freeipa/ipatests/test_ipapython/test_dn.py

1447 lines
48 KiB
Python
Raw Normal View History

import contextlib
import pytest
from cryptography import x509
import six
from ipapython.dn import DN, RDN, AVA, str2dn, dn2str, DECODING_ERROR
from ipapython import dn_ctypes
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 six.PY3:
unicode = str
def cmp(a, b):
if a == b:
assert not a < b
assert not a > b
assert not a != b
assert a <= b
assert a >= b
return 0
elif a < b:
assert not a > b
assert a != b
assert a <= b
assert not a >= b
return -1
else:
assert a > b
assert a != b
assert not a <= b
assert a >= b
return 1
pytestmark = pytest.mark.tier0
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 expected_class(klass, component):
if klass is AVA:
if component == 'self':
return AVA
elif klass is RDN:
if component == 'self':
return RDN
elif component == 'AVA':
return AVA
elif klass is DN:
if component == 'self':
return DN
elif component == 'AVA':
return AVA
elif component == 'RDN':
return RDN
raise ValueError("class %s with component '%s' unknown" % (klass.__name__, component))
class TestAVA:
@pytest.fixture(autouse=True)
def ava_setup(self):
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.attr1 = 'cn'
self.value1 = 'Bob'
self.str_ava1 = '%s=%s' % (self.attr1, self.value1)
self.ava1 = AVA(self.attr1, self.value1)
self.attr2 = 'ou'
self.value2 = 'People'
self.str_ava2 = '%s=%s' % (self.attr2, self.value2)
self.ava2 = AVA(self.attr2, self.value2)
self.attr3 = 'c'
self.value3 = 'US'
self.str_ava3 = '%s=%s' % (self.attr3, self.value3)
self.ava3 = AVA(self.attr3, self.value3)
def assertExpectedClass(self, klass, obj, component):
assert obj.__class__ is expected_class(klass, component)
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 test_create(self):
# Create with attr,value pair
ava1 = AVA(self.attr1, self.value1)
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1 == self.ava1
# Create with "attr=value" string
ava1 = AVA(self.str_ava1)
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1 == self.ava1
# Create with tuple (attr, value)
ava1 = AVA((self.attr1, self.value1))
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1 == self.ava1
# Create with list [attr, value]
ava1 = AVA([self.attr1, self.value1])
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1 == self.ava1
# Create with no args should fail
with pytest.raises(TypeError):
AVA()
# Create with more than 3 args should fail
with pytest.raises(TypeError):
AVA(self.attr1, self.value1, self.attr1, self.attr1)
# Create with 1 arg which is not string should fail
with pytest.raises(TypeError):
AVA(1)
# Create with malformed AVA string should fail
with pytest.raises(ValueError):
AVA("cn")
# Create with non-string parameters, should convert
ava1 = AVA(1, self.value1)
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1.attr == u'1'
ava1 = AVA((1, self.value1))
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1.attr == u'1'
ava1 = AVA(self.attr1, 1)
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1.value == u'1'
ava1 = AVA((self.attr1, 1))
self.assertExpectedClass(AVA, ava1, 'self')
assert ava1.value == u'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
def test_indexing(self):
ava1 = AVA(self.ava1)
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 ava1[self.attr1] == self.value1
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 ava1[0] == self.attr1
assert ava1[1] == self.value1
with pytest.raises(KeyError):
ava1['foo'] # pylint: disable=pointless-statement
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
with pytest.raises(KeyError):
ava1[3] # pylint: disable=pointless-statement
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 test_properties(self):
ava1 = AVA(self.ava1)
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 ava1.attr == self.attr1
assert isinstance(ava1.attr, unicode)
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 ava1.value == self.value1
assert isinstance(ava1.value, unicode)
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 test_str(self):
ava1 = AVA(self.ava1)
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 str(ava1) == self.str_ava1
assert isinstance(str(ava1), str)
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 test_cmp(self):
# Equality
ava1 = AVA(self.attr1, self.value1)
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 ava1 == self.ava1
assert ava1 == self.ava1
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 ava1 == self.str_ava1
assert ava1 == self.str_ava1
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
result = cmp(ava1, self.ava1)
assert result == 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
# Upper case attr should still be equal
ava1 = AVA(self.attr1.upper(), self.value1)
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 ava1.attr != self.attr1
assert ava1.value == self.value1
assert ava1 == self.ava1
assert ava1 == self.ava1
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
result = cmp(ava1, self.ava1)
assert result == 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
# Upper case value should still be equal
ava1 = AVA(self.attr1, self.value1.upper())
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 ava1.attr == self.attr1
assert ava1.value != self.value1
assert ava1 == self.ava1
assert ava1 == self.ava1
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
result = cmp(ava1, self.ava1)
assert result == 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
# Make ava1's attr greater
with pytest.raises(AttributeError):
ava1.attr = self.attr1 + "1"
ava1 = AVA(self.attr1 + "1", self.value1.upper())
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 ava1 != self.ava1
assert ava1 != self.ava1
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
result = cmp(ava1, self.ava1)
assert result == 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
result = cmp(self.ava1, ava1)
assert result == -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
# Reset ava1's attr, should be equal again
with pytest.raises(AttributeError):
ava1.attr = self.attr1
ava1 = AVA(self.attr1, self.value1.upper())
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
result = cmp(ava1, self.ava1)
assert result == 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
# Make ava1's value greater
# attr will be equal, this tests secondary comparision component
with pytest.raises(AttributeError):
ava1.value = self.value1 + "1"
ava1 = AVA(self.attr1, self.value1 + "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
result = cmp(ava1, self.ava1)
assert result == 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
result = cmp(self.ava1, ava1)
assert result == -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
def test_hashing(self):
# create AVA's that are equal but differ in case
ava1 = AVA((self.attr1.lower(), self.value1.upper()))
ava2 = AVA((self.attr1.upper(), self.value1.lower()))
