IntenseWebs/freeipa
3
0
Fork 0
mirror of https://salsa.debian.org/freeipa-team/freeipa.git synced 2025-02-25 18:55:28 -06:00
Code Issues Packages Projects Releases Wiki Activity
2bf68115ce
freeipa/ipalib/plugins/dns.py

2892 lines
96 KiB
Python
Raw Normal View History

Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
# Authors:
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
# Martin Kosek <mkosek@redhat.com>
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
# Pavel Zuna <pzuna@redhat.com>
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
#
# Copyright (C) 2010 Red Hat
# see file 'COPYING' for use and warranty information
#
Change FreeIPA license to GPLv3+ The changes include: * Change license blobs in source files to mention GPLv3+ not GPLv2 only * Add GPLv3+ license text * Package COPYING not LICENSE as the license blobs (even the old ones) mention COPYING specifically, it is also more common, I think https://fedorahosted.org/freeipa/ticket/239
2010-12-09 06:59:11 -06:00
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
Change FreeIPA license to GPLv3+ The changes include: * Change license blobs in source files to mention GPLv3+ not GPLv2 only * Add GPLv3+ license text * Package COPYING not LICENSE as the license blobs (even the old ones) mention COPYING specifically, it is also more common, I think https://fedorahosted.org/freeipa/ticket/239
2010-12-09 06:59:11 -06:00
# along with this program. If not, see <http://www.gnu.org/licenses/>.
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
Allow relative DNS name in NS validator Precallback validator was failing when a zone-relative name was used as a NS record (for example record "ns" in a zone "example.com"). However, this is valid in BIND and we should allow it as well. Imports in dns module had to be switched to absolute imports (available from Python 2.5) to deal with a conflict of IPA dns module and dnspython module. https://fedorahosted.org/freeipa/ticket/2630
2012-05-29 08:06:31 -05:00
from __future__ import absolute_import
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
import netaddr
import time
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
import re
Allow relative DNS name in NS validator Precallback validator was failing when a zone-relative name was used as a NS record (for example record "ns" in a zone "example.com"). However, this is valid in BIND and we should allow it as well. Imports in dns module had to be switched to absolute imports (available from Python 2.5) to deal with a conflict of IPA dns module and dnspython module. https://fedorahosted.org/freeipa/ticket/2630
2012-05-29 08:06:31 -05:00
import dns.name
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
from ipalib.request import context
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
from ipalib import api, errors, output
from ipalib import Command
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
from ipalib.parameters import Flag, Bool, Int, Decimal, Str, StrEnum, Any
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
from ipalib.plugins.baseldap import *
from ipalib import _, ngettext
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
from ipalib.util import (validate_zonemgr, normalize_zonemgr,
Fill new DNS zone update policy by default For security reasons, dynamic updates are not enabled for new DNS zones. In order to enable the dynamic zone securely, user needs to allow dynamic updates and create a zone update policy. The policy is not easy to construct for regular users, we should rather fill it by default and let users just switch the policy on or off. https://fedorahosted.org/freeipa/ticket/2441
2012-06-04 10:53:34 -05:00
validate_hostname, validate_dns_label, validate_domain_name,
get_dns_forward_zone_update_policy, get_dns_reverse_zone_update_policy)
Replace DNS client based on acutil with python-dns IPA client and server tool set used authconfig acutil module to for client DNS operations. This is not optimal DNS interface for several reasons: - does not provide native Python object oriented interface but but rather C-like interface based on functions and structures which is not easy to use and extend - acutil is not meant to be used by third parties besides authconfig and thus can break without notice Replace the acutil with python-dns package which has a feature rich interface for dealing with all different aspects of DNS including DNSSEC. The main target of this patch is to replace all uses of acutil DNS library with a use python-dns. In most cases, even though the larger parts of the code are changed, the actual functionality is changed only in the following cases: - redundant DNS checks were removed from verify_fqdn function in installutils to make the whole DNS check simpler and less error-prone. Logging was improves for the remaining checks - improved logging for ipa-client-install DNS discovery https://fedorahosted.org/freeipa/ticket/2730 https://fedorahosted.org/freeipa/ticket/1837
2012-05-11 07:38:09 -05:00
from ipapython.ipautil import valid_ip, CheckedIPAddress, is_host_resolvable
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _("""
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Domain Name System (DNS)
Manage DNS zone and resource records.
Add help for new structured DNS framework DNS Test Day shown that the new RR specific DNS options and the concepts behind them may not be easily understood. This patch adds an explanation of the new DNS framework for structured options to make it easier for the user to understand and use the new options. https://fedorahosted.org/freeipa/ticket/2382
2012-03-01 07:05:50 -06:00
USING STRUCTURED PER-TYPE OPTIONS
There are many structured DNS RR types where DNS data stored in LDAP server
is not just a scalar value, for example an IP address or a domain name, but
a data structure which may be often complex. A good example is a LOC record
[RFC1876] which consists of many mandatory and optional parts (degrees,
minutes, seconds of latitude and longitude, altitude or precision).
It may be difficult to manipulate such DNS records without making a mistake
and entering an invalid value. DNS module provides an abstraction over these
raw records and allows to manipulate each RR type with specific options. For
each supported RR type, DNS module provides a standard option to manipulate
a raw records with format --<rrtype>-rec, e.g. --mx-rec, and special options
for every part of the RR structure with format --<rrtype>-<partname>, e.g.
--mx-preference and --mx-exchanger.
When adding a record, either RR specific options or standard option for a raw
value can be used, they just should not be combined in one add operation. When
modifying an existing entry, new RR specific options can be used to change
one part of a DNS record, where the standard option for raw value is used
to specify the modified value. The following example demonstrates
a modification of MX record preference form 0 to 1 in a record without
modifying the exchanger:
ipa dnsrecord-mod --mx-rec="0 mx.example.com." --mx-preference=1
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
EXAMPLES:
Add new zone:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnszone-add example.com --name-server=nameserver.example.com \\
--admin-email=admin@example.com
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Per-domain DNS record permissions IPA implements read/write permissions for DNS record or zones. Provided set of permissions and privileges can, however, only grant access to the whole DNS tree, which may not be appropriate. Administrators may miss more fine-grained permissions allowing them to delegate access per-zone. Create a new IPA auxiliary objectclass ipaDNSZone allowing a managedBy attribute for a DNS zone. This attribute will hold a group DN (in this case a permission) which allows its members to read or write in a zone. Member permissions in given zone will only have 2 limitations: 1) Members cannot delete the zone 2) Members cannot edit managedBy attribute Current DNS deny ACI used to enforce read access is removed so that DNS privileges are based on allow ACIs only, which is much more flexible approach as deny ACIs have always precedence and limit other extensions. Per-zone access is allowed in 3 generic ACIs placed in cn=dns,$SUFFIX so that no special ACIs has to be added to DNS zones itselves. 2 new commands have been added which allows an administrator to create the system permission allowing the per-zone access and fill a zone's managedBy attribute: * dnszone-add-permission: Add per-zone permission * dnszone-remove-permission: Remove per-zone permission https://fedorahosted.org/freeipa/ticket/2511
2012-06-27 06:10:10 -05:00
Add system permission that can be used for per-zone privilege delegation:
ipa dnszone-add-permission example.com
Fix DNS zone --allow-dynupdate option behavior --allow-dynupdate was implemented as a Flag parameter type, which is not convenient for LDAP attributes. When a DNS zone with permitted dynamic updates was modified and the --allow-dynupdate flag was not set, dynamic updates were turned off. This patch changes the option type to Bool parameter type which behaves according to user expectations when modifying the zone. https://fedorahosted.org/freeipa/ticket/2039
2011-11-09 07:23:15 -06:00
Modify the zone to allow dynamic updates for hosts own records in realm EXAMPLE.COM:
Fill new DNS zone update policy by default For security reasons, dynamic updates are not enabled for new DNS zones. In order to enable the dynamic zone securely, user needs to allow dynamic updates and create a zone update policy. The policy is not easy to construct for regular users, we should rather fill it by default and let users just switch the policy on or off. https://fedorahosted.org/freeipa/ticket/2441
2012-06-04 10:53:34 -05:00
ipa dnszone-mod example.com --dynamic-update=TRUE
This is the equivalent of:
ipa dnszone-mod example.com --dynamic-update=TRUE \\
--update-policy="grant EXAMPLE.COM krb5-self * A; grant EXAMPLE.COM krb5-self * AAAA; grant EXAMPLE.COM krb5-self * SSHFP;"
Fix DNS zone --allow-dynupdate option behavior --allow-dynupdate was implemented as a Flag parameter type, which is not convenient for LDAP attributes. When a DNS zone with permitted dynamic updates was modified and the --allow-dynupdate flag was not set, dynamic updates were turned off. This patch changes the option type to Bool parameter type which behaves according to user expectations when modifying the zone. https://fedorahosted.org/freeipa/ticket/2039
2011-11-09 07:23:15 -06:00
Query and transfer ACLs for DNS zones Provide a way to specify BIND allow-query and allow-transfer ACLs for DNS zones. IMPORTANT: new bind-dyndb-ldap adds a zone transfer ability. To avoid zone information leaks to unintended places, allow-transfer ACL for every zone is by default set to none and has to be explicitly enabled by an Administrator. This is done both for new DNS zones and old DNS zones during RPM update via new DNS upgrade plugin. https://fedorahosted.org/freeipa/ticket/1211
2012-02-24 02:30:39 -06:00
Modify the zone to allow zone transfers for local network only:
ipa dnszone-mod example.com --allow-transfer=10.0.0.0/8
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
Add new reverse zone specified by network IP address:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnszone-add --name-from-ip=80.142.15.0/24 \\
--name-server=nameserver.example.com
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
Fixes for the DNS plugin https://fedorahosted.org/freeipa/ticket/730
2011-01-07 03:29:16 -06:00
Add second nameserver for example.com:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnsrecord-add example.com @ --ns-rec=nameserver2.example.com
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Validate MX records https://fedorahosted.org/freeipa/ticket/967
2011-02-15 03:40:27 -06:00
Add a mail server for example.com:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnsrecord-add example.com @ --mx-rec="10 mail1"
Add another record using MX record specific options:
ipa dnsrecord-add example.com @ --mx-preference=20 --mx-exchanger=mail2
Add another record using interactive mode (started when dnsrecord-add, dnsrecord-mod,
or dnsrecord-del are executed with no options):
ipa dnsrecord-add example.com @
Please choose a type of DNS resource record to be added
The most common types for this type of zone are: NS, MX, LOC
DNS resource record type: MX
MX Preference: 30
MX Exchanger: mail3
Record name: example.com
MX record: 10 mail1, 20 mail2, 30 mail3
NS record: nameserver.example.com., nameserver2.example.com.
