05:00
Extends backend.Connectible.create_connection.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if bind_dn is None:
bind_dn = DN()
assert isinstance(bind_dn, DN)
2009-03-31 13:11:44 -05:00
if tls_cacertfile is not None:
_ldap.set_option(_ldap.OPT_X_TLS_CACERTFILE, tls_cacertfile)
if tls_certfile is not None:
_ldap.set_option(_ldap.OPT_X_TLS_CERTFILE, tls_certfile)
if tls_keyfile is not None:
_ldap.set_option(_ldap.OPT_X_TLS_KEYFILE, tls_keyfile)
2011-01-10 10:17:20 -06:00
if debug_level:
2011-01-25 11:46:26 -06:00
_ldap.set_option(_ldap.OPT_DEBUG_LEVEL, debug_level)
2011-01-10 10:17:20 -06:00
try:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
conn = IPASimpleLDAPObject(self.ldap_uri)
if self.ldap_uri.startswith('ldapi://') and ccache:
conn.set_option(_ldap.OPT_HOST_NAME, api.env.host)
minssf = conn.get_option(_ldap.OPT_X_SASL_SSF_MIN)
maxssf = conn.get_option(_ldap.OPT_X_SASL_SSF_MAX)
# Always connect with at least an SSF of 56, confidentiality
# This also protects us from a broken ldap.conf
if minssf < 56:
minssf = 56
conn.set_option(_ldap.OPT_X_SASL_SSF_MIN, minssf)
if maxssf < minssf:
conn.set_option(_ldap.OPT_X_SASL_SSF_MAX, minssf)
if ccache is not None:
os.environ['KRB5CCNAME'] = ccache
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
conn.sasl_interactive_bind_s(None, SASL_AUTH)
principal = krbV.CCache(name=ccache,
context=krbV.default_context()).principal().name
setattr(context, 'principal', principal)
else:
# no kerberos ccache, use simple bind or external sasl
if autobind:
pent = pwd.getpwuid(os.geteuid())
auth_tokens = _ldap.sasl.external(pent.pw_name)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
conn.sasl_interactive_bind_s(None, auth_tokens)
else:
conn.simple_bind_s(bind_dn, bind_pw)
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
return conn
def destroy_connection(self):
"""Disconnect from LDAP server."""
try:
self.conn.unbind_s()
except _ldap.LDAPError:
# ignore when trying to unbind multiple times
pass
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
def normalize_dn(self, dn):
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
Normalize distinguished name by assuring it ends with
the base_dn.
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
Note: You don't have to normalize DN's before passing them to
ldap2 methods. It's done internally for you.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if not dn.endswith(self.base_dn):
# DN's are mutable, don't use in-place addtion (+=) which would
# modify the dn passed in with unintended side-effects. Addition
# returns a new DN object which is the concatenation of the two.
dn = dn + self.base_dn
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
return dn
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def make_dn_from_attr(self, attr, value, parent_dn=None):
"""
Make distinguished name from attribute.
Keyword arguments:
parent_dn -- DN of the parent entry (default '')
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if parent_dn is None:
parent_dn = DN()
assert isinstance(parent_dn, DN)
parent_dn = self.normalize_dn(parent_dn)
if isinstance(value, (list, tuple)):
value = value[0]
return DN((attr, value), parent_dn)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def make_dn(self, entry_attrs, primary_key='cn', parent_dn=None):
2009-03-31 13:11:44 -05:00
"""
Make distinguished name from entry attributes.
Keyword arguments:
primary_key -- attribute from which to make RDN (default 'cn')
parent_dn -- DN of the parent entry (default '')
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
2009-03-31 13:11:44 -05:00
assert primary_key in entry_attrs
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if parent_dn is None:
parent_dn = DN()
parent_dn = self.normalize_dn(parent_dn)
return DN((primary_key, entry_attrs[primary_key]), parent_dn)
def add_entry(self, dn, entry_attrs, normalize=True):
2009-03-31 13:11:44 -05:00
"""Create a new entry."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
if normalize:
dn = self.normalize_dn(dn)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
# remove all None or [] values, python-ldap hates'em
entry_attrs = dict(
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
# FIXME, shouldn't these values be an error?
(k, v) for (k, v) in entry_attrs.iteritems()
if v is not None and v != []
)
2009-03-31 13:11:44 -05:00
try:
self.conn.add_s(dn, list(entry_attrs.iteritems()))
2009-03-31 13:11:44 -05:00
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
# generating filters for find_entry
# some examples:
# f1 = ldap2.make_filter_from_attr(u'firstName', u'Pavel')
# f2 = ldap2.make_filter_from_attr(u'lastName', u'Zuna')
# f = ldap2.combine_filters([f1, f2], ldap2.MATCH_ALL)
# # f should be (&(firstName=Pavel)(lastName=Zuna))
# # it should be equivalent to:
# entry_attrs = {u'firstName': u'Pavel', u'lastName': u'Zuna'}
# f = ldap2.make_filter(entry_attrs, rules=ldap2.MATCH_ALL)
def combine_filters(self, filters, rules='|'):
"""
Combine filters into one for ldap2.find_entries.
Keyword arguments:
rules -- see ldap2.make_filter
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
2009-03-31 13:11:44 -05:00
assert isinstance(filters, (list, tuple))
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
filters = [f for f in filters if f]
if filters and rules == self.MATCH_NONE: # unary operator
return '(%s%s)' % (self.MATCH_NONE,
self.combine_filters(filters, self.MATCH_ANY))
2009-03-31 13:11:44 -05:00
if len(filters) > 1:
flt = '(%s' % rules
else:
flt = ''
for f in filters:
if not f.startswith('('):
f = '(%s)' % f
2009-03-31 13:11:44 -05:00
flt = '%s%s' % (flt, f)
if len(filters) > 1:
flt = '%s)' % flt
return flt
def make_filter_from_attr(self, attr, value, rules='|', exact=True,
leading_wildcard=True, trailing_wildcard=True):
2009-03-31 13:11:44 -05:00
"""
Make filter for ldap2.find_entries from attribute.
Keyword arguments:
rules -- see ldap2.make_filter
exact -- boolean, True - make filter as (attr=value)
False - make filter as (attr=*value*)
leading_wildcard -- boolean, True - allow heading filter wildcard when exact=False
False - forbid heading filter wildcard when exact=False
trailing_wildcard -- boolean, True - allow trailing filter wildcard when exact=False
False - forbid trailing filter wildcard when exact=False
2009-03-31 13:11:44 -05:00
"""
if isinstance(value, (list, tuple)):
make_filter_rules = self.MATCH_ANY if rules == self.MATCH_NONE else rules
flts = [ self.make_filter_from_attr(attr, v, exact=exact,
leading_wildcard=leading_wildcard,
trailing_wildcard=trailing_wildcard) for v in value ]
2009-03-31 13:11:44 -05:00
return self.combine_filters(flts, rules)
elif value is not None:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
value = _ldap_filter.escape_filter_chars(value_to_utf8(value))
if not exact:
template = '%s'
if leading_wildcard:
template = '*' + template
if trailing_wildcard:
template = template + '*'
value = template % value
if rules == self.MATCH_NONE:
return '(!(%s=%s))' % (attr, value)
return '(%s=%s)' % (attr, value)
return ''
2009-03-31 13:11:44 -05:00
def make_filter(self, entry_attrs, attrs_list=None, rules='|', exact=True,
leading_wildcard=True, trailing_wildcard=True):
2009-03-31 13:11:44 -05:00
"""
Make filter for ldap2.find_entries from entry attributes.
Keyword arguments:
attrs_list -- list of attributes to use, all if None (default None)
rules -- specifies how to determine a match (default ldap2.MATCH_ANY)
exact -- boolean, True - make filter as (attr=value)
False - make filter as (attr=*value*)
leading_wildcard -- boolean, True - allow heading filter wildcard when exact=False
False - forbid heading filter wildcard when exact=False
trailing_wildcard -- boolean, True - allow trailing filter wildcard when exact=False
False - forbid trailing filter wildcard when exact=False
2009-03-31 13:11:44 -05:00
rules can be one of the following:
ldap2.MATCH_NONE - match entries that do not match any attribute
ldap2.MATCH_ALL - match entries that match all attributes
ldap2.MATCH_ANY - match entries that match any of attribute
"""
make_filter_rules = self.MATCH_ANY if rules == self.MATCH_NONE else rules
2009-03-31 13:11:44 -05:00
flts = []
if attrs_list is None:
for (k, v) in entry_attrs.iteritems():
flts.append(
self.make_filter_from_attr(k, v, make_filter_rules, exact,
leading_wildcard, trailing_wildcard)
)
2009-03-31 13:11:44 -05:00
else:
for a in attrs_list:
value = entry_attrs.get(a, None)
if value is not None:
flts.append(
self.make_filter_from_attr(a, value, make_filter_rules, exact,
leading_wildcard, trailing_wildcard)
)
2009-03-31 13:11:44 -05:00
return self.combine_filters(flts, rules)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def find_entries(self, filter=None, attrs_list=None, base_dn=None,
scope=_ldap.SCOPE_SUBTREE, time_limit=None, size_limit=None,
normalize=True, search_refs=False):
2009-03-31 13:11:44 -05:00
"""
Return a list of entries and indication of whether the results were
truncated ([(dn, entry_attrs)], truncated) matching specified search
parameters followed by truncated flag. If the truncated flag is True,
search hit a server limit and its results are incomplete.