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
# AVAs that are equal should hash to the same value.
assert ava1 == ava2
assert hash(ava1) == hash(ava2)
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
# Different AVA objects with the same value should
# map to 1 common key and 1 member in a set. The key and
# member are based on the object's 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
ava1_a = AVA(self.ava1)
ava1_b = AVA(self.ava1)
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
ava2_a = AVA(self.ava2)
ava2_b = AVA(self.ava2)
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
ava3_a = AVA(self.ava3)
ava3_b = AVA(self.ava3)
assert ava1_a == ava1_b
assert ava2_a == ava2_b
assert ava3_a == ava3_b
d = dict()
s = set()
d[ava1_a] = str(ava1_a)
d[ava1_b] = str(ava1_b)
d[ava2_a] = str(ava2_a)
d[ava2_b] = str(ava2_b)
s.add(ava1_a)
s.add(ava1_b)
s.add(ava2_a)
s.add(ava2_b)
assert len(d) == 2
assert len(s) == 2
assert sorted(d) == sorted([ava1_a, ava2_a])
assert sorted(s) == sorted([ava1_a, ava2_a])
assert ava1_a in d
assert ava1_b in d
assert ava2_a in d
assert ava2_b in d
assert ava3_a not in d
assert ava3_b not in d
assert ava1_a in s
assert ava1_b in s
assert ava2_a in s
assert ava2_b in s
assert ava3_a not in s
assert ava3_b not in s
class TestRDN:
@pytest.fixture(autouse=True)
def rdn_setup(self):
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
# ava1 must sort before ava2
self.attr1 = 'cn'
self.value1 = 'Bob'
self.str_ava1 = '%s=%s' % (self.attr1, self.value1)
self.ava1 = AVA(self.attr1, self.value1)
self.str_rdn1 = '%s=%s' % (self.attr1, self.value1)
self.rdn1 = RDN((self.attr1, self.value1))
self.attr2 = 'ou'
self.value2 = 'people'
self.str_ava2 = '%s=%s' % (self.attr2, self.value2)
self.ava2 = AVA(self.attr2, self.value2)
self.str_rdn2 = '%s=%s' % (self.attr2, self.value2)
self.rdn2 = RDN((self.attr2, self.value2))
self.str_ava3 = '%s=%s+%s=%s' % (self.attr1, self.value1, self.attr2, self.value2)
self.str_rdn3 = '%s=%s+%s=%s' % (self.attr1, self.value1, self.attr2, self.value2)
self.rdn3 = RDN(self.ava1, self.ava2)
def assertExpectedClass(self, klass, obj, component):
assert obj.__class__ is expected_class(klass, component)
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 test_create(self):
# Create with single attr,value pair
rdn1 = RDN((self.attr1, self.value1))
assert len(rdn1) == 1
assert rdn1 == self.rdn1
self.assertExpectedClass(RDN, rdn1, 'self')
for i in range(0, len(rdn1)):
self.assertExpectedClass(RDN, rdn1[i], 'AVA')
assert rdn1[0] == self.ava1
# Create with multiple attr,value pairs
rdn3 = RDN((self.attr1, self.value1), (self.attr2, self.value2))
assert len(rdn3) == 2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
assert rdn3[0] == self.ava1
assert rdn3[1] == self.ava2
# Create with multiple attr,value pairs passed as lists
rdn3 = RDN([self.attr1, self.value1], [self.attr2, self.value2])
assert len(rdn3) == 2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
assert rdn3[0] == self.ava1
assert rdn3[1] == self.ava2
# Create with multiple attr,value pairs but reverse
# constructor parameter ordering. RDN canonical ordering
# should remain the same
rdn3 = RDN((self.attr2, self.value2), (self.attr1, self.value1))
assert len(rdn3) == 2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
assert rdn3[0] == self.ava1
assert rdn3[1] == self.ava2
# Create with single AVA object
rdn1 = RDN(self.ava1)
assert len(rdn1) == 1
assert rdn1 == self.rdn1
self.assertExpectedClass(RDN, rdn1, 'self')
for i in range(0, len(rdn1)):
self.assertExpectedClass(RDN, rdn1[i], 'AVA')
assert rdn1[0] == self.ava1
# Create with multiple AVA objects
rdn3 = RDN(self.ava1, self.ava2)
assert len(rdn3) == 2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
assert rdn3[0] == self.ava1
assert rdn3[1] == self.ava2
# Create with multiple AVA objects but reverse constructor
# parameter ordering. RDN canonical ordering should remain
# the same
rdn3 = RDN(self.ava2, self.ava1)
assert len(rdn3) == 2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
assert rdn3[0] == self.ava1
assert rdn3[1] == self.ava2
# Create with single string with 1 AVA
rdn1 = RDN(self.str_rdn1)
assert len(rdn1) == 1
assert rdn1 == self.rdn1
self.assertExpectedClass(RDN, rdn1, 'self')
for i in range(0, len(rdn1)):
self.assertExpectedClass(RDN, rdn1[i], 'AVA')
assert rdn1[0] == self.ava1
# Create with single string with 2 AVA's
rdn3 = RDN(self.str_rdn3)
assert len(rdn3) == 2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
assert rdn3[0] == self.ava1
assert rdn3[1] == self.ava2
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 test_properties(self):
rdn1 = RDN(self.rdn1)
rdn2 = RDN(self.rdn2)
rdn3 = RDN(self.rdn3)
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 rdn1.attr == self.attr1
assert isinstance(rdn1.attr, unicode)
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 rdn1.value == self.value1
assert isinstance(rdn1.value, unicode)
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 rdn2.attr == self.attr2
assert isinstance(rdn2.attr, unicode)
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 rdn2.value == self.value2
assert isinstance(rdn2.value, unicode)
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 rdn3.attr == self.attr1
assert isinstance(rdn3.attr, unicode)
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 rdn3.value == self.value1
assert isinstance(rdn3.value, unicode)
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 test_str(self):