Validate MX records https://fedorahosted.org/freeipa/ticket/967
2011-02-15 03:40:27 -06:00
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Delete previously added nameserver from example.com:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnsrecord-del example.com @ --ns-rec=nameserver2.example.com.
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Add LOC record for example.com:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnsrecord-add example.com @ --loc-rec="49 11 42.4 N 16 36 29.6 E 227.64m"
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
Add new A record for www.example.com. Create a reverse record in appropriate
reverse zone as well. In this case a PTR record "2" pointing to www.example.com.
will be created in zone 15.142.80.in-addr.arpa.
ipa dnsrecord-add example.com www --a-rec=80.142.15.2 --a-create-reverse
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Add new PTR record for www.example.com
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnsrecord-add 15.142.80.in-addr.arpa. 2 --ptr-rec=www.example.com.
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Add example of DNS SRV record and a simple validator https://fedorahosted.org/freeipa/ticket/846
2011-01-26 08:31:50 -06:00
Add new SRV records for LDAP servers. Three quarters of the requests
should go to fast.example.com, one quarter to slow.example.com. If neither
is available, switch to backup.example.com.
ipa dnsrecord-add example.com _ldap._tcp --srv-rec="0 3 389 fast.example.com"
ipa dnsrecord-add example.com _ldap._tcp --srv-rec="0 1 389 slow.example.com"
ipa dnsrecord-add example.com _ldap._tcp --srv-rec="1 1 389 backup.example.com"
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
The interactive mode can be used for easy modification:
ipa dnsrecord-mod example.com _ldap._tcp
No option to modify specific record provided.
Current DNS record contents:
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
SRV record: 0 3 389 fast.example.com, 0 1 389 slow.example.com, 1 1 389 backup.example.com
Enforce sizelimit in permission-find, post_callback returns truncated We actually perform two searches in permission-find. The first looks for matches within the permission object itself. The second looks at matches in the underlying aci. We need to break out in two places. The first is if we find enough matches in the permission itself. The second when we are appending matches from acis. The post_callback() definition needed to be modified to return the truncated value so a plugin author can modify that value. https://fedorahosted.org/freeipa/ticket/2322
2012-05-23 10:00:24 -05:00
Modify SRV record '0 3 389 fast.example.com'? Yes/No (default No):
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Modify SRV record '0 1 389 slow.example.com'? Yes/No (default No): y
SRV Priority [0]: (keep the default value)
SRV Weight [1]: 2 (modified value)
SRV Port [389]: (keep the default value)
SRV Target [slow.example.com]: (keep the default value)
1 SRV record skipped. Only one value per DNS record type can be modified at one time.
Record name: _ldap._tcp
SRV record: 0 3 389 fast.example.com, 1 1 389 backup.example.com, 0 2 389 slow.example.com
After this modification, three fifths of the requests should go to
fast.example.com and two fifths to slow.example.com.
An example of the interactive mode for dnsrecord-del command:
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
ipa dnsrecord-del example.com www
No option to delete specific record provided.
Delete all? Yes/No (default No): (do not delete all records)
Current DNS record contents:
A record: 1.2.3.4, 11.22.33.44
ticket 1659 - invalid i18n string in dns.py dns.py at line 976 has an invalid i18n string and cannot be processed during message extraction causing message catalog generation to fail. The format parameters are trapped inside the i18n string. Also it's not necessary to promote the i18n string literal to unicode via the u prefix because the _() function returns unicode.
2011-08-17 17:51:38 -05:00
Delete A record '1.2.3.4'? Yes/No (default No):
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Delete A record '11.22.33.44'? Yes/No (default No): y
Record name: www
A record: 1.2.3.4 (A record 11.22.33.44 has been deleted)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Show zone example.com:
ipa dnszone-show example.com
Fix style and grammatical issues in built-in command help. There is a rather large API.txt change but it is only due to changes in the doc string in parameters. ticket 729
2011-03-04 10:08:54 -06:00
Find zone with "example" in its domain name:
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
ipa dnszone-find example
Find records for resources with "www" in their name in zone example.com:
ipa dnsrecord-find example.com www
Find A records with value 10.10.0.1 in zone example.com
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
ipa dnsrecord-find example.com --a-rec=10.10.0.1
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Show records for resource www in zone example.com
ipa dnsrecord-show example.com www
Add DNS conditional forwarding Add ability configure per-zone forwarder for DNS zones. Any data in such zone will then be considered as non-authoritative and all queries will be sent to specified forwarder. https://fedorahosted.org/freeipa/ticket/2108
2012-02-14 04:10:22 -06:00
Forward all request for a zone external.com to another nameserver using
a "first" policy (it will send the queries to the selected forwarder and if
not answered it will use global resolvers):
ipa dnszone-add external.com
ipa dnszone-mod external.com --forwarder=10.20.0.1 \\
--forward-policy=first
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Delete zone example.com with all resource records:
ipa dnszone-del example.com
Resolve a host name to see if it exists (will add default IPA domain
if one is not included):
ipa dns-resolve www.example.com
ipa dns-resolve www
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
Improve user awareness about dnsconfig Global DNS configuration is a nice tool to maintain a common DNS settings stored in LDAP which are then used for all enrolled IPA servers. However, the settings stored in LDAP override local settings in named.conf on DNS servers. This patch adds more information about global DNS configuration options in install scripts and DNS module help. https://fedorahosted.org/freeipa/ticket/2525
2012-03-15 07:51:59 -05:00
GLOBAL DNS CONFIGURATION
DNS configuration passed to command line install script is stored in a local
configuration file on each IPA server where DNS service is configured. These
local settings can be overridden with a common configuration stored in LDAP
server:
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
Show global DNS configuration:
ipa dnsconfig-show
Modify global DNS configuration and set a list of global forwarders:
ipa dnsconfig-mod --forwarder=10.0.0.1
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
""")
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
# supported resource record types
_record_types = (
u'A', u'AAAA', u'A6', u'AFSDB', u'APL', u'CERT', u'CNAME', u'DHCID', u'DLV',
Remove obsolete record types from DNS https://fedorahosted.org/freeipa/ticket/923
2011-02-10 14:17:21 -06:00
u'DNAME', u'DNSKEY', u'DS', u'HIP', u'IPSECKEY', u'KEY', u'KX', u'LOC',
u'MX', u'NAPTR', u'NS', u'NSEC', u'NSEC3', u'NSEC3PARAM', u'PTR',
u'RRSIG', u'RP', u'SIG', u'SPF', u'SRV', u'SSHFP', u'TA', u'TKEY',
u'TSIG', u'TXT',
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
# DNS zone record identificator
_dns_zone_record = u'@'
# most used record types, always ask for those in interactive prompt
_top_record_types = ('A', 'AAAA', )
_rev_top_record_types = ('PTR', )
_zone_top_record_types = ('NS', 'MX', 'LOC', )
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
# attributes derived from record types
_record_attributes = [str('%srecord' % t.lower()) for t in _record_types]
# supported DNS classes, IN = internet, rest is almost never used
_record_classes = (u'IN', u'CS', u'CH', u'HS')
Add --zonemgr/--admin-mail validator Do at least a basic validation of DNS zone manager mail address. Do not require '@' to be in the mail address as the SOA record stores this value without it and people may be used to configure it that way. '@' is always removed by the installer/dns plugin before the DNS zone is created. https://fedorahosted.org/freeipa/ticket/1966
2011-10-24 11:35:48 -05:00
def _rname_validator(ugettext, zonemgr):
try:
validate_zonemgr(zonemgr)
except ValueError, e:
return unicode(e)
return None
Change parameters to use only default_from for dynamic default values. Replace all occurences of create_default with equivalent default_from and remove create_default from the framework. This is needed for proper parameter validation, as there is no way to tell which parameters to validate prior to calling create_default, because create_default does not provide information about which parameters are used for generating the default value.
2012-03-15 03:32:37 -05:00
def _create_zone_serial():
Enable SOA serial autoincrement SOA serial autoincrement is a requirement for major DNS features, e.g. zone transfers or DNSSEC. Enable it by default in named.conf both for new and upgraded installations. Name of the bind-dyndb-ldap option is "serial_autoincrement". From now on, idnsSOAserial attribute also has to be put to replication agreement exclude list as serial will be incremented on each DNS server separately and won't be shared. Exclude list has to be updated both for new replication agreements and the current ones. Minimum number of connections for bind-dyndb-ldap has been rised to 4 connections, the setting will be updated during package upgrade. https://fedorahosted.org/freeipa/ticket/2554
2012-06-28 09:46:48 -05:00
"""
Generate serial number for zones. bind-dyndb-ldap expects unix time in
to be used for SOA serial.
SOA serial in a date format would also work, but it may be set to far
future when many DNS updates are done per day (more than 100). Unix
timestamp is more resilient to this issue.