2009-03-31 13:11:44 -05:00
Keyword arguments:
attrs_list -- list of attributes to return, all if None (default None)
base_dn -- dn of the entry at which to start the search (default '')
scope -- search scope, see LDAP docs (default ldap2.SCOPE_SUBTREE)
time_limit -- time limit in seconds (default use IPA config values)
size_limit -- size (number of entries returned) limit (default use IPA config values)
normalize -- normalize the DN (default True)
search_refs -- allow search references to be returned (default skips these entries)
2009-03-31 13:11:44 -05:00
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if base_dn is None:
base_dn = DN()
assert isinstance(base_dn, DN)
if normalize:
base_dn = self.normalize_dn(base_dn)
if not filter:
filter = '(objectClass=*)'
res = []
truncated = False
2009-03-31 13:11:44 -05:00
if time_limit is None or size_limit is None:
(cdn, config) = self.get_ipa_config()
if time_limit is None:
time_limit = config.get('ipasearchtimelimit', [-1])[0]
if size_limit is None:
size_limit = config.get('ipasearchrecordslimit', [0])[0]
if time_limit == 0:
time_limit = -1
if not isinstance(size_limit, int):
size_limit = int(size_limit)
if not isinstance(time_limit, float):
time_limit = float(time_limit)
if attrs_list:
attrs_list = list(set(attrs_list))
2009-03-31 13:11:44 -05:00
# pass arguments to python-ldap
try:
id = self.conn.search_ext(
base_dn, scope, filter, attrs_list, timeout=time_limit,
sizelimit=size_limit
)
while True:
(objtype, res_list) = self.conn.result(id, 0)
if not res_list:
break
if objtype == _ldap.RES_SEARCH_ENTRY or \
(search_refs and objtype == _ldap.RES_SEARCH_REFERENCE):
res.append(res_list[0])
2009-03-31 13:11:44 -05:00
except (_ldap.ADMINLIMIT_EXCEEDED, _ldap.TIMELIMIT_EXCEEDED,
_ldap.SIZELIMIT_EXCEEDED), e:
truncated = True
2009-03-31 13:11:44 -05:00
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
if not res and not truncated:
2009-06-10 07:22:09 -05:00
raise errors.NotFound(reason='no such entry')
2009-03-31 13:11:44 -05:00
if attrs_list and ('memberindirect' in attrs_list or '*' in attrs_list):
for r in res:
if not 'member' in r[1]:
continue
else:
members = r[1]['member']
indirect = self.get_members(r[0], members, membertype=MEMBERS_INDIRECT,
time_limit=time_limit, size_limit=size_limit, normalize=normalize)
if len(indirect) > 0:
r[1]['memberindirect'] = indirect
if attrs_list and ('memberofindirect' in attrs_list or '*' in attrs_list):
for r in res:
if 'memberof' in r[1]:
memberof = r[1]['memberof']
del r[1]['memberof']
elif 'memberOf' in r[1]:
memberof = r[1]['memberOf']
del r[1]['memberOf']
else:
continue
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
(direct, indirect) = self.get_memberof(r[0], memberof, time_limit=time_limit,
size_limit=size_limit, normalize=normalize)
if len(direct) > 0:
r[1]['memberof'] = direct
if len(indirect) > 0:
r[1]['memberofindirect'] = indirect
return (res, truncated)
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
def find_entry_by_attr(self, attr, value, object_class, attrs_list=None, base_dn=None):
"""
Find entry (dn, entry_attrs) by attribute and object class.