rdn1 = RDN(self.rdn1)
rdn2 = RDN(self.rdn2)
rdn3 = RDN(self.rdn3)
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 str(rdn1) == self.str_rdn1
assert isinstance(str(rdn1), str)
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 str(rdn2) == self.str_rdn2
assert isinstance(str(rdn2), str)
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 str(rdn3) == self.str_rdn3
assert isinstance(str(rdn3), str)
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 test_cmp(self):
# Equality
rdn1 = RDN((self.attr1, self.value1))
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 rdn1 == self.rdn1
assert rdn1 == self.rdn1
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 rdn1 == self.str_rdn1
assert rdn1 == self.str_rdn1
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
result = cmp(rdn1, self.rdn1)
assert result == 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
# Make rdn1's attr greater
rdn1 = RDN((self.attr1 + "1", self.value1))
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 rdn1 != self.rdn1
assert rdn1 != self.rdn1
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
result = cmp(rdn1, self.rdn1)
assert result == 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
result = cmp(self.rdn1, rdn1)
assert result == -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
# Reset rdn1's attr, should be equal again
rdn1 = RDN((self.attr1, self.value1))
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
result = cmp(rdn1, self.rdn1)
assert result == 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
# Make rdn1's value greater
# attr will be equal, this tests secondary comparision component
rdn1 = RDN((self.attr1, self.value1 + "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
result = cmp(rdn1, self.rdn1)
assert result == 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
result = cmp(self.rdn1, rdn1)
assert result == -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
# Make sure rdn's with more ava's are greater
result = cmp(self.rdn1, self.rdn3)
assert result == -1
result = cmp(self.rdn3, self.rdn1)
assert result == 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
def test_indexing(self):
rdn1 = RDN(self.rdn1)
rdn2 = RDN(self.rdn2)
rdn3 = RDN(self.rdn3)
assert rdn1[0] == self.ava1
assert rdn1[self.ava1.attr] == self.ava1.value
with pytest.raises(KeyError):
rdn1['foo'] # pylint: disable=pointless-statement
assert rdn2[0] == self.ava2
assert rdn2[self.ava2.attr] == self.ava2.value
with pytest.raises(KeyError):
rdn2['foo'] # pylint: disable=pointless-statement
assert rdn3[0] == self.ava1
assert rdn3[self.ava1.attr] == self.ava1.value
assert rdn3[1] == self.ava2
assert rdn3[self.ava2.attr] == self.ava2.value
with pytest.raises(KeyError):
rdn3['foo'] # pylint: disable=pointless-statement
assert rdn1.attr == self.attr1
assert rdn1.value == self.value1
with pytest.raises(TypeError):
rdn3[1.0] # pylint: disable=pointless-statement
# Slices
assert rdn3[0:1] == [self.ava1]
assert rdn3[:] == [self.ava1, self.ava2]
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 test_assignments(self):
rdn = RDN((self.attr1, self.value1))
with pytest.raises(TypeError):
# pylint: disable=unsupported-assignment-operation
rdn[0] = self.ava2
def test_iter(self):
rdn1 = RDN(self.rdn1)
rdn2 = RDN(self.rdn2)
rdn3 = RDN(self.rdn3)
assert len(rdn1) == 1
assert rdn1[:] == [self.ava1]
for i, ava in enumerate(rdn1):
if i == 0:
assert ava == self.ava1
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
else:
pytest.fail(
"got iteration index %d, but len=%d" % (i, len(rdn1)))
assert len(rdn2) == 1
assert rdn2[:] == [self.ava2]
for i, ava in enumerate(rdn2):
if i == 0:
assert ava == self.ava2
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
else:
pytest.fail(
"got iteration index %d, but len=%d" % (i, len(rdn2)))
assert len(rdn3) == 2
assert rdn3[:] == [self.ava1, self.ava2]
for i, ava in enumerate(rdn3):
if i == 0:
assert ava == self.ava1
elif i == 1:
assert ava == self.ava2
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
else:
pytest.fail(
"got iteration index %d, but len=%d" % (i, len(rdn3)))
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 test_concat(self):
rdn1 = RDN((self.attr1, self.value1))
rdn2 = RDN((self.attr2, self.value2))
# in-place addtion
rdn1 += rdn2
assert rdn1 == self.rdn3
self.assertExpectedClass(RDN, rdn1, 'self')
for i in range(0, len(rdn1)):
self.assertExpectedClass(RDN, rdn1[i], 'AVA')
rdn1 = RDN((self.attr1, self.value1))
rdn1 += self.ava2
assert rdn1 == self.rdn3
self.assertExpectedClass(RDN, rdn1, 'self')
for i in range(0, len(rdn1)):
self.assertExpectedClass(RDN, rdn1[i], 'AVA')
rdn1 = RDN((self.attr1, self.value1))
rdn1 += self.str_ava2
assert rdn1 == self.rdn3
self.assertExpectedClass(RDN, rdn1, 'self')
for i in range(0, len(rdn1)):
self.assertExpectedClass(RDN, rdn1[i], 'AVA')
# concatenation
rdn1 = RDN((self.attr1, self.value1))
rdn3 = rdn1 + rdn2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
rdn3 = rdn1 + self.ava2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
rdn3 = rdn1 + self.str_ava2
assert rdn3 == self.rdn3
self.assertExpectedClass(RDN, rdn3, 'self')
for i in range(0, len(rdn3)):
self.assertExpectedClass(RDN, rdn3[i], 'AVA')
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 test_hashing(self):
# create RDN's that are equal but differ in case
rdn1 = RDN((self.attr1.lower(), self.value1.upper()))
rdn2 = RDN((self.attr1.upper(), self.value1.lower()))