"""
return int(time.time())
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
def _reverse_zone_name(netstr):
net = netaddr.IPNetwork(netstr)
items = net.ip.reverse_dns.split('.')
if net.version == 4:
return u'.'.join(items[4 - net.prefixlen / 8:])
elif net.version == 6:
return u'.'.join(items[32 - net.prefixlen / 4:])
else:
return None
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _validate_ipaddr(ugettext, ipaddr, ip_version=None):
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
try:
Use stricter semantics when checking IP address for DNS records https://fedorahosted.org/freeipa/ticket/2379
2012-02-15 04:11:24 -06:00
ip = netaddr.IPAddress(ipaddr, flags=netaddr.INET_PTON)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if ip_version is not None:
if ip.version != ip_version:
return _('invalid IP address version (is %(value)d, must be %(required_value)d)!') \
% dict(value=ip.version, required_value=ip_version)
Prevent stacktrace when DNS AAAA record is added This patch fixes a stacktrace that is printed out when a IPv6 AAAA record with subnet prefix length (e.g. /64) is added. The same error message as when IPv4 record with subnet prefix length is used. https://fedorahosted.org/freeipa/ticket/1115
2011-03-22 05:00:06 -05:00
except (netaddr.AddrFormatError, ValueError):
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
return _('invalid IP address format')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return None
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _validate_ip4addr(ugettext, ipaddr):
return _validate_ipaddr(ugettext, ipaddr, 4)
def _validate_ip6addr(ugettext, ipaddr):
return _validate_ipaddr(ugettext, ipaddr, 6)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
def _validate_ipnet(ugettext, ipnet):
try:
net = netaddr.IPNetwork(ipnet)
Validate name_from_ip parameter of dnszone. ticket 1627
2011-09-26 04:00:32 -05:00
except (netaddr.AddrFormatError, ValueError, UnboundLocalError):
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
return _('invalid IP network format')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return None
Query and transfer ACLs for DNS zones Provide a way to specify BIND allow-query and allow-transfer ACLs for DNS zones. IMPORTANT: new bind-dyndb-ldap adds a zone transfer ability. To avoid zone information leaks to unintended places, allow-transfer ACL for every zone is by default set to none and has to be explicitly enabled by an Administrator. This is done both for new DNS zones and old DNS zones during RPM update via new DNS upgrade plugin. https://fedorahosted.org/freeipa/ticket/1211
2012-02-24 02:30:39 -06:00
def _validate_bind_aci(ugettext, bind_acis):
if not bind_acis:
return
bind_acis = bind_acis.split(';')
if bind_acis[-1]:
return _('each ACL element must be terminated with a semicolon')
else:
bind_acis.pop(-1)
for bind_aci in bind_acis:
if bind_aci in ("any", "none"):
continue
if bind_aci in ("localhost", "localnets"):
return _('ACL name "%s" is not supported') % bind_aci
if bind_aci.startswith('!'):
bind_aci = bind_aci[1:]
try:
ip = CheckedIPAddress(bind_aci, parse_netmask=True,
allow_network=True)
except (netaddr.AddrFormatError, ValueError), e:
return unicode(e)
except UnboundLocalError:
return _(u"invalid address format")
def _normalize_bind_aci(bind_acis):
if not bind_acis:
return
bind_acis = bind_acis.split(';')
normalized = []
for bind_aci in bind_acis:
if not bind_aci:
continue
if bind_aci in ("any", "none", "localhost", "localnets"):
normalized.append(bind_aci)
continue
prefix = ""
if bind_aci.startswith('!'):
bind_aci = bind_aci[1:]
prefix = "!"
try:
ip = CheckedIPAddress(bind_aci, parse_netmask=True,
allow_network=True)
if '/' in bind_aci: # addr with netmask
netmask = "/%s" % ip.prefixlen
else:
netmask = ""
normalized.append(u"%s%s%s" % (prefix, str(ip), netmask))
continue
except:
normalized.append(bind_aci)
continue
acis = u';'.join(normalized)
acis += u';'
return acis
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
def _bind_hostname_validator(ugettext, value):
Add example of DNS SRV record and a simple validator https://fedorahosted.org/freeipa/ticket/846
2011-01-26 08:31:50 -06:00
try:
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
# Allow domain name which is not fully qualified. These are supported
# in bind and then translated as <non-fqdn-name>.<domain>.
validate_hostname(value, check_fqdn=False)
except ValueError, e:
return _('invalid domain-name: %s') \
% unicode(e)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
return None
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
def _dns_record_name_validator(ugettext, value):
if value == _dns_zone_record:
return
try:
map(lambda label:validate_dns_label(label, allow_underscore=True), \
value.split(u'.'))
except ValueError, e:
return unicode(e)
Allow port numbers for idnsForwarders Let user enter custom ports for zone conditional forwarders or global forwarders in dnsconfig. Ports can be specified in a standard BIND format: IP_ADDRESS [port PORT] https://fedorahosted.org/freeipa/ticket/2462
2012-03-20 09:35:54 -05:00
def _validate_bind_forwarder(ugettext, forwarder):
ip_address, sep, port = forwarder.partition(u' port ')
ip_address_validation = _validate_ipaddr(ugettext, ip_address)
if ip_address_validation is not None:
return ip_address_validation
if sep:
try:
port = int(port)
if port < 0 or port > 65535:
raise ValueError()
except ValueError:
return _('%(port)s is not a valid port' % dict(port=port))
return None
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
def _domain_name_validator(ugettext, value):
try:
validate_domain_name(value)
except ValueError, e:
return unicode(e)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _hostname_validator(ugettext, value):
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
try:
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
validate_hostname(value)
except ValueError, e:
return _('invalid domain-name: %s') \
% unicode(e)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
return None
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _normalize_hostname(domain_name):
"""Make it fully-qualified"""
if domain_name[-1] != '.':
return domain_name + '.'
else:
return domain_name
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
def is_forward_record(zone, str_address):
addr = netaddr.IPAddress(str_address)
if addr.version == 4:
result = api.Command['dnsrecord_find'](zone, arecord=str_address)
elif addr.version == 6:
result = api.Command['dnsrecord_find'](zone, aaaarecord=str_address)
else:
raise ValueError('Invalid address family')
return result['count'] > 0
def add_forward_record(zone, name, str_address):
addr = netaddr.IPAddress(str_address)
try:
if addr.version == 4:
api.Command['dnsrecord_add'](zone, name, arecord=str_address)
elif addr.version == 6:
api.Command['dnsrecord_add'](zone, name, aaaarecord=str_address)
else:
raise ValueError('Invalid address family')
except errors.EmptyModlist:
pass # the entry already exists and matches
def get_reverse_zone(ipaddr, prefixlen=None):
ip = netaddr.IPAddress(ipaddr)
revdns = unicode(ip.reverse_dns)
if prefixlen is None:
revzone = u''
result = api.Command['dnszone_find']()['result']
for zone in result:
zonename = zone['idnsname'][0]
if revdns.endswith(zonename) and len(zonename) > len(revzone):
revzone = zonename
else:
if ip.version == 4:
pos = 4 - prefixlen / 8
elif ip.version == 6:
pos = 32 - prefixlen / 4
items = ip.reverse_dns.split('.')
revzone = u'.'.join(items[pos:])
try:
api.Command['dnszone_show'](revzone)
except errors.NotFound:
revzone = u''
if len(revzone) == 0:
raise errors.NotFound(
reason=_('DNS reverse zone for IP address %(addr)s not found') % dict(addr=ipaddr)
)
revname = revdns[:-len(revzone)-1]
return revzone, revname
def add_records_for_host_validation(option_name, host, domain, ip_addresses, check_forward=True, check_reverse=True):
Improve FQDN handling in DNS and host plugins DNS and host plugin does not work well with domain names ending with dot. host plugin creates a record with two fqdn attributes when such hostname is created which then has to be manually fixed. DNS plugin handled zones with and without trailing dot as two distinct zones, which may lead to issues when both zones are created. This patch sanitizes approach to FQDNs in both DNS and host plugin. Hostnames are now always normalized to the form without trailing dot as this form did not work before and it would keep hostname form consistent without changes in our server/client enrollment process. As DNS zones always worked in both forms this patch rather makes sure that the plugin works with both forms of one zone and prevents creating 2 identical zones with just different format. https://fedorahosted.org/freeipa/ticket/2420
2012-02-28 02:13:13 -06:00
try:
api.Command['dnszone_show'](domain)['result']
except errors.NotFound:
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
raise errors.NotFound(
reason=_('DNS zone %(zone)s not found') % dict(zone=domain)
)
if not isinstance(ip_addresses, (tuple, list)):
ip_addresses = [ip_addresses]
for ip_address in ip_addresses:
try:
ip = CheckedIPAddress(ip_address, match_local=False)
except Exception, e:
raise errors.ValidationError(name=option_name, error=unicode(e))
if check_forward:
if is_forward_record(domain, unicode(ip)):
raise errors.DuplicateEntry(
message=_(u'IP address %(ip)s is already assigned in domain %(domain)s.')\
% dict(ip=str(ip), domain=domain))
if check_reverse:
try:
prefixlen = None
if not ip.defaultnet:
prefixlen = ip.prefixlen
# we prefer lookup of the IP through the reverse zone
revzone, revname = get_reverse_zone(ip, prefixlen)
reverse = api.Command['dnsrecord_find'](revzone, idnsname=revname)
if reverse['count'] > 0:
raise errors.DuplicateEntry(
message=_(u'Reverse record for IP address %(ip)s already exists in reverse zone %(zone)s.')\
% dict(ip=str(ip), zone=revzone))
except errors.NotFound:
pass
def add_records_for_host(host, domain, ip_addresses, add_forward=True, add_reverse=True):
if not isinstance(ip_addresses, (tuple, list)):
ip_addresses = [ip_addresses]
for ip_address in ip_addresses:
ip = CheckedIPAddress(ip_address, match_local=False)
if add_forward:
add_forward_record(domain, host, unicode(ip))
if add_reverse:
try:
prefixlen = None
if not ip.defaultnet:
prefixlen = ip.prefixlen
revzone, revname = get_reverse_zone(ip, prefixlen)
addkw = { 'ptrrecord' : host + "." + domain }
api.Command['dnsrecord_add'](revzone, revname, **addkw)
except errors.EmptyModlist:
# the entry already exists and matches
pass
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class DNSRecord(Str):
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
# a list of parts that create the actual raw DNS record
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
parts = None
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
# an optional list of parameters used in record-specific operations
extra = None
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
supported = True
# supported RR types: https://fedorahosted.org/bind-dyndb-ldap/browser/doc/schema
label_format = _("%s record")
part_label_format = "%s %s"
doc_format = _('Comma-separated list of raw %s records')
option_group_format = _('%s Record')
see_rfc_msg = _("(see RFC %s for details)")
part_name_format = "%s_part_%s"
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
extra_name_format = "%s_extra_%s"
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
cli_name_format = "%s_%s"
format_error_msg = None
kwargs = Str.kwargs + (
('validatedns', bool, True),
('normalizedns', bool, True),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
# should be replaced in subclasses
rrtype = None
rfc = None
def __init__(self, name=None, *rules, **kw):
if self.rrtype not in _record_types:
raise ValueError("Unknown RR type: %s. Must be one of %s" % \
(str(self.rrtype), ", ".join(_record_types)))
if not name:
name = "%srecord*" % self.rrtype.lower()
kw.setdefault('cli_name', '%s_rec' % self.rrtype.lower())
kw.setdefault('label', self.label_format % self.rrtype)
kw.setdefault('doc', self.doc_format % self.rrtype)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
kw.setdefault('option_group', self.option_group_format % self.rrtype)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
kw['csv'] = True
if not self.supported:
kw['flags'] = ('no_option',)
super(DNSRecord, self).__init__(name, *rules, **kw)
def _get_part_values(self, value):
values = value.split()
if len(values) != len(self.parts):
return None
return tuple(values)
Fix NSEC record conversion NSEC record needs special treatment as it is not composed from a fixed set of DNS parts divided by space, but it contains a multivalued DNS part "types" containing a list of RR types it covers. There was already a special method for parsing raw NSEC record to DNS parts, but the other direction was missing. This patch adds special NSEC convertor to fix this issue. https://fedorahosted.org/freeipa/ticket/2307
2012-02-03 03:41:21 -06:00
def _part_values_to_string(self, values, index):
self._validate_parts(values)
return u" ".join(super(DNSRecord, self)._convert_scalar(v, index) \
for v in values if v is not None)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def get_parts_from_kw(self, kw, raise_on_none=True):
part_names = tuple(self.part_name_format % (self.rrtype.lower(), part.name) \
for part in self.parts)
vals = tuple(kw.get(part_name) for part_name in part_names)
if all(val is None for val in vals):
return
if raise_on_none:
for val_id,val in enumerate(vals):
if val is None and self.parts[val_id].required:
cli_name = self.cli_name_format % (self.rrtype.lower(), self.parts[val_id].name)
raise errors.ConversionError(name=self.name,
error=_("'%s' is a required part of DNS record") % cli_name)
return vals
def _validate_parts(self, parts):
if len(parts) != len(self.parts):
raise errors.ValidationError(name=self.name,
error=_("Invalid number of parts!"))