Keyword arguments:
attrs_list - list of attributes to return, all if None (default None)
base_dn - dn of the entry at which to start the search (default '')
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if base_dn is None:
base_dn = DN()
assert isinstance(base_dn, DN)
search_kw = {attr: value, 'objectClass': object_class}
filter = self.make_filter(search_kw, rules=self.MATCH_ALL)
(entries, truncated) = self.find_entries(filter, attrs_list, base_dn)
if len(entries) > 1:
raise errors.SingleMatchExpected(found=len(entries))
else:
if truncated:
raise errors.LimitsExceeded()
else:
return entries[0]
def get_entry(self, dn, attrs_list=None, time_limit=None,
size_limit=None, normalize=True):
2009-03-31 13:11:44 -05:00
"""
Get entry (dn, entry_attrs) by dn.
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
Keyword arguments:
attrs_list - list of attributes to return, all if None (default None)
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
(entry, truncated) = self.find_entries(
None, attrs_list, dn, self.SCOPE_BASE, time_limit=time_limit,
size_limit=size_limit, normalize=normalize
)
if truncated:
raise errors.LimitsExceeded()
return entry[0]
2009-03-31 13:11:44 -05:00
config_defaults = {'ipasearchtimelimit': [2], 'ipasearchrecordslimit': [0]}
def get_ipa_config(self, attrs_list=None):
2009-03-31 13:11:44 -05:00
"""Returns the IPA configuration entry (dn, entry_attrs)."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
odn = api.Object.config.get_dn()
assert isinstance(odn, DN)
assert isinstance(api.env.basedn, DN)
cdn = DN(odn, api.env.basedn)
try:
config_entry = getattr(context, 'config_entry')
return (cdn, copy.deepcopy(config_entry))
except AttributeError:
# Not in our context yet
pass
try:
(entry, truncated) = self.find_entries(
None, attrs_list, base_dn=cdn, scope=self.SCOPE_BASE,
time_limit=2, size_limit=10
)
if truncated:
raise errors.LimitsExceeded()
(cdn, config_entry) = entry[0]
except errors.NotFound:
config_entry = {}
for a in self.config_defaults:
if a not in config_entry:
config_entry[a] = self.config_defaults[a]
setattr(context, 'config_entry', copy.deepcopy(config_entry))
return (cdn, config_entry)
2009-03-31 13:11:44 -05:00
def has_upg(self):
"""Returns True/False whether User-Private Groups are enabled.
This is determined based on whether the UPG Template exists.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
upg_dn = DN(('cn', 'UPG Definition'), ('cn', 'Definitions'), ('cn', 'Managed Entries'),
('cn', 'etc'), api.env.basedn)
try:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
upg_entry = self.conn.search_s(upg_dn, _ldap.SCOPE_BASE,
attrlist=['*'])[0]
disable_attr = '(objectclass=disable)'
if 'originfilter' in upg_entry[1]:
org_filter = upg_entry[1]['originfilter']
return not bool(re.search(r'%s' % disable_attr, org_filter[0]))
else:
return False
except _ldap.NO_SUCH_OBJECT, e:
return False
def get_effective_rights(self, dn, entry_attrs):
"""Returns the rights the currently bound user has for the given DN.