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
# RDNs that are equal should hash to the same value.
assert rdn1 == rdn2
assert hash(rdn1) == hash(rdn2)
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 TestDN:
@pytest.fixture(autouse=True)
def dn_setup(self):
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
# ava1 must sort before ava2
self.attr1 = 'cn'
self.value1 = u'Bob'
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.str_ava1 = '%s=%s' % (self.attr1, self.value1)
self.ava1 = AVA(self.attr1, self.value1)
self.str_rdn1 = '%s=%s' % (self.attr1, self.value1)
self.rdn1 = RDN((self.attr1, self.value1))
self.attr2 = 'ou'
self.value2 = u'people'
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.str_ava2 = '%s=%s' % (self.attr2, self.value2)
self.ava2 = AVA(self.attr2, self.value2)
self.str_rdn2 = '%s=%s' % (self.attr2, self.value2)
self.rdn2 = RDN((self.attr2, self.value2))
self.str_dn1 = self.str_rdn1
self.dn1 = DN(self.rdn1)
self.str_dn2 = self.str_rdn2
self.dn2 = DN(self.rdn2)
self.str_dn3 = '%s,%s' % (self.str_rdn1, self.str_rdn2)
self.dn3 = DN(self.rdn1, self.rdn2)
self.base_rdn1 = RDN(('dc', 'redhat'))
self.base_rdn2 = RDN(('dc', 'com'))
self.base_dn = DN(self.base_rdn1, self.base_rdn2)
self.container_rdn1 = RDN(('cn', 'sudorules'))
self.container_rdn2 = RDN(('cn', 'sudo'))
self.container_dn = DN(self.container_rdn1, self.container_rdn2)
self.base_container_dn = DN((self.attr1, self.value1),
self.container_dn, self.base_dn)
ou = x509.NameAttribute(
x509.NameOID.ORGANIZATIONAL_UNIT_NAME, self.value2)
cn = x509.NameAttribute(x509.NameOID.COMMON_NAME, self.value1)
c = x509.NameAttribute(x509.NameOID.COUNTRY_NAME, 'AU')
st = x509.NameAttribute(
x509.NameOID.STATE_OR_PROVINCE_NAME, 'Queensland')
self.x500name = x509.Name([ou, cn])
self.x500nameMultiRDN = x509.Name([
x509.RelativeDistinguishedName([c, st]),
x509.RelativeDistinguishedName([cn]),
])
self.x500nameMultiRDN2 = x509.Name([
x509.RelativeDistinguishedName([st, c]),
x509.RelativeDistinguishedName([cn]),
])
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 assertExpectedClass(self, klass, obj, component):
assert obj.__class__ is expected_class(klass, component)
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 test_create(self):
# Create with single attr,value pair
dn1 = DN((self.attr1, self.value1))
assert len(dn1) == 1
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[0].attr, unicode)
assert isinstance(dn1[0].value, unicode)
assert dn1[0] == self.rdn1
# Create with single attr,value pair passed as a tuple
dn1 = DN((self.attr1, self.value1))
assert len(dn1) == 1
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
# Creation with multiple attr,value string pairs should fail
with pytest.raises(ValueError):
dn1 = DN(self.attr1, self.value1, self.attr2, self.value2)
# Create with multiple attr,value pairs passed as tuples & lists
dn1 = DN((self.attr1, self.value1), [self.attr2, self.value2])
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
assert dn1[1] == self.rdn2
# Create with multiple attr,value pairs passed as tuple and RDN
dn1 = DN((self.attr1, self.value1), RDN((self.attr2, self.value2)))
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
assert dn1[1] == self.rdn2
# Create with multiple attr,value pairs but reverse
# constructor parameter ordering. RDN ordering should also be
# reversed because DN's are a ordered sequence of RDN's
dn1 = DN((self.attr2, self.value2), (self.attr1, self.value1))
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn2
assert dn1[1] == self.rdn1
# Create with single RDN object
dn1 = DN(self.rdn1)
assert len(dn1) == 1
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
# Create with multiple RDN objects, assure ordering is preserved.
dn1 = DN(self.rdn1, self.rdn2)