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _convert_scalar(self, value, index=None):
if isinstance(value, (tuple, list)):
Fix NSEC record conversion NSEC record needs special treatment as it is not composed from a fixed set of DNS parts divided by space, but it contains a multivalued DNS part "types" containing a list of RR types it covers. There was already a special method for parsing raw NSEC record to DNS parts, but the other direction was missing. This patch adds special NSEC convertor to fix this issue. https://fedorahosted.org/freeipa/ticket/2307
2012-02-03 03:41:21 -06:00
return self._part_values_to_string(value, index)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return super(DNSRecord, self)._convert_scalar(value, index)
def normalize(self, value):
if self.normalizedns: #pylint: disable=E1101
if isinstance(value, (tuple, list)):
value = tuple(
self._normalize_parts(v) for v in value \
if v is not None
)
elif value is not None:
value = (self._normalize_parts(value),)
return super(DNSRecord, self).normalize(value)
def _normalize_parts(self, value):
"""
Normalize a DNS record value using normalizers for its parts.
"""
if self.parts is None:
return value
try:
values = self._get_part_values(value)
if not values:
return value
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
converted_values = [ part._convert_scalar(values[part_id]) \
if values[part_id] is not None else None
for part_id, part in enumerate(self.parts)
]
new_values = [ part.normalize(converted_values[part_id]) \
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
for part_id, part in enumerate(self.parts) ]
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
value = self._convert_scalar(new_values)
except Exception:
# cannot normalize, rather return original value than fail
pass
return value
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _rule_validatedns(self, _, value):
if not self.validatedns: #pylint: disable=E1101
return
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if value is None:
return
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
if value is None:
return
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if not self.supported:
return _('DNS RR type "%s" is not supported by bind-dyndb-ldap plugin') \
% self.rrtype
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if self.parts is None:
return
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
# validate record format
values = self._get_part_values(value)
if not values:
if not self.format_error_msg:
part_names = [part.name.upper() for part in self.parts]
if self.rfc:
see_rfc_msg = " " + self.see_rfc_msg % self.rfc
else:
see_rfc_msg = ""
return _('format must be specified as "%(format)s" %(rfcs)s') \
% dict(format=" ".join(part_names), rfcs=see_rfc_msg)
else:
return self.format_error_msg
# validate every part
for part_id, part in enumerate(self.parts):
val = part.normalize(values[part_id])
val = part.convert(val)
part.validate(val)
return None
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
def _convert_dnsrecord_part(self, part):
"""
All parts of DNSRecord need to be processed and modified before they
can be added to global DNS API. For example a prefix need to be added
before part name so that the name is unique in the global namespace.
"""
name = self.part_name_format % (self.rrtype.lower(), part.name)
cli_name = self.cli_name_format % (self.rrtype.lower(), part.name)
label = self.part_label_format % (self.rrtype, unicode(part.label))
option_group = self.option_group_format % self.rrtype
flags = list(part.flags) + ['dnsrecord_part', 'virtual_attribute',]
if not part.required:
flags.append('dnsrecord_optional')
return part.clone_rename(name,
cli_name=cli_name,
label=label,
required=False,
option_group=option_group,
flags=flags,
hint=self.name,) # name of parent RR param
def _convert_dnsrecord_extra(self, extra):
"""
Parameters for special per-type behavior need to be processed in the
same way as record parts in _convert_dnsrecord_part().
"""
name = self.extra_name_format % (self.rrtype.lower(), extra.name)
cli_name = self.cli_name_format % (self.rrtype.lower(), extra.name)
label = self.part_label_format % (self.rrtype, unicode(extra.label))
option_group = self.option_group_format % self.rrtype
flags = list(extra.flags) + ['dnsrecord_extra', 'virtual_attribute',]
return extra.clone_rename(name,
cli_name=cli_name,
label=label,
required=False,
option_group=option_group,
flags=flags,
hint=self.name,) # name of parent RR param
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def get_parts(self):
if self.parts is None:
return tuple()
Mark optional DNS record parts All DNS record part options in dnsrecord commands need to be optional so that all of them are not required in every dnsrecord command. However, FreeIPA API then does not include an information which DNS record part options are optional in term of creating a new DNS record. For example, LOC record option "latitude seconds" is not needed to add a new LOC record. This patch adds a flag "dnsrecord_optional" to all such options so that this information is available for any other UI reading the FreeIPA API. https://fedorahosted.org/freeipa/ticket/2208
2012-01-19 05:14:41 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
return tuple(self._convert_dnsrecord_part(part) for part in self.parts)
Mark optional DNS record parts All DNS record part options in dnsrecord commands need to be optional so that all of them are not required in every dnsrecord command. However, FreeIPA API then does not include an information which DNS record part options are optional in term of creating a new DNS record. For example, LOC record option "latitude seconds" is not needed to add a new LOC record. This patch adds a flag "dnsrecord_optional" to all such options so that this information is available for any other UI reading the FreeIPA API. https://fedorahosted.org/freeipa/ticket/2208
2012-01-19 05:14:41 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
def get_extra(self):
if self.extra is None:
return tuple()
Mark optional DNS record parts All DNS record part options in dnsrecord commands need to be optional so that all of them are not required in every dnsrecord command. However, FreeIPA API then does not include an information which DNS record part options are optional in term of creating a new DNS record. For example, LOC record option "latitude seconds" is not needed to add a new LOC record. This patch adds a flag "dnsrecord_optional" to all such options so that this information is available for any other UI reading the FreeIPA API. https://fedorahosted.org/freeipa/ticket/2208
2012-01-19 05:14:41 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
return tuple(self._convert_dnsrecord_extra(extra) for extra in self.extra)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
def __get_part_param(self, backend, part, output_kw, default=None):
name = self.part_name_format % (self.rrtype.lower(), part.name)
label = self.part_label_format % (self.rrtype, unicode(part.label))
optional = not part.required
while True:
try:
raw = backend.textui.prompt(label,
optional=optional,
default=default)
if not raw.strip():
raw = default
output_kw[name] = part(raw)
break
except (errors.ValidationError, errors.ConversionError), e:
backend.textui.print_prompt_attribute_error(
unicode(label), unicode(e.error))
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def prompt_parts(self, backend, mod_dnsvalue=None):
mod_parts = None
if mod_dnsvalue is not None:
mod_parts = self._get_part_values(mod_dnsvalue)
user_options = {}
if self.parts is None:
return user_options
for part_id, part in enumerate(self.parts):
if mod_parts:
default = mod_parts[part_id]
else:
default = None
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
self.__get_part_param(backend, part, user_options, default)
return user_options
def prompt_missing_parts(self, backend, kw, prompt_optional=False):
user_options = {}
if self.parts is None:
return user_options
for part in self.parts:
name = self.part_name_format % (self.rrtype.lower(), part.name)
label = self.part_label_format % (self.rrtype, unicode(part.label))
if name in kw:
continue
optional = not part.required
if optional and not prompt_optional:
continue
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
default = part.get_default(**kw)
self.__get_part_param(backend, part, user_options, default)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
return user_options
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
# callbacks for per-type special record behavior
def dnsrecord_add_pre_callback(self, ldap, dn, entry_attrs, attrs_list, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
def dnsrecord_add_post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
class ForwardRecord(DNSRecord):
extra = (
Flag('create_reverse?',
label=_('Create reverse'),
doc=_('Create reverse record for this IP Address'),
flags=['no_update']
),
)
def dnsrecord_add_pre_callback(self, ldap, dn, entry_attrs, attrs_list, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
reverse_option = self._convert_dnsrecord_extra(self.extra[0])
if options.get(reverse_option.name):
records = entry_attrs.get(self.name, [])
if not records:
# --<rrtype>-create-reverse is set, but there are not records
raise errors.RequirementError(name=self.name)
for record in records:
add_records_for_host_validation(self.name, keys[-1], keys[-2], record,
check_forward=False,
check_reverse=True)
setattr(context, '%s_reverse' % self.name, entry_attrs.get(self.name))
def dnsrecord_add_post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
rev_records = getattr(context, '%s_reverse' % self.name, [])
if rev_records:
# make sure we don't run this post callback action again in nested
# commands, line adding PTR record in add_records_for_host
delattr(context, '%s_reverse' % self.name)
for record in rev_records:
try:
add_records_for_host(keys[-1], keys[-2], record,
add_forward=False, add_reverse=True)
except Exception, e:
raise errors.NonFatalError(
reason=_('Cannot create reverse record for "%(value)s": %(exc)s') \
% dict(value=record, exc=unicode(e)))
class ARecord(ForwardRecord):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
rrtype = 'A'
rfc = 1035
parts = (
Str('ip_address',
_validate_ip4addr,
label=_('IP Address'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class A6Record(DNSRecord):
rrtype = 'A6'
rfc = 3226
Add data field for A6 record Since A6 is an obsolete RR type, no DNS part option was created. This is, however, not consistent with the rest of per-type API and may cause problems. This patch adds at least a DNS part for raw A6 record data so that the record type is treated consistently. This patch also fixes interactive mode for A6 records. Their data were not detected correctly as dnsrecord_add didn't expect a number in DNS part option name. https://fedorahosted.org/freeipa/ticket/2309
2012-02-03 07:52:34 -06:00
parts = (
Str('data',
label=_('Record data'),
),
)
def _get_part_values(self, value):
# A6 RR type is obsolete and only a raw interface is provided
return (value,)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
class AAAARecord(ForwardRecord):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