Returns 2 attributes, the attributeLevelRights for the given list of
attributes and the entryLevelRights for the entry itself.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
principal = getattr(context, 'principal')
(binddn, attrs) = self.find_entry_by_attr("krbprincipalname", principal, "krbPrincipalAux")
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(binddn, DN)
sctrl = [GetEffectiveRightsControl(True, "dn: " + str(binddn))]
self.conn.set_option(_ldap.OPT_SERVER_CONTROLS, sctrl)
(dn, attrs) = self.get_entry(dn, entry_attrs)
# remove the control so subsequent operations don't include GER
self.conn.set_option(_ldap.OPT_SERVER_CONTROLS, [])
return (dn, attrs)
def can_write(self, dn, attr):
"""Returns True/False if the currently bound user has write permissions
on the attribute. This only operates on a single attribute at a time.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
(dn, attrs) = self.get_effective_rights(dn, [attr])
if 'attributelevelrights' in attrs:
attr_rights = attrs.get('attributelevelrights')[0].decode('UTF-8')
(attr, rights) = attr_rights.split(':')
if 'w' in rights:
return True
return False
def can_read(self, dn, attr):
"""Returns True/False if the currently bound user has read permissions
on the attribute. This only operates on a single attribute at a time.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
(dn, attrs) = self.get_effective_rights(dn, [attr])
if 'attributelevelrights' in attrs:
attr_rights = attrs.get('attributelevelrights')[0].decode('UTF-8')
(attr, rights) = attr_rights.split(':')
if 'r' in rights:
return True
return False
#
# Entry-level effective rights
#
# a - Add
# d - Delete
# n - Rename the DN
# v - View the entry
#
def can_delete(self, dn):
"""Returns True/False if the currently bound user has delete permissions
on the entry.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
(dn, attrs) = self.get_effective_rights(dn, ["*"])
if 'entrylevelrights' in attrs:
entry_rights = attrs['entrylevelrights'][0].decode('UTF-8')
if 'd' in entry_rights:
return True
return False
def can_add(self, dn):
"""Returns True/False if the currently bound user has add permissions
on the entry.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
(dn, attrs) = self.get_effective_rights(dn, ["*"])
if 'entrylevelrights' in attrs:
entry_rights = attrs['entrylevelrights'][0].decode('UTF-8')
if 'a' in entry_rights:
return True
return False
2009-03-31 13:11:44 -05:00
def update_entry_rdn(self, dn, new_rdn, del_old=True):
"""
Update entry's relative distinguished name.
Keyword arguments:
del_old -- delete old RDN value (default True)
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
assert isinstance(new_rdn, RDN)
2009-03-31 13:11:44 -05:00
dn = self.normalize_dn(dn)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if dn[0] == new_rdn:
raise errors.EmptyModlist()
2009-03-31 13:11:44 -05:00
try:
self.conn.rename_s(dn, new_rdn, delold=int(del_old))
time.sleep(.3) # Give memberOf plugin a chance to work
2009-03-31 13:11:44 -05:00
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
def _generate_modlist(self, dn, entry_attrs, normalize):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
# get original entry
(dn, entry_attrs_old) = self.get_entry(dn, entry_attrs.keys(), normalize=normalize)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
# generate modlist
# for multi value attributes: no MOD_REPLACE to handle simultaneous
# updates better
# for single value attribute: always MOD_REPLACE
2009-03-31 13:11:44 -05:00
modlist = []
for (k, v) in entry_attrs.iteritems():
if v is None and k in entry_attrs_old:
2009-06-10 07:22:09 -05:00
modlist.append((_ldap.MOD_DELETE, k, None))
2009-03-31 13:11:44 -05:00
else:
if not isinstance(v, (list, tuple)):
v = [v]
v = set(filter(lambda value: value is not None, v))
2009-06-10 07:22:09 -05:00
old_v = set(entry_attrs_old.get(k.lower(), []))
2009-03-31 13:11:44 -05:00
# FIXME: Convert all values to either unicode, DN or str
# before detecting value changes (see IPASimpleLDAPObject for
# supported types).
# This conversion will set a common ground for the comparison.
#
# This fix can be removed when ticket 2265 is fixed and our
# encoded entry_attrs' types will match get_entry result
try:
v = set(unicode_from_utf8(self.conn.encode(value))
if not isinstance(value, (DN, str, unicode))
else value for value in v)
except Exception, e:
# Rather let the value slip in modlist than let ldap2 crash
self.error("Cannot convert attribute '%s' for modlist "
"for modlist comparison: %s", k, e)
2009-03-31 13:11:44 -05:00
adds = list(v.difference(old_v))
rems = list(old_v.difference(v))
is_single_value = self.get_single_value(k)
value_count = len(old_v) + len(adds) - len(rems)
if is_single_value and value_count > 1:
raise errors.OnlyOneValueAllowed(attr=k)
force_replace = False
if k in self._FORCE_REPLACE_ON_UPDATE_ATTRS or is_single_value:
force_replace = True
elif len(v) > 0 and len(v.intersection(old_v)) == 0:
force_replace = True
2009-03-31 13:11:44 -05:00
if adds:
if force_replace:
modlist.append((_ldap.MOD_REPLACE, k, adds))
else:
modlist.append((_ldap.MOD_ADD, k, adds))
2009-03-31 13:11:44 -05:00
if rems:
if not force_replace:
modlist.append((_ldap.MOD_DELETE, k, rems))
2009-03-31 13:11:44 -05:00
return modlist
def update_entry(self, dn, entry_attrs, normalize=True):
2009-03-31 13:11:44 -05:00
"""
Update entry's attributes.