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
assert dn1[1] == self.rdn2
# Create with multiple RDN objects in different order, assure
# ordering is preserved.
dn1 = DN(self.rdn2, self.rdn1)
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn2
assert dn1[1] == self.rdn1
# Create with single string with 1 RDN
dn1 = DN(self.str_rdn1)
assert len(dn1) == 1
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
# Create with single string with 2 RDN's
dn1 = DN(self.str_dn3)
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
assert dn1[1] == self.rdn2
# Create with a python-cryptography 'Name'
dn1 = DN(self.x500name)
assert len(dn1) == 2
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
assert isinstance(dn1[i].attr, unicode)
assert isinstance(dn1[i].value, unicode)
assert dn1[0] == self.rdn1
assert dn1[1] == self.rdn2
# Create from 'Name' with multi-valued RDN
dn1 = DN(self.x500nameMultiRDN)
assert len(dn1) == 2
assert len(dn1[1]) == 2
assert AVA('c', 'au') in dn1[1]
assert AVA('st', 'queensland') in dn1[1]
assert len(dn1[0]) == 1
assert self.ava1 in dn1[0]
# Create with RDN, and 2 DN's (e.g. attr + container + base)
dn1 = DN((self.attr1, self.value1), self.container_dn, self.base_dn)
assert len(dn1) == 5
dn_str = ','.join([str(self.rdn1),
str(self.container_rdn1), str(self.container_rdn2),
str(self.base_rdn1), str(self.base_rdn2)])
assert str(dn1) == dn_str
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 test_str(self):
dn1 = DN(self.dn1)
dn2 = DN(self.dn2)
dn3 = DN(self.dn3)
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 str(dn1) == self.str_dn1
assert isinstance(str(dn1), str)
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 str(dn2) == self.str_dn2
assert isinstance(str(dn2), str)
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 str(dn3) == self.str_dn3
assert isinstance(str(dn3), str)
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 test_cmp(self):
# Equality
dn1 = DN((self.attr1, self.value1))
assert dn1 == self.dn1
assert dn1 == self.dn1
assert dn1 == self.str_dn1
assert dn1 == self.str_dn1
result = cmp(dn1, self.dn1)
assert result == 0
# Make dn1's attr greater
with pytest.raises(AttributeError):
dn1[0].attr = self.attr1 + "1"
dn1 = DN((self.attr1 + "1", self.value1))
assert dn1 != self.dn1
assert dn1 != self.dn1
result = cmp(dn1, self.dn1)
assert result == 1
result = cmp(self.dn1, dn1)
assert result == -1
# Reset dn1's attr, should be equal again
with pytest.raises(AttributeError):
dn1[0].attr = self.attr1
dn1 = DN((self.attr1, self.value1))
result = cmp(dn1, self.dn1)
assert result == 0
# Make dn1's value greater
# attr will be equal, this tests secondary comparision component
with pytest.raises(AttributeError):
dn1[0].value = self.value1 + "1"
dn1 = DN((self.attr1, self.value1 + "1"))
result = cmp(dn1, self.dn1)
assert result == 1
result = cmp(self.dn1, dn1)
assert result == -1
# Make sure dn's with more rdn's are greater
result = cmp(self.dn1, self.dn3)
assert result == -1
result = cmp(self.dn3, self.dn1)
assert result == 1
# Test startswith, endswith
container_dn = DN(self.container_dn)
base_container_dn = DN(self.base_container_dn)
assert base_container_dn.startswith(self.rdn1)
assert base_container_dn.startswith(self.dn1)
assert base_container_dn.startswith(self.dn1 + container_dn)
assert not base_container_dn.startswith(self.dn2)
assert not base_container_dn.startswith(self.rdn2)
assert base_container_dn.startswith((self.dn1))
assert base_container_dn.startswith((self.rdn1))
assert not base_container_dn.startswith((self.rdn2))
assert base_container_dn.startswith((self.rdn2, self.rdn1))
assert base_container_dn.startswith((self.dn1, self.dn2))
assert base_container_dn.endswith(self.base_dn)
assert base_container_dn.endswith(container_dn + self.base_dn)
assert not base_container_dn.endswith(DN(self.base_rdn1))
assert base_container_dn.endswith(DN(self.base_rdn2))
assert base_container_dn.endswith(
(DN(self.base_rdn1), DN(self.base_rdn2)))
# Test "in" membership
assert self.container_rdn1 in container_dn
# pylint: disable=comparison-with-itself
assert container_dn in container_dn
assert self.base_rdn1 not in container_dn
assert self.container_rdn1 in base_container_dn
assert container_dn in base_container_dn
assert container_dn + self.base_dn in base_container_dn
assert self.dn1 + container_dn + self.base_dn in base_container_dn
assert self.dn1 + container_dn + self.base_dn == base_container_dn
assert self.container_rdn1 not in self.base_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 test_eq_multi_rdn(self):
dn1 = DN(self.ava1, 'ST=Queensland+C=AU')
dn2 = DN(self.ava1, 'C=AU+ST=Queensland')
assert dn1 == dn2
# ensure AVAs get sorted when constructing from x509.Name
dn3 = DN(self.x500nameMultiRDN)
dn4 = DN(self.x500nameMultiRDN2)
assert dn3 == dn4
# ensure AVAs get sorted in the same way regardless of what
# the DN was constructed from
assert dn1 == dn3
assert dn1 == dn4
assert dn2 == dn3
assert dn2 == dn4
def test_indexing(self):
dn1 = DN(self.dn1)
dn2 = DN(self.dn2)
dn3 = DN(self.dn3)
assert dn1[0] == self.rdn1
assert dn1[self.rdn1.attr] == self.rdn1.value
with pytest.raises(KeyError):
dn1['foo'] # pylint: disable=pointless-statement
assert dn2[0] == self.rdn2
assert dn2[self.rdn2.attr] == self.rdn2.value