rrtype = 'AAAA'
rfc = 3596
parts = (
Str('ip_address',
_validate_ip6addr,
label=_('IP Address'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class AFSDBRecord(DNSRecord):
rrtype = 'AFSDB'
rfc = 1183
parts = (
Int('subtype?',
label=_('Subtype'),
minvalue=0,
maxvalue=65535,
),
Str('hostname',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Hostname'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class APLRecord(DNSRecord):
rrtype = 'APL'
rfc = 3123
supported = False
class CERTRecord(DNSRecord):
rrtype = 'CERT'
rfc = 4398
parts = (
Int('type',
label=_('Certificate Type'),
minvalue=0,
maxvalue=65535,
),
Int('key_tag',
label=_('Key Tag'),
minvalue=0,
maxvalue=65535,
),
Int('algorithm',
label=_('Algorithm'),
minvalue=0,
maxvalue=255,
),
Str('certificate_or_crl',
label=_('Certificate/CRL'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class CNAMERecord(DNSRecord):
rrtype = 'CNAME'
rfc = 1035
parts = (
Str('hostname',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Hostname'),
doc=_('A hostname which this alias hostname points to'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class DHCIDRecord(DNSRecord):
rrtype = 'DHCID'
rfc = 4701
supported = False
class DLVRecord(DNSRecord):
rrtype = 'DLV'
rfc = 4431
supported = False
class DNAMERecord(DNSRecord):
rrtype = 'DNAME'
rfc = 2672
parts = (
Str('target',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Target'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class DNSKEYRecord(DNSRecord):
rrtype = 'DNSKEY'
rfc = 4034
supported = False
class DSRecord(DNSRecord):
rrtype = 'DS'
rfc = 4034
parts = (
Int('key_tag',
label=_('Key Tag'),
minvalue=0,
maxvalue=65535,
),
Int('algorithm',
label=_('Algorithm'),
minvalue=0,
maxvalue=255,
),
Int('digest_type',
label=_('Digest Type'),
minvalue=0,
maxvalue=255,
),
Str('digest',
label=_('Digest'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class HIPRecord(DNSRecord):
rrtype = 'HIP'
rfc = 5205
supported = False
class KEYRecord(DNSRecord):
rrtype = 'KEY'
rfc = 2535
parts = (
Int('flags',
label=_('Flags'),
minvalue=0,
maxvalue=65535,
),
Int('protocol',
label=_('Protocol'),
minvalue=0,
maxvalue=255,
),
Int('algorithm',
label=_('Algorithm'),
minvalue=0,
maxvalue=255,
),
Str('public_key',
label=_('Public Key'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class IPSECKEYRecord(DNSRecord):
rrtype = 'IPSECKEY'
rfc = 4025
supported = False
class KXRecord(DNSRecord):
rrtype = 'KX'
rfc = 2230
parts = (
Int('preference',
label=_('Preference'),
doc=_('Preference given to this exchanger. Lower values are more preferred'),
minvalue=0,
maxvalue=65535,
),
Str('exchanger',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Exchanger'),
doc=_('A host willing to act as a key exchanger'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class LOCRecord(DNSRecord):
rrtype = 'LOC'
rfc = 1876
parts = (
Int('lat_deg',
label=_('Degrees Latitude'),
minvalue=0,
maxvalue=90,
),
Int('lat_min?',
label=_('Minutes Latitude'),
minvalue=0,
maxvalue=59,
),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
Decimal('lat_sec?',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Seconds Latitude'),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
minvalue='0.0',
maxvalue='59.999',
precision=3,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
StrEnum('lat_dir',
label=_('Direction Latitude'),
values=(u'N', u'S',),
),
Int('lon_deg',
Add help for new structured DNS framework DNS Test Day shown that the new RR specific DNS options and the concepts behind them may not be easily understood. This patch adds an explanation of the new DNS framework for structured options to make it easier for the user to understand and use the new options. https://fedorahosted.org/freeipa/ticket/2382
2012-03-01 07:05:50 -06:00
label=_('Degrees Longitude'),
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
minvalue=0,
maxvalue=180,
),
Int('lon_min?',
Add help for new structured DNS framework DNS Test Day shown that the new RR specific DNS options and the concepts behind them may not be easily understood. This patch adds an explanation of the new DNS framework for structured options to make it easier for the user to understand and use the new options. https://fedorahosted.org/freeipa/ticket/2382
2012-03-01 07:05:50 -06:00
label=_('Minutes Longitude'),
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
minvalue=0,
maxvalue=59,
),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
Decimal('lon_sec?',
Add help for new structured DNS framework DNS Test Day shown that the new RR specific DNS options and the concepts behind them may not be easily understood. This patch adds an explanation of the new DNS framework for structured options to make it easier for the user to understand and use the new options. https://fedorahosted.org/freeipa/ticket/2382
2012-03-01 07:05:50 -06:00
label=_('Seconds Longitude'),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
minvalue='0.0',
maxvalue='59.999',
precision=3,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
StrEnum('lon_dir',
Add help for new structured DNS framework DNS Test Day shown that the new RR specific DNS options and the concepts behind them may not be easily understood. This patch adds an explanation of the new DNS framework for structured options to make it easier for the user to understand and use the new options. https://fedorahosted.org/freeipa/ticket/2382
2012-03-01 07:05:50 -06:00
label=_('Direction Longitude'),
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
values=(u'E', u'W',),
),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
Decimal('altitude',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Altitude'),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
minvalue='-100000.00',
maxvalue='42849672.95',
precision=2,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
Decimal('size?',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Size'),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
minvalue='0.0',
maxvalue='90000000.00',
precision=2,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
Decimal('h_precision?',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Horizontal Precision'),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
minvalue='0.0',
maxvalue='90000000.00',
precision=2,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
Decimal('v_precision?',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Vertical Precision'),
Replace float with Decimal Having float type as a base type for floating point parameters in ipalib introduces several issues, e.g. problem with representation or value comparison. Python language provides a Decimal type which help overcome these issues. This patch replaces a float type and Float parameter with a decimal.Decimal type in Decimal parameter. A precision attribute was added to Decimal parameter that can be used to limit a number of decimal places in parameter representation. This approach fixes a problem with API.txt validation where comparison of float values may fail on different architectures due to float representation error. In order to safely transfer the parameter value over RPC it is being converted to string which is then converted back to decimal.Decimal number on a server side. https://fedorahosted.org/freeipa/ticket/2260
2012-01-17 04:19:00 -06:00
minvalue='0.0',
maxvalue='90000000.00',
precision=2,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
format_error_msg = _("""format must be specified as
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
"d1 [m1 [s1]] {"N"|"S"} d2 [m2 [s2]] {"E"|"W"} alt["m"] [siz["m"] [hp["m"] [vp["m"]]]]"
where:
d1: [0 .. 90] (degrees latitude)
d2: [0 .. 180] (degrees longitude)
m1, m2: [0 .. 59] (minutes latitude/longitude)
s1, s2: [0 .. 59.999] (seconds latitude/longitude)
alt: [-100000.00 .. 42849672.95] BY .01 (altitude in meters)
siz, hp, vp: [0 .. 90000000.00] (size/precision in meters)
See RFC 1876 for details""")
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _get_part_values(self, value):
regex = re.compile(\
r'(?P<d1>\d{1,2}\s+)(?P<m1>\d{1,2}\s+)?(?P<s1>\d{1,2}\.?\d{1,3}?\s+)?'\
r'(?P<dir1>[N|S])\s+'\
r'(?P<d2>\d{1,3}\s+)(?P<m2>\d{1,2}\s+)?(?P<s2>\d{1,2}\.?\d{1,3}?\s+)?'\
r'(?P<dir2>[W|E])\s+'\
r'(?P<alt>-?\d{1,8}\.?\d{1,2}?)m?\s*'\
r'(?P<siz>\d{1,8}\.?\d{1,2}?)?m?\s*'\
r'(?P<hp>\d{1,8}\.?\d{1,2}?)?m?\s*(?P<vp>\d{1,8}\.?\d{1,2}?)?m?\s*$')
m = regex.match(value)
if m is None:
return None
return tuple(x.strip() if x is not None else x for x in m.groups())
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def _validate_parts(self, parts):
super(LOCRecord, self)._validate_parts(parts)
# create part_name -> part_id map first
part_name_map = dict((part.name, part_id) \
for part_id,part in enumerate(self.parts))
requirements = ( ('lat_sec', 'lat_min'),
('lon_sec', 'lon_min'),
('h_precision', 'size'),
('v_precision', 'h_precision', 'size') )
for req in requirements:
target_part = req[0]
if parts[part_name_map[target_part]] is not None:
required_parts = req[1:]
if any(parts[part_name_map[part]] is None for part in required_parts):
target_cli_name = self.cli_name_format % (self.rrtype.lower(), req[0])
required_cli_names = [ self.cli_name_format % (self.rrtype.lower(), part)
for part in req[1:] ]
error = _("'%(required)s' must not be empty when '%(name)s' is set") % \
dict(required=', '.join(required_cli_names),
name=target_cli_name)
raise errors.ValidationError(name=self.name, error=error)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class MXRecord(DNSRecord):
rrtype = 'MX'
rfc = 1035
parts = (
Int('preference',
label=_('Preference'),
doc=_('Preference given to this exchanger. Lower values are more preferred'),
minvalue=0,
maxvalue=65535,
),
Str('exchanger',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Exchanger'),
doc=_('A host willing to act as a mail exchanger'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class NSRecord(DNSRecord):
rrtype = 'NS'
rfc = 1035
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
parts = (
Str('hostname',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Hostname'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class NSECRecord(DNSRecord):
rrtype = 'NSEC'
rfc = 4034
format_error_msg = _('format must be specified as "NEXT TYPE1 '\
'[TYPE2 [TYPE3 [...]]]" (see RFC 4034 for details)')
_allowed_types = (u'SOA',) + _record_types
parts = (
Str('next',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_bind_hostname_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Next Domain Name'),
),
Fix NSEC record conversion NSEC record needs special treatment as it is not composed from a fixed set of DNS parts divided by space, but it contains a multivalued DNS part "types" containing a list of RR types it covers. There was already a special method for parsing raw NSEC record to DNS parts, but the other direction was missing. This patch adds special NSEC convertor to fix this issue. https://fedorahosted.org/freeipa/ticket/2307
2012-02-03 03:41:21 -06:00
StrEnum('types+',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Type Map'),
values=_allowed_types,