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
An attribute value set to None deletes all current values.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
if normalize:
dn = self.normalize_dn(dn)
2009-03-31 13:11:44 -05:00
# generate modlist
modlist = self._generate_modlist(dn, entry_attrs, normalize)
2009-03-31 13:11:44 -05:00
if not modlist:
raise errors.EmptyModlist()
2009-03-31 13:11:44 -05:00
# pass arguments to python-ldap
try:
self.conn.modify_s(dn, modlist)
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
def delete_entry(self, dn, normalize=True):
2009-03-31 13:11:44 -05:00
"""Delete entry."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
if normalize:
dn = self.normalize_dn(dn)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
2009-03-31 13:11:44 -05:00
try:
self.conn.delete_s(dn)
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
2009-06-10 07:22:09 -05:00
def modify_password(self, dn, new_pass, old_pass=''):
2009-03-31 13:11:44 -05:00
"""Set user password."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
dn = self.normalize_dn(dn)
# The python-ldap passwd command doesn't verify the old password
# so we'll do a simple bind to validate it.
if old_pass != '':
try:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
conn = IPASimpleLDAPObject(self.ldap_uri)
conn.simple_bind_s(dn, old_pass)
conn.unbind()
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
try:
self.conn.passwd_s(dn, old_pass, new_pass)
2009-03-31 13:11:44 -05:00
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
def add_entry_to_group(self, dn, group_dn, member_attr='member', allow_same=False):
"""
Add entry designaed by dn to group group_dn in the member attribute
member_attr.
Adding a group as a member of itself is not allowed unless allow_same
is True.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
assert isinstance(group_dn, DN)
self.debug("add_entry_to_group: dn=%s group_dn=%s member_attr=%s", dn, group_dn, member_attr)
2009-03-31 13:11:44 -05:00
# check if the entry exists
(dn, entry_attrs) = self.get_entry(dn, ['objectclass'])
2009-03-31 13:11:44 -05:00
# get group entry
(group_dn, group_entry_attrs) = self.get_entry(group_dn, [member_attr])
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.debug("add_entry_to_group: group_entry_attrs=%s", group_entry_attrs)
# check if we're not trying to add group into itself
if dn == group_dn and not allow_same:
raise errors.SameGroupError()
2009-03-31 13:11:44 -05:00
# add dn to group entry's `member_attr` attribute
members = group_entry_attrs.get(member_attr, [])
members.append(dn)
group_entry_attrs[member_attr] = members
# update group entry
try:
self.update_entry(group_dn, group_entry_attrs)
except errors.EmptyModlist:
raise errors.AlreadyGroupMember()
2009-03-31 13:11:44 -05:00
def remove_entry_from_group(self, dn, group_dn, member_attr='member'):
"""Remove entry from group."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
assert isinstance(group_dn, DN)
self.debug("remove_entry_from_group: dn=%s group_dn=%s member_attr=%s", dn, group_dn, member_attr)
2009-03-31 13:11:44 -05:00
# get group entry
(group_dn, group_entry_attrs) = self.get_entry(group_dn, [member_attr])
2009-03-31 13:11:44 -05:00
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.debug("remove_entry_from_group: group_entry_attrs=%s", group_entry_attrs)
2009-03-31 13:11:44 -05:00
# remove dn from group entry's `member_attr` attribute
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
members = group_entry_attrs.get(member_attr, [])
assert all([isinstance(x, DN) for x in members])
2009-03-31 13:11:44 -05:00
try:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
members.remove(dn)
2009-03-31 13:11:44 -05:00
except ValueError:
raise errors.NotGroupMember()
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
group_entry_attrs[member_attr] = members
2009-03-31 13:11:44 -05:00
# update group entry
self.update_entry(group_dn, group_entry_attrs)
def get_members(self, group_dn, members, attr_list=[], membertype=MEMBERS_ALL, time_limit=None, size_limit=None, normalize=True):
"""Do a memberOf search of groupdn and return the attributes in
attr_list (an empty list returns all attributes).
membertype = MEMBERS_ALL all members returned
membertype = MEMBERS_DIRECT only direct members are returned
membertype = MEMBERS_INDIRECT only inherited members are returned
Members may be included in a group as a result of being a member
of a group that is a member of the group being queried.