with pytest.raises(KeyError):
dn2['foo'] # pylint: disable=pointless-statement
assert dn3[0] == self.rdn1
assert dn3[self.rdn1.attr] == self.rdn1.value
assert dn3[1] == self.rdn2
assert dn3[self.rdn2.attr] == self.rdn2.value
with pytest.raises(KeyError):
dn3['foo'] # pylint: disable=pointless-statement
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
with pytest.raises(TypeError):
dn3[1.0] # pylint: disable=pointless-statement
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 test_assignments(self):
dn = DN('t=0,t=1,t=2,t=3,t=4,t=5,t=6,t=7,t=8,t=9')
with pytest.raises(TypeError):
# pylint: disable=unsupported-assignment-operation
dn[0] = RDN('t=a')
with pytest.raises(TypeError):
# pylint: disable=unsupported-assignment-operation
dn[0:1] = [RDN('t=a'), RDN('t=b')]
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 test_iter(self):
dn1 = DN(self.dn1)
dn2 = DN(self.dn2)
dn3 = DN(self.dn3)
assert len(dn1) == 1
assert dn1[:] == self.rdn1
for i, ava in enumerate(dn1):
if i == 0:
assert ava == self.rdn1
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
else:
pytest.fail(
"got iteration index %d, but len=%d" % (i, len(self.rdn1)))
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 len(dn2) == 1
assert dn2[:] == self.rdn2
for i, ava in enumerate(dn2):
if i == 0:
assert ava == self.rdn2
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
else:
pytest.fail(
"got iteration index %d, but len=%d" % (i, len(self.rdn2)))
assert len(dn3) == 2
assert dn3[:] == DN(self.rdn1, self.rdn2)
for i, ava in enumerate(dn3):
if i == 0:
assert ava == self.rdn1
elif i == 1:
assert ava == self.rdn2
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
else:
pytest.fail(
"got iteration index %d, but len=%d" % (i, len(dn3)))
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 test_concat(self):
dn1 = DN((self.attr1, self.value1))
dn2 = DN([self.attr2, self.value2])
# in-place addtion
dn1 += dn2
assert dn1 == self.dn3
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
dn1 = DN((self.attr1, self.value1))
dn1 += self.rdn2
assert dn1 == self.dn3
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
dn1 = DN((self.attr1, self.value1))
dn1 += self.dn2
assert dn1 == self.dn3
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
dn1 = DN((self.attr1, self.value1))
dn1 += self.str_dn2
assert dn1 == self.dn3
self.assertExpectedClass(DN, dn1, 'self')
for i in range(0, len(dn1)):
self.assertExpectedClass(DN, dn1[i], 'RDN')
for j in range(0, len(dn1[i])):
self.assertExpectedClass(DN, dn1[i][j], 'AVA')
# concatenation
dn1 = DN((self.attr1, self.value1))
dn3 = dn1 + dn2
assert dn3 == self.dn3
self.assertExpectedClass(DN, dn3, 'self')
for i in range(0, len(dn3)):
self.assertExpectedClass(DN, dn3[i], 'RDN')
for j in range(0, len(dn3[i])):
self.assertExpectedClass(DN, dn3[i][j], 'AVA')
dn1 = DN((self.attr1, self.value1))
dn3 = dn1 + self.rdn2
assert dn3 == self.dn3
self.assertExpectedClass(DN, dn3, 'self')
for i in range(0, len(dn3)):
self.assertExpectedClass(DN, dn3[i], 'RDN')
for j in range(0, len(dn3[i])):
self.assertExpectedClass(DN, dn3[i][j], 'AVA')
dn3 = dn1 + self.str_rdn2
assert dn3 == self.dn3
self.assertExpectedClass(DN, dn3, 'self')
for i in range(0, len(dn3)):
self.assertExpectedClass(DN, dn3[i], 'RDN')
self.assertExpectedClass(DN, dn3[i][0], 'AVA')
dn3 = dn1 + self.str_dn2
assert dn3 == self.dn3
self.assertExpectedClass(DN, dn3, 'self')
self.assertExpectedClass(DN, dn3, 'self')
for i in range(0, len(dn3)):
self.assertExpectedClass(DN, dn3[i], 'RDN')
for j in range(0, len(dn3[i])):
self.assertExpectedClass(DN, dn3[i][j], 'AVA')
dn3 = dn1 + self.dn2
assert dn3 == self.dn3
self.assertExpectedClass(DN, dn3, 'self')
self.assertExpectedClass(DN, dn3, 'self')
for i in range(0, len(dn3)):
self.assertExpectedClass(DN, dn3[i], 'RDN')
for j in range(0, len(dn3[i])):
self.assertExpectedClass(DN, dn3[i][j], 'AVA')
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 test_find(self):
# -10 -9 -8 -7 -6 -5 -4 -3 -2 -1
dn = DN('t=0,t=1,cn=bob,t=3,t=4,t=5,cn=bob,t=7,t=8,t=9')
pat = DN('cn=bob')
# forward
assert dn.find(pat) == 2
assert dn.find(pat, 1) == 2
assert dn.find(pat, 1, 3) == 2
assert dn.find(pat, 2, 3) == 2
assert dn.find(pat, 6) == 6
assert dn.find(pat, 7) == -1
assert dn.find(pat, 1, 2) == -1
with pytest.raises(ValueError):
assert dn.index(pat, 7) == -1
with pytest.raises(ValueError):
assert dn.index(pat, 1, 2) == -1
# reverse
assert dn.rfind(pat) == 6
assert dn.rfind(pat, -4) == 6
assert dn.rfind(pat, 6) == 6
assert dn.rfind(pat, 6, 8) == 6
assert dn.rfind(pat, 6, 8) == 6
assert dn.rfind(pat, -8) == 6
assert dn.rfind(pat, -8, -4) == 6
assert dn.rfind(pat, -8, -5) == 2
assert dn.rfind(pat, 7) == -1
assert dn.rfind(pat, -3) == -1
with pytest.raises(ValueError):
assert dn.rindex(pat, 7) == -1
with pytest.raises(ValueError):
assert dn.rindex(pat, -3) == -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
def test_replace(self):
# pylint: disable=no-member
dn = DN('t=0,t=1,t=2,t=3,t=4,t=5,t=6,t=7,t=8,t=9')
with pytest.raises(AttributeError):
dn.replace # pylint: disable=pointless-statement
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 test_hashing(self):
# create DN's that are equal but differ in case
dn1 = DN((self.attr1.lower(), self.value1.upper()))
dn2 = DN((self.attr1.upper(), self.value1.lower()))