Fix NSEC record conversion NSEC record needs special treatment as it is not composed from a fixed set of DNS parts divided by space, but it contains a multivalued DNS part "types" containing a list of RR types it covers. There was already a special method for parsing raw NSEC record to DNS parts, but the other direction was missing. This patch adds special NSEC convertor to fix this issue. https://fedorahosted.org/freeipa/ticket/2307
2012-02-03 03:41:21 -06:00
csv=True,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _get_part_values(self, value):
values = value.split()
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if len(values) < 2:
return None
Validate NAPTR records https://fedorahosted.org/freeipa/ticket/840
2011-02-18 04:00:36 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return (values[0], tuple(values[1:]))
Validate NAPTR records https://fedorahosted.org/freeipa/ticket/840
2011-02-18 04:00:36 -06:00
Fix NSEC record conversion NSEC record needs special treatment as it is not composed from a fixed set of DNS parts divided by space, but it contains a multivalued DNS part "types" containing a list of RR types it covers. There was already a special method for parsing raw NSEC record to DNS parts, but the other direction was missing. This patch adds special NSEC convertor to fix this issue. https://fedorahosted.org/freeipa/ticket/2307
2012-02-03 03:41:21 -06:00
def _part_values_to_string(self, values, index):
self._validate_parts(values)
values_flat = [values[0],] # add "next" part
types = values[1]
if not isinstance(types, (list, tuple)):
types = [types,]
values_flat.extend(types)
return u" ".join(Str._convert_scalar(self, v, index) \
for v in values_flat if v is not None)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class NSEC3Record(DNSRecord):
rrtype = 'NSEC3'
rfc = 5155
supported = False
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class NSEC3PARAMRecord(DNSRecord):
rrtype = 'NSEC3PARAM'
rfc = 5155
supported = False
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _validate_naptr_flags(ugettext, flags):
allowed_flags = u'SAUP'
flags = flags.replace('"','').replace('\'','')
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
for flag in flags:
if flag not in allowed_flags:
return _('flags must be one of "S", "A", "U", or "P"')
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class NAPTRRecord(DNSRecord):
rrtype = 'NAPTR'
rfc = 2915
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
parts = (
Int('order',
label=_('Order'),
minvalue=0,
maxvalue=65535,
),
Int('preference',
label=_('Preference'),
minvalue=0,
maxvalue=65535,
),
Str('flags',
_validate_naptr_flags,
label=_('Flags'),
normalizer=lambda x:x.upper()
),
Str('service',
label=_('Service'),
),
Str('regexp',
label=_('Regular Expression'),
),
Str('replacement',
label=_('Replacement'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class PTRRecord(DNSRecord):
rrtype = 'PTR'
rfc = 1035
parts = (
Str('hostname',
_hostname_validator,
normalizer=_normalize_hostname,
label=_('Hostname'),
doc=_('The hostname this reverse record points to'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class RPRecord(DNSRecord):
rrtype = 'RP'
rfc = 1183
supported = False
Add SRV record target validator Add missing SRV record target validator to filter out possible user errors. https://fedorahosted.org/freeipa/ticket/2308
2012-02-03 07:25:53 -06:00
def _srv_target_validator(ugettext, value):
if value == u'.':
# service not available
return
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
return _bind_hostname_validator(ugettext, value)
Add SRV record target validator Add missing SRV record target validator to filter out possible user errors. https://fedorahosted.org/freeipa/ticket/2308
2012-02-03 07:25:53 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class SRVRecord(DNSRecord):
rrtype = 'SRV'
rfc = 2782
parts = (
Int('priority',
label=_('Priority'),
minvalue=0,
maxvalue=65535,
),
Int('weight',
label=_('Weight'),
minvalue=0,
maxvalue=65535,
),
Int('port',
label=_('Port'),
minvalue=0,
maxvalue=65535,
),
Str('target',
Add SRV record target validator Add missing SRV record target validator to filter out possible user errors. https://fedorahosted.org/freeipa/ticket/2308
2012-02-03 07:25:53 -06:00
_srv_target_validator,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
label=_('Target'),
doc=_('The domain name of the target host or \'.\' if the service is decidedly not available at this domain'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def _sig_time_validator(ugettext, value):
time_format = "%Y%m%d%H%M%S"
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
try:
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
time.strptime(value, time_format)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
except ValueError:
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return _('the value does not follow "YYYYMMDDHHMMSS" time format')
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class SIGRecord(DNSRecord):
rrtype = 'SIG'
rfc = 2535
_allowed_types = tuple([u'SOA'] + [x for x in _record_types if x != u'SIG'])
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
parts = (
StrEnum('type_covered',
label=_('Type Covered'),
values=_allowed_types,
),
Int('algorithm',
label=_('Algorithm'),
minvalue=0,
maxvalue=255,
),
Int('labels',
label=_('Labels'),
minvalue=0,
maxvalue=255,
),
Int('original_ttl',
label=_('Original TTL'),
minvalue=0,
),
Str('signature_expiration',
_sig_time_validator,
label=_('Signature Expiration'),
),
Str('signature_inception',
_sig_time_validator,
label=_('Signature Inception'),
),
Int('key_tag',
label=_('Key Tag'),
minvalue=0,
maxvalue=65535,
),
Str('signers_name',
label=_('Signer\'s Name'),
),
Str('signature',
label=_('Signature'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class SPFRecord(DNSRecord):
rrtype = 'SPF'
rfc = 4408
supported = False
class RRSIGRecord(SIGRecord):
rrtype = 'RRSIG'
rfc = 4034
class SSHFPRecord(DNSRecord):
rrtype = 'SSHFP'
rfc = 4255
parts = (
Int('algorithm',
label=_('Algorithm'),
minvalue=0,
maxvalue=255,
),
Int('fp_type',
label=_('Fingerprint Type'),
minvalue=0,
maxvalue=255,
),
Str('fingerprint',
label=_('Fingerprint'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class TARecord(DNSRecord):
rrtype = 'TA'
supported = False
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class TKEYRecord(DNSRecord):
rrtype = 'TKEY'
supported = False
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class TSIGRecord(DNSRecord):
rrtype = 'TSIG'
supported = False
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class TXTRecord(DNSRecord):
rrtype = 'TXT'
rfc = 1035
parts = (
Str('data',
label=_('Text Data'),
),
)
Improve DNS record data validation Implement missing validators for DNS RR types so that we can capture at least basic user errors. Additionally, a normalizer creating a fully-qualified domain name has been implemented for several RRs where name server may mis-interpret the domain name otherwise. Unit tests exercising these new validators for the most common RR types have been added. This patch also consolidates hard-coded values in DNS test to one place. https://fedorahosted.org/freeipa/ticket/1106
2011-11-09 10:35:52 -06:00
Fix TXT record parsing TXT record validation fails to parse the record if it contains spaces. Standard DNS part parser uses a space to divide record parts. A special parser thus need to be implemented for this RR type. https://fedorahosted.org/freeipa/ticket/2306
2012-02-03 03:01:27 -06:00
def _get_part_values(self, value):
# ignore any space in TXT record
return (value,)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
_dns_records = (
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
ARecord(),
AAAARecord(),
A6Record(),
AFSDBRecord(),
APLRecord(),
CERTRecord(),
CNAMERecord(),
DHCIDRecord(),
DLVRecord(),
DNAMERecord(),
DNSKEYRecord(),
DSRecord(),
HIPRecord(),
IPSECKEYRecord(),
KEYRecord(),
KXRecord(),
LOCRecord(),
MXRecord(),
NAPTRRecord(),
NSRecord(),
NSECRecord(),
NSEC3Record(),
NSEC3PARAMRecord(),
PTRRecord(),
RRSIGRecord(),
RPRecord(),
SIGRecord(),
SPFRecord(),
SRVRecord(),
SSHFPRecord(),
TARecord(),
TKEYRecord(),
TSIGRecord(),
TXTRecord(),
)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def __dns_record_options_iter():
for opt in (Any('dnsrecords?',
label=_('Records'),
flags=['no_create', 'no_search', 'no_update'],),
Str('dnstype?',
label=_('Record type'),
flags=['no_create', 'no_search', 'no_update'],),
Str('dnsdata?',
label=_('Record data'),
flags=['no_create', 'no_search', 'no_update'],)):
# These 3 options are used in --structured format. They are defined
# rather in takes_params than has_output_params because of their
# order - they should be printed to CLI before any DNS part param
yield opt
for option in _dns_records:
yield option
for part in option.get_parts():
yield part
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
for extra in option.get_extra():
yield extra
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
_dns_record_options = tuple(__dns_record_options_iter())
Improve dnsrecord-add interactive mode When an invalid record type is entered during dnsrecord-add interactive mode, user is provided with a list of allowed values (record types). However, the provided list contains also unsupported record types (APL, DHCID, etc.) and any attempt to add such records would end with error. This patch limits the list to supported record types only. https://fedorahosted.org/freeipa/ticket/2378
2012-02-28 01:48:08 -06:00
_dns_supported_record_types = tuple(record.rrtype for record in _dns_records \
if record.supported)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Improve DNS PTR record validation Current PTR validation is unclear and may misled the user. This patch improves the validation process so that the eventual exception is clearer. New check that the PTR record is fully qualified has been added to ensure that the reverse zone resolution behaves as expected. Additionally, several strings in the DNS plugin were prepared for localization. https://fedorahosted.org/freeipa/ticket/1129
2011-04-01 05:01:39 -05:00
# dictionary of valid reverse zone -> number of address components
_valid_reverse_zones = {
'.in-addr.arpa.' : 4,
'.ip6.arpa.' : 32,
}
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
def zone_is_reverse(zone_name):
for rev_zone_name in _valid_reverse_zones.keys():
if zone_name.endswith(rev_zone_name):
return True
return False
Allow relative DNS name in NS validator Precallback validator was failing when a zone-relative name was used as a NS record (for example record "ns" in a zone "example.com"). However, this is valid in BIND and we should allow it as well. Imports in dns module had to be switched to absolute imports (available from Python 2.5) to deal with a conflict of IPA dns module and dnspython module. https://fedorahosted.org/freeipa/ticket/2630
2012-05-29 08:06:31 -05:00
def check_ns_rec_resolvable(zone, name):
if not name.endswith('.'):
# this is a DNS name relative to the zone
zone = dns.name.from_text(zone)
name = unicode(dns.name.from_text(name, origin=zone))
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
try:
return api.Command['dns_resolve'](name)
except errors.NotFound:
Fix translatable strings in ipalib plugins. Needed for xgettext/pygettext processing.
2011-02-23 15:47:49 -06:00
raise errors.NotFound(
reason=_('Nameserver \'%(host)s\' does not have a corresponding A/AAAA record') % {'host': name}
)
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
Port installer and host plugin to the new DNS plugin * move ipa dns-resolve to the new plugin * port the installer and the host plugin to the new interface * remove the old plugin
2011-01-12 14:02:05 -06:00
def dns_container_exists(ldap):
try:
ldap.get_entry(api.env.container_dns, [])
except errors.NotFound:
return False
return True
Fill new DNS zone update policy by default For security reasons, dynamic updates are not enabled for new DNS zones. In order to enable the dynamic zone securely, user needs to allow dynamic updates and create a zone update policy. The policy is not easy to construct for regular users, we should rather fill it by default and let users just switch the policy on or off. https://fedorahosted.org/freeipa/ticket/2441
2012-06-04 10:53:34 -05:00
def default_zone_update_policy(zone):
if zone_is_reverse(zone):
return get_dns_reverse_zone_update_policy(api.env.realm, zone)
else:
return get_dns_forward_zone_update_policy(api.env.realm)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
class dnszone(LDAPObject):
"""
DNS Zone, container for resource records.
"""
container_dn = api.env.container_dns
Fixed object_name and object_name_plural internationalization The object_name, object_name_plural and messages that use these attributes have been converted to support translation. The label attribute in the Param class has been modified to accept unicode string. Ticket #1435
2011-07-12 11:01:25 -05:00
object_name = _('DNS zone')
object_name_plural = _('DNS zones')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
object_class = ['top', 'idnsrecord', 'idnszone']
Per-domain DNS record permissions IPA implements read/write permissions for DNS record or zones. Provided set of permissions and privileges can, however, only grant access to the whole DNS tree, which may not be appropriate. Administrators may miss more fine-grained permissions allowing them to delegate access per-zone. Create a new IPA auxiliary objectclass ipaDNSZone allowing a managedBy attribute for a DNS zone. This attribute will hold a group DN (in this case a permission) which allows its members to read or write in a zone. Member permissions in given zone will only have 2 limitations: 1) Members cannot delete the zone 2) Members cannot edit managedBy attribute Current DNS deny ACI used to enforce read access is removed so that DNS privileges are based on allow ACIs only, which is much more flexible approach as deny ACIs have always precedence and limit other extensions. Per-zone access is allowed in 3 generic ACIs placed in cn=dns,$SUFFIX so that no special ACIs has to be added to DNS zones itselves. 2 new commands have been added which allows an administrator to create the system permission allowing the per-zone access and fill a zone's managedBy attribute: * dnszone-add-permission: Add per-zone permission * dnszone-remove-permission: Remove per-zone permission https://fedorahosted.org/freeipa/ticket/2511
2012-06-27 06:10:10 -05:00
possible_objectclasses = ['ipadnszone']
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
default_attributes = [
'idnsname', 'idnszoneactive', 'idnssoamname', 'idnssoarname',
'idnssoaserial', 'idnssoarefresh', 'idnssoaretry', 'idnssoaexpire',
Add DNS conditional forwarding Add ability configure per-zone forwarder for DNS zones. Any data in such zone will then be considered as non-authoritative and all queries will be sent to specified forwarder. https://fedorahosted.org/freeipa/ticket/2108
2012-02-14 04:10:22 -06:00
'idnssoaminimum', 'idnsallowquery', 'idnsallowtransfer',
'idnsforwarders', 'idnsforwardpolicy'
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
] + _record_attributes
Fixed entity labels. The entity labels in the following locations have been fixed: - search facet title: plural - details facet title: singular - association facet title: singular - breadcrumb: plural - adder dialog title: singular - deleter dialog title: plural Some entity labels have been changed into the correct plural form. Unused file install/ui/test/data/i18n_messages.json has been removed. Ticket #1249 Ticket #1387
2011-06-24 11:39:48 -05:00
label = _('DNS Zones')
Fixed label capitalization The CSS text-transform sometimes produces incorrect capitalization, so the code has been modified to use translated labels that already contain the correct capitalization. Ticket #1424
2011-07-13 21:10:47 -05:00
label_singular = _('DNS Zone')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
takes_params = (
Str('idnsname',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_domain_name_validator,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='name',
label=_('Zone name'),
doc=_('Zone name (FQDN)'),
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
default_from=lambda name_from_ip: _reverse_zone_name(name_from_ip),
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
normalizer=lambda value: value.lower(),
primary_key=True,
),
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
Str('name_from_ip?', _validate_ipnet,
label=_('Reverse zone IP network'),
doc=_('IP network to create reverse zone name from'),
Fix LDAP object parameter encoding Parameters in LDAP objects missed an information if they are real LDAP attributes or not. Real LDAP attributes are written to entry_attrs dictionary in plugin callbacks and are being encoded. This causes issues when plugin callbacks does not expect that the parameters values are already encoded for submission to LDAP. This patch introduces a new flag "noattribute" used to mark that a parameter is not an LDAP attribute and thus should not be encoded or added to entry_attrs. Param documentation is improved to describe the meaning of this and other Param flags or attributes. https://fedorahosted.org/freeipa/ticket/2097
2011-11-14 10:03:44 -06:00
flags=('virtual_attribute',),
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
),
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Str('idnssoamname',
cli_name='name_server',
Fix style and grammatical issues in built-in command help. There is a rather large API.txt change but it is only due to changes in the doc string in parameters. ticket 729
2011-03-04 10:08:54 -06:00
label=_('Authoritative nameserver'),
Improve dnszone-add error message Check that NS address passed in dnszone-add is a domain name and not an IP address. Make this clear also the parameter help. https://fedorahosted.org/freeipa/ticket/1567
2011-08-04 02:59:01 -05:00
doc=_('Authoritative nameserver domain name'),
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
),
Str('idnssoarname',
Add --zonemgr/--admin-mail validator Do at least a basic validation of DNS zone manager mail address. Do not require '@' to be in the mail address as the SOA record stores this value without it and people may be used to configure it that way. '@' is always removed by the installer/dns plugin before the DNS zone is created. https://fedorahosted.org/freeipa/ticket/1966
2011-10-24 11:35:48 -05:00
_rname_validator,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='admin_email',
label=_('Administrator e-mail address'),
doc=_('Administrator e-mail address'),
Change default DNS zone manager to hostmaster Change our default zone manager to hostmaster@<domain> (as per RFC 2142 recommendation). https://fedorahosted.org/freeipa/ticket/1981
2011-11-23 09:09:29 -06:00
default_from=lambda idnsname: 'hostmaster.%s' % idnsname,
Improve zonemgr validator and normalizer The validator has been improved to support better both SOA format (e-mail address in a domain name format, without '@') and standard e-mail format. Allow '\.' character in a SOA format encoding the standard '.' in the local-part of an e-mail. Normalization code has been moved to one common function. https://fedorahosted.org/freeipa/ticket/2053
2011-11-23 09:03:51 -06:00
normalizer=normalize_zonemgr,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
),
Mark some attributes required to match the schema. This makes no changes to the functionality in the command-line or GUI because these all have defaults anyway. This is mostly to show them properly in the UI and prevent someone from trying to erase the value (and getting a nasty schema error in response). https://fedorahosted.org/freeipa/ticket/2015
2011-11-28 11:31:45 -06:00
Int('idnssoaserial',
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='serial',
label=_('SOA serial'),
doc=_('SOA record serial number'),
Better doc for idnssoaminimum
2011-02-18 06:53:27 -06:00
minvalue=1,
Change parameters to use only default_from for dynamic default values. Replace all occurences of create_default with equivalent default_from and remove create_default from the framework. This is needed for proper parameter validation, as there is no way to tell which parameters to validate prior to calling create_default, because create_default does not provide information about which parameters are used for generating the default value.
2012-03-15 03:32:37 -05:00
default_from=_create_zone_serial,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
autofill=True,
),
Mark some attributes required to match the schema. This makes no changes to the functionality in the command-line or GUI because these all have defaults anyway. This is mostly to show them properly in the UI and prevent someone from trying to erase the value (and getting a nasty schema error in response). https://fedorahosted.org/freeipa/ticket/2015
2011-11-28 11:31:45 -06:00
Int('idnssoarefresh',
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='refresh',
label=_('SOA refresh'),
doc=_('SOA record refresh time'),
Better doc for idnssoaminimum
2011-02-18 06:53:27 -06:00
minvalue=0,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
default=3600,
autofill=True,
),
Mark some attributes required to match the schema. This makes no changes to the functionality in the command-line or GUI because these all have defaults anyway. This is mostly to show them properly in the UI and prevent someone from trying to erase the value (and getting a nasty schema error in response). https://fedorahosted.org/freeipa/ticket/2015
2011-11-28 11:31:45 -06:00
Int('idnssoaretry',
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='retry',
label=_('SOA retry'),
doc=_('SOA record retry time'),
Better doc for idnssoaminimum
2011-02-18 06:53:27 -06:00
minvalue=0,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
default=900,
autofill=True,
),
Mark some attributes required to match the schema. This makes no changes to the functionality in the command-line or GUI because these all have defaults anyway. This is mostly to show them properly in the UI and prevent someone from trying to erase the value (and getting a nasty schema error in response). https://fedorahosted.org/freeipa/ticket/2015
2011-11-28 11:31:45 -06:00
Int('idnssoaexpire',
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='expire',
label=_('SOA expire'),
doc=_('SOA record expire time'),
default=1209600,
Better doc for idnssoaminimum
2011-02-18 06:53:27 -06:00
minvalue=0,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
autofill=True,
),
Mark some attributes required to match the schema. This makes no changes to the functionality in the command-line or GUI because these all have defaults anyway. This is mostly to show them properly in the UI and prevent someone from trying to erase the value (and getting a nasty schema error in response). https://fedorahosted.org/freeipa/ticket/2015
2011-11-28 11:31:45 -06:00
Int('idnssoaminimum',
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='minimum',
label=_('SOA minimum'),
Better doc for idnssoaminimum
2011-02-18 06:53:27 -06:00
doc=_('How long should negative responses be cached'),
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
default=3600,
Better doc for idnssoaminimum
2011-02-18 06:53:27 -06:00
minvalue=0,
maxvalue=10800,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
autofill=True,
),
Int('dnsttl?',
cli_name='ttl',
label=_('SOA time to live'),
doc=_('SOA record time to live'),
),
StrEnum('dnsclass?',
cli_name='class',
label=_('SOA class'),
doc=_('SOA record class'),
values=_record_classes,
),
Str('idnsupdatepolicy?',
cli_name='update_policy',
label=_('BIND update policy'),
doc=_('BIND update policy'),
Fill new DNS zone update policy by default For security reasons, dynamic updates are not enabled for new DNS zones. In order to enable the dynamic zone securely, user needs to allow dynamic updates and create a zone update policy. The policy is not easy to construct for regular users, we should rather fill it by default and let users just switch the policy on or off. https://fedorahosted.org/freeipa/ticket/2441
2012-06-04 10:53:34 -05:00
default_from=lambda idnsname: default_zone_update_policy(idnsname),
autofill=True
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
),
Fix DNS zone --allow-dynupdate option behavior --allow-dynupdate was implemented as a Flag parameter type, which is not convenient for LDAP attributes. When a DNS zone with permitted dynamic updates was modified and the --allow-dynupdate flag was not set, dynamic updates were turned off. This patch changes the option type to Bool parameter type which behaves according to user expectations when modifying the zone. https://fedorahosted.org/freeipa/ticket/2039
2011-11-09 07:23:15 -06:00
Bool('idnszoneactive?',
Added some fields to DNS2 plugin Field idnszoneactive is marked as optional, because it is set to true by default (see class dnszone_add). https://fedorahosted.org/freeipa/ticket/601
2010-12-21 05:18:05 -06:00
cli_name='zone_active',
label=_('Active zone'),
doc=_('Is zone active?'),
flags=['no_create', 'no_update'],
attribute=True,
),
Fix DNS zone --allow-dynupdate option behavior --allow-dynupdate was implemented as a Flag parameter type, which is not convenient for LDAP attributes. When a DNS zone with permitted dynamic updates was modified and the --allow-dynupdate flag was not set, dynamic updates were turned off. This patch changes the option type to Bool parameter type which behaves according to user expectations when modifying the zone. https://fedorahosted.org/freeipa/ticket/2039
2011-11-09 07:23:15 -06:00
Bool('idnsallowdynupdate?',
cli_name='dynamic_update',
Added some fields to DNS2 plugin Field idnszoneactive is marked as optional, because it is set to true by default (see class dnszone_add). https://fedorahosted.org/freeipa/ticket/601
2010-12-21 05:18:05 -06:00
label=_('Dynamic update'),
Fix style and grammatical issues in built-in command help. There is a rather large API.txt change but it is only due to changes in the doc string in parameters. ticket 729
2011-03-04 10:08:54 -06:00
doc=_('Allow dynamic updates.'),
Added some fields to DNS2 plugin Field idnszoneactive is marked as optional, because it is set to true by default (see class dnszone_add). https://fedorahosted.org/freeipa/ticket/601
2010-12-21 05:18:05 -06:00
attribute=True,
Fix DNS zone --allow-dynupdate option behavior --allow-dynupdate was implemented as a Flag parameter type, which is not convenient for LDAP attributes. When a DNS zone with permitted dynamic updates was modified and the --allow-dynupdate flag was not set, dynamic updates were turned off. This patch changes the option type to Bool parameter type which behaves according to user expectations when modifying the zone. https://fedorahosted.org/freeipa/ticket/2039
2011-11-09 07:23:15 -06:00
default=False,
autofill=True
Added some fields to DNS2 plugin Field idnszoneactive is marked as optional, because it is set to true by default (see class dnszone_add). https://fedorahosted.org/freeipa/ticket/601
2010-12-21 05:18:05 -06:00
),
Query and transfer ACLs for DNS zones Provide a way to specify BIND allow-query and allow-transfer ACLs for DNS zones. IMPORTANT: new bind-dyndb-ldap adds a zone transfer ability. To avoid zone information leaks to unintended places, allow-transfer ACL for every zone is by default set to none and has to be explicitly enabled by an Administrator. This is done both for new DNS zones and old DNS zones during RPM update via new DNS upgrade plugin. https://fedorahosted.org/freeipa/ticket/1211
2012-02-24 02:30:39 -06:00
Str('idnsallowquery?',
_validate_bind_aci,
normalizer=_normalize_bind_aci,
cli_name='allow_query',
label=_('Allow query'),
doc=_('Semicolon separated list of IP addresses or networks which are allowed to issue queries'),
default=u'any;', # anyone can issue queries by default
autofill=True,
),
Str('idnsallowtransfer?',
_validate_bind_aci,
normalizer=_normalize_bind_aci,
cli_name='allow_transfer',
label=_('Allow transfer'),
doc=_('Semicolon separated list of IP addresses or networks which are allowed to transfer the zone'),
default=u'none;', # no one can issue queries by default
autofill=True,
),
Add DNS conditional forwarding Add ability configure per-zone forwarder for DNS zones. Any data in such zone will then be considered as non-authoritative and all queries will be sent to specified forwarder. https://fedorahosted.org/freeipa/ticket/2108
2012-02-14 04:10:22 -06:00
Str('idnsforwarders*',
Allow port numbers for idnsForwarders Let user enter custom ports for zone conditional forwarders or global forwarders in dnsconfig. Ports can be specified in a standard BIND format: IP_ADDRESS [port PORT] https://fedorahosted.org/freeipa/ticket/2462
2012-03-20 09:35:54 -05:00
_validate_bind_forwarder,
Add DNS conditional forwarding Add ability configure per-zone forwarder for DNS zones. Any data in such zone will then be considered as non-authoritative and all queries will be sent to specified forwarder. https://fedorahosted.org/freeipa/ticket/2108
2012-02-14 04:10:22 -06:00
cli_name='forwarder',
label=_('Zone forwarders'),
Fix dnszone-mod --forwader option help string Help should not point to global forwarders but rather to per-zone conditional forwarders. https://fedorahosted.org/freeipa/ticket/2717
2012-06-05 03:42:43 -05:00
doc=_('A list of per-zone forwarders. A custom port can be specified '
Allow port numbers for idnsForwarders Let user enter custom ports for zone conditional forwarders or global forwarders in dnsconfig. Ports can be specified in a standard BIND format: IP_ADDRESS [port PORT] https://fedorahosted.org/freeipa/ticket/2462
2012-03-20 09:35:54 -05:00
'for each forwarder using a standard format "IP_ADDRESS port PORT"'),
Add DNS conditional forwarding Add ability configure per-zone forwarder for DNS zones. Any data in such zone will then be considered as non-authoritative and all queries will be sent to specified forwarder. https://fedorahosted.org/freeipa/ticket/2108
2012-02-14 04:10:22 -06:00
csv=True,
),
StrEnum('idnsforwardpolicy?',
cli_name='forward_policy',
label=_('Forward policy'),
values=(u'only', u'first',),
),
Add API for PTR sync control New version of bind-dyndb-ldap plugin have an ability to automatically update machine reverse address when its forward address is updated via GSS-TSIG update. The reverse zone must be managed by FreeIPA as well in order of this feature to work. As it would not be secure to enable this behaviour for all zones there is a global attribute that can enable PTR sync for all zones and also a per-zone attribute that can enable for chosen zones only. This patch adds an API for this control. https://fedorahosted.org/freeipa/ticket/2176
2012-02-14 04:11:41 -06:00
Bool('idnsallowsyncptr?',
cli_name='allow_sync_ptr',
label=_('Allow PTR sync'),
doc=_('Allow synchronization of forward (A, AAAA) and reverse (PTR) records in the zone'),
),
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
)
Improve FQDN handling in DNS and host plugins DNS and host plugin does not work well with domain names ending with dot. host plugin creates a record with two fqdn attributes when such hostname is created which then has to be manually fixed. DNS plugin handled zones with and without trailing dot as two distinct zones, which may lead to issues when both zones are created. This patch sanitizes approach to FQDNs in both DNS and host plugin. Hostnames are now always normalized to the form without trailing dot as this form did not work before and it would keep hostname form consistent without changes in our server/client enrollment process. As DNS zones always worked in both forms this patch rather makes sure that the plugin works with both forms of one zone and prevents creating 2 identical zones with just different format. https://fedorahosted.org/freeipa/ticket/2420
2012-02-28 02:13:13 -06:00
def get_dn(self, *keys, **options):
zone = keys[-1]
dn = super(dnszone, self).get_dn(zone, **options)
try:
self.backend.get_entry(dn, [''])
except errors.NotFound:
if zone.endswith(u'.'):
zone = zone[:-1]
else:
zone = zone + u'.'
test_dn = super(dnszone, self).get_dn(zone, **options)
try:
(dn, entry_attrs) = self.backend.get_entry(test_dn, [''])
except errors.NotFound:
pass
return dn
Per-domain DNS record permissions IPA implements read/write permissions for DNS record or zones. Provided set of permissions and privileges can, however, only grant access to the whole DNS tree, which may not be appropriate. Administrators may miss more fine-grained permissions allowing them to delegate access per-zone. Create a new IPA auxiliary objectclass ipaDNSZone allowing a managedBy attribute for a DNS zone. This attribute will hold a group DN (in this case a permission) which allows its members to read or write in a zone. Member permissions in given zone will only have 2 limitations: 1) Members cannot delete the zone 2) Members cannot edit managedBy attribute Current DNS deny ACI used to enforce read access is removed so that DNS privileges are based on allow ACIs only, which is much more flexible approach as deny ACIs have always precedence and limit other extensions. Per-zone access is allowed in 3 generic ACIs placed in cn=dns,$SUFFIX so that no special ACIs has to be added to DNS zones itselves. 2 new commands have been added which allows an administrator to create the system permission allowing the per-zone access and fill a zone's managedBy attribute: * dnszone-add-permission: Add per-zone permission * dnszone-remove-permission: Remove per-zone permission https://fedorahosted.org/freeipa/ticket/2511
2012-06-27 06:10:10 -05:00
def permission_name(self, zone):
return u"Manage DNS zone %s" % zone
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnszone)
class dnszone_add(LDAPCreate):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Create new DNS zone (SOA record).')
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
takes_options = LDAPCreate.takes_options + (
Flag('force',
Need force option in DNS zone adder dialog When adding a new DNS zone in the WebUI, IPA server will verify whether the nameserver is in DNS. Sometimes it is necessary to skip the verification. This patch adds a --force option already available in CLI which can skip this the verification. https://fedorahosted.org/freeipa/ticket/1105
2011-04-01 10:09:07 -05:00
label=_('Force'),
Fix style and grammatical issues in built-in command help. There is a rather large API.txt change but it is only due to changes in the doc string in parameters. ticket 729
2011-03-04 10:08:54 -06:00
doc=_('Force DNS zone creation even if nameserver not in DNS.'),
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
),
Str('ip_address?', _validate_