Returns a list of DNs.
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(group_dn, DN)
if membertype not in [MEMBERS_ALL, MEMBERS_DIRECT, MEMBERS_INDIRECT]:
return None
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.debug("get_members: group_dn=%s members=%s membertype=%s", group_dn, members, membertype)
search_group_dn = _ldap_filter.escape_filter_chars(str(group_dn))
searchfilter = "(memberof=%s)" % search_group_dn
attr_list.append("member")
# Verify group membership
results = []
if membertype == MEMBERS_ALL or membertype == MEMBERS_INDIRECT:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
user_container_dn = DN(api.env.container_user, api.env.basedn) # FIXME, initialize once
host_container_dn = DN(api.env.container_host, api.env.basedn)
checkmembers = set(DN(x) for x in members)
checked = set()
while checkmembers:
member_dn = checkmembers.pop()
checked.add(member_dn)
# No need to check entry types that are not nested for
# additional members
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if member_dn.endswith(user_container_dn) or \
member_dn.endswith(host_container_dn):
results.append([member_dn, {}])
continue
try:
(result, truncated) = self.find_entries(searchfilter,
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
attr_list, member_dn, time_limit=time_limit,
size_limit=size_limit, scope=_ldap.SCOPE_BASE,
normalize=normalize)
if truncated:
raise errors.LimitsExceeded()
results.append(list(result[0]))
for m in result[0][1].get('member', []):
# This member may contain other members, add it to our
# candidate list
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if m not in checked:
checkmembers.add(m)
except errors.NotFound:
pass
if membertype == MEMBERS_ALL:
entries = []
for e in results:
entries.append(e[0])
return entries
(dn, group) = self.get_entry(group_dn, ['dn', 'member'],
size_limit=size_limit, time_limit=time_limit)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
real_members = group.get('member', [])
entries = []
for e in results:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
if e[0] not in real_members and e[0] not in entries:
if membertype == MEMBERS_INDIRECT:
entries.append(e[0])
else:
if membertype == MEMBERS_DIRECT:
entries.append(e[0])
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.debug("get_members: result=%s", entries)
return entries
def get_memberof(self, entry_dn, memberof, time_limit=None, size_limit=None, normalize=True):
"""
Examine the objects that an entry is a member of and determine if they
are a direct or indirect member of that group.
entry_dn: dn of the entry we want the direct/indirect members of
memberof: the memberOf attribute for entry_dn
Returns two memberof lists: (direct, indirect)
"""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(entry_dn, DN)
self.debug("get_memberof: entry_dn=%s memberof=%s", entry_dn, memberof)
if not type(memberof) in (list, tuple):
return ([], [])
if len(memberof) == 0:
return ([], [])
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
search_entry_dn = _ldap_filter.escape_filter_chars(str(entry_dn))
attr_list = ["dn", "memberof"]
searchfilter = "(|(member=%s)(memberhost=%s)(memberuser=%s))" % (
search_entry_dn, search_entry_dn, search_entry_dn)
# Search only the groups for which the object is a member to
# determine if it is directly or indirectly associated.
results = []
for group in memberof:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(group, DN)
try:
(result, truncated) = self.find_entries(searchfilter, attr_list,
group, time_limit=time_limit,size_limit=size_limit,
scope=_ldap.SCOPE_BASE, normalize=normalize)
results.extend(list(result))
except errors.NotFound:
pass
direct = []