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
# DNs that are equal should hash to the same value.
assert dn1 == dn2
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
# Good, everyone's equal, now verify their hash 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 hash(dn1) == hash(dn2)
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
# Different DN objects with the same value should
# map to 1 common key and 1 member in a set. The key and
# member are based on the object's 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
dn1_a = DN(self.dn1)
dn1_b = DN(self.dn1)
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
dn2_a = DN(self.dn2)
dn2_b = DN(self.dn2)
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
dn3_a = DN(self.dn3)
dn3_b = DN(self.dn3)
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 dn1_a == dn1_b
assert dn2_a == dn2_b
assert dn3_a == dn3_b
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
d = dict()
s = set()
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
d[dn1_a] = str(dn1_a)
d[dn1_b] = str(dn1_b)
d[dn2_a] = str(dn2_a)
d[dn2_b] = str(dn2_b)
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
s.add(dn1_a)
s.add(dn1_b)
s.add(dn2_a)
s.add(dn2_b)
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 len(d) == 2
assert len(s) == 2
assert sorted(d) == sorted([dn1_a, dn2_a])
assert sorted(s) == sorted([dn1_a, dn2_a])
assert dn1_a in d
assert dn1_b in d
assert dn2_a in d
assert dn2_b in d
assert dn3_a not in d
assert dn3_b not in d
assert dn1_a in s
assert dn1_b in s
assert dn2_a in s
assert dn2_b in s
assert dn3_a not in s
assert dn3_b not in s
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 test_x500_text(self):
# null DN x500 ordering and LDAP ordering are the same
nulldn = DN()
assert nulldn.ldap_text() == nulldn.x500_text()
# reverse a DN with a single RDN
assert self.dn1.ldap_text() == self.dn1.x500_text()
# reverse a DN with 2 RDNs
dn3_x500 = self.dn3.x500_text()
dn3_rev = DN(self.rdn2, self.rdn1)
assert dn3_rev.ldap_text() == dn3_x500
# reverse a longer DN
longdn_x500 = self.base_container_dn.x500_text()
longdn_rev = DN(longdn_x500)
l = len(self.base_container_dn)
for i in range(l):
assert longdn_rev[i] == self.base_container_dn[l - 1 - i]
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 TestEscapes:
@pytest.fixture(autouse=True)
def escapes_setup(self):
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.privilege = 'R,W privilege'
self.dn_str_hex_escape = 'cn=R\\2cW privilege,cn=privileges,cn=pbac,dc=idm,dc=lab,dc=bos,dc=redhat,dc=com'
self.dn_str_backslash_escape = 'cn=R\\,W privilege,cn=privileges,cn=pbac,dc=idm,dc=lab,dc=bos,dc=redhat,dc=com'
def test_escape(self):
dn = DN(self.dn_str_hex_escape)
assert dn['cn'] == self.privilege
assert dn[0].value == self.privilege
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 = DN(self.dn_str_backslash_escape)
assert dn['cn'] == self.privilege
assert dn[0].value == self.privilege
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 TestInternationalization:
@pytest.fixture(autouse=True)
def i18n_setup(self):
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
# Hello in Arabic
self.arabic_hello_utf8 = (b'\xd9\x85\xd9\x83\xd9\x8a\xd9\x84' +
b'\xd8\xb9\x20\xd9\x85\xd8\xa7\xd9' +
b'\x84\xd9\x91\xd8\xb3\xd9\x84\xd8\xa7')
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.arabic_hello_unicode = self.arabic_hello_utf8.decode('utf-8')
def assert_equal_utf8(self, obj, b):
if six.PY2:
assert str(obj) == b
else:
assert str(obj) == b.decode('utf-8')
@contextlib.contextmanager
def fail_py3(self, exception_type):
try:
yield
except exception_type:
if six.PY2:
raise
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 test_i18n(self):
actual = self.arabic_hello_unicode.encode('utf-8')
expected = self.arabic_hello_utf8
assert actual == expected
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
# AVA's
# test attr i18n
ava1 = AVA(self.arabic_hello_unicode, 'foo')
assert isinstance(ava1.attr, unicode)
assert isinstance(ava1.value, unicode)
assert ava1.attr == self.arabic_hello_unicode
self.assert_equal_utf8(ava1, self.arabic_hello_utf8 + b'=foo')
with self.fail_py3(TypeError):
ava1 = AVA(self.arabic_hello_utf8, 'foo')
if six.PY2:
assert isinstance(ava1.attr, unicode)
assert isinstance(ava1.value, unicode)
assert ava1.attr == self.arabic_hello_unicode
self.assert_equal_utf8(ava1, self.arabic_hello_utf8 + b'=foo')
# test value i18n
ava1 = AVA('cn', self.arabic_hello_unicode)
assert isinstance(ava1.attr, unicode)
assert isinstance(ava1.value, unicode)
assert ava1.value == self.arabic_hello_unicode
self.assert_equal_utf8(ava1, b'cn=' + self.arabic_hello_utf8)
with self.fail_py3(TypeError):
ava1 = AVA('cn', self.arabic_hello_utf8)
if six.PY2:
assert isinstance(ava1.attr, unicode)
assert isinstance(ava1.value, unicode)
assert ava1.value == self.arabic_hello_unicode
self.assert_equal_utf8(ava1, b'cn=' + self.arabic_hello_utf8)
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
# RDN's
# test attr i18n
rdn1 = RDN((self.arabic_hello_unicode, 'foo'))
assert isinstance(rdn1.attr, unicode)
assert isinstance(rdn1.value, unicode)
assert rdn1.attr == self.arabic_hello_unicode
self.assert_equal_utf8(rdn1, self.arabic_hello_utf8 + b'=foo')
with self.fail_py3(TypeError):
rdn1 = RDN((self.arabic_hello_utf8, 'foo'))
if six.PY2:
assert isinstance(rdn1.attr, unicode)
assert isinstance(rdn1.value, unicode)
assert rdn1.attr == self.arabic_hello_unicode
assert str(rdn1) == self.arabic_hello_utf8 + b'=foo'
# test value i18n
rdn1 = RDN(('cn', self.arabic_hello_unicode))
assert isinstance(rdn1.attr, unicode)
assert isinstance(rdn1.value, unicode)
assert rdn1.value == self.arabic_hello_unicode
self.assert_equal_utf8(rdn1, b'cn=' + self.arabic_hello_utf8)
with self.fail_py3(TypeError):
rdn1 = RDN(('cn', self.arabic_hello_utf8))
if six.PY2:
assert isinstance(rdn1.attr, unicode)
assert isinstance(rdn1.value, unicode)
assert rdn1.value == self.arabic_hello_unicode
assert str(rdn1) == b'cn=' + self.arabic_hello_utf8
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's
# test attr i18n
dn1 = DN((self.arabic_hello_unicode, 'foo'))
assert isinstance(dn1[0].attr, unicode)
assert isinstance(dn1[0].value, unicode)
assert dn1[0].attr == self.arabic_hello_unicode
self.assert_equal_utf8(dn1, self.arabic_hello_utf8 + b'=foo')
with self.fail_py3(TypeError):
dn1 = DN((self.arabic_hello_utf8, 'foo'))
if six.PY2:
assert isinstance(dn1[0].attr, unicode)
assert isinstance(dn1[0].value, unicode)
assert dn1[0].attr == self.arabic_hello_unicode
assert str(dn1) == self.arabic_hello_utf8 + b'=foo'
# test value i18n
dn1 = DN(('cn', self.arabic_hello_unicode))
assert isinstance(dn1[0].attr, unicode)
assert isinstance(dn1[0].value, unicode)
assert dn1[0].value == self.arabic_hello_unicode
self.assert_equal_utf8(dn1, b'cn=' + self.arabic_hello_utf8)
with self.fail_py3(TypeError):
dn1 = DN(('cn', self.arabic_hello_utf8))
if six.PY2:
assert isinstance(dn1[0].attr, unicode)
assert isinstance(dn1[0].value, unicode)
assert dn1[0].value == self.arabic_hello_unicode
assert str(dn1) == b'cn=' + self.arabic_hello_utf8
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
# 1: LDAP_AVA_STRING
# 4: LDAP_AVA_NONPRINTABLE
@pytest.mark.parametrize(
'dnstring,expected',
[
('', []),
('cn=bob', [[('cn', 'bob', 1)]]),
('cn=Bob', [[('cn', 'Bob', 1)]]),
(u'cn=b\xf6b', [[('cn', u'b\xf6b', 4)]]),
('cn=bob,sn=builder', [[('cn', 'bob', 1)], [('sn', 'builder', 1)]]),
(u'cn=b\xf6b,sn=builder', [
[('cn', u'b\xf6b', 4)], [('sn', 'builder', 1)]
]),
('cn=bob+sn=builder', [[('cn', 'bob', 1), ('sn', 'builder', 1)]]),
('dc=ipa,dc=example', [[('dc', 'ipa', 1)], [('dc', 'example', 1)]]),
('cn=R\\,W privilege', [[('cn', 'R,W privilege', 1)]]),
]
)
def test_str2dn2str(dnstring, expected):
dn = str2dn(dnstring)
assert dn == expected
assert dn2str(dn) == dnstring
assert dn_ctypes.str2dn(dnstring) == dn
assert dn_ctypes.dn2str(dn) == dnstring
@pytest.mark.parametrize(
'dnstring',
[
'cn',
'cn=foo,',
'cn=foo+bar',
]
)
def test_str2dn_errors(dnstring):
with pytest.raises(DECODING_ERROR):
str2dn(dnstring)
with pytest.raises(dn_ctypes.DECODING_ERROR):
dn_ctypes.str2dn(dnstring)
def test_dn2str_special():
dnstring = 'cn=R\\2cW privilege'
dnstring2 = 'cn=R\\,W privilege'
expected = [[('cn', 'R,W privilege', 1)]]
dn = str2dn(dnstring)
assert dn == expected
assert dn2str(dn) == dnstring2
assert dn_ctypes.str2dn(dnstring) == dn
assert dn_ctypes.dn2str(dn) == dnstring2