# If there is an exception here, it is likely due to a failure in
# referential integrity. All members should have corresponding
# memberOf entries.
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
indirect = list(memberof)
for r in results:
direct.append(r[0])
try:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
indirect.remove(r[0])
except ValueError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.info('Failed to remove indirect entry %s from %s' % r[0], entry_dn)
raise e
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.debug("get_memberof: result direct=%s indirect=%s", direct, indirect)
return (direct, indirect)
2009-03-31 13:11:44 -05:00
def set_entry_active(self, dn, active):
"""Mark entry active/inactive."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
assert isinstance(active, bool)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
2009-03-31 13:11:44 -05:00
# get the entry in question
(dn, entry_attrs) = self.get_entry(dn, ['nsaccountlock'])
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
# check nsAccountLock attribute
account_lock_attr = entry_attrs.get('nsaccountlock', ['false'])
2009-03-31 13:11:44 -05:00
account_lock_attr = account_lock_attr[0].lower()
if active:
if account_lock_attr == 'false':
raise errors.AlreadyActive()
2009-03-31 13:11:44 -05:00
else:
if account_lock_attr == 'true':
raise errors.AlreadyInactive()
# LDAP expects string instead of Bool but it also requires it to be TRUE or FALSE,
# not True or False as Python stringification does. Thus, we uppercase it.
account_lock_attr = str(not active).upper()
entry_attrs['nsaccountlock'] = account_lock_attr
self.update_entry(dn, entry_attrs)
2009-03-31 13:11:44 -05:00
def activate_entry(self, dn):
"""Mark entry active."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
self.set_entry_active(dn, True)
def deactivate_entry(self, dn):
"""Mark entry inactive."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
self.set_entry_active(dn, False)
def remove_principal_key(self, dn):
"""Remove a kerberos principal key."""
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
dn = self.normalize_dn(dn)
# We need to do this directly using the LDAP library because we
# don't have read access to krbprincipalkey so we need to delete
# it in the blind.
mod = [(_ldap.MOD_REPLACE, 'krbprincipalkey', None),
(_ldap.MOD_REPLACE, 'krblastpwdchange', None)]
try:
self.conn.modify_s(dn, mod)
except _ldap.LDAPError, e:
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
self.handle_errors(e)
2009-03-31 13:11:44 -05:00
# CrudBackend methods
def _get_normalized_entry_for_crud(self, dn, attrs_list=None):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
(dn, entry_attrs) = self.get_entry(dn, attrs_list)
entry_attrs['dn'] = dn
2009-03-31 13:11:44 -05:00
return entry_attrs
def create(self, **kw):
"""
Create a new entry and return it as one dict (DN included).
Extends CrudBackend.create.
"""
assert 'dn' in kw
dn = kw['dn']
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
2009-03-31 13:11:44 -05:00
del kw['dn']
self.add_entry(dn, kw)
return self._get_normalized_entry_for_crud(dn)
def retrieve(self, primary_key, attributes):
"""
Get entry by primary_key (DN) as one dict (DN included).
Extends CrudBackend.retrieve.
"""
return self._get_normalized_entry_for_crud(primary_key, attributes)
def update(self, primary_key, **kw):
"""
Update entry's attributes and return it as one dict (DN included).
Extends CrudBackend.update.
"""
self.update_entry(primary_key, kw)
return self._get_normalized_entry_for_crud(primary_key)
def delete(self, primary_key):
"""
Delete entry by primary_key (DN).
2009-04-20 12:59:41 -05:00
2009-03-31 13:11:44 -05:00
Extends CrudBackend.delete.
"""
self.delete_entry(primary_key)
def search(self, **kw):
"""
Return a list of entries (each entry is one dict, DN included) matching
the specified criteria.
Keyword arguments:
filter -- search filter (default: '')
attrs_list -- list of attributes to return, all if None (default None)
base_dn -- dn of the entry at which to start the search (default '')
scope -- search scope, see LDAP docs (default ldap2.SCOPE_SUBTREE)
Extends CrudBackend.search.
"""
# get keyword arguments
filter = kw.pop('filter', None)
attrs_list = kw.pop('attrs_list', None)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
base_dn = kw.pop('base_dn', DN())
assert isinstance(base_dn, DN)
2009-03-31 13:11:44 -05:00
scope = kw.pop('scope', self.SCOPE_SUBTREE)
# generate filter
filter_tmp = self.make_filter(kw)
if filter:
filter = self.combine_filters((filter, filter_tmp), self.MATCH_ALL)
else:
filter = filter_tmp
if not filter:
filter = '(objectClass=*)'
# find entries and normalize the output for CRUD
output = []
(entries, truncated) = self.find_entries(
filter, attrs_list, base_dn, scope
)
2009-03-31 13:11:44 -05:00
for (dn, entry_attrs) in entries:
entry_attrs['dn'] = [dn]
output.append(entry_attrs)
if truncated:
return (-1, output)
return (len(output), output)
2009-03-31 13:11:44 -05:00
api.register(ldap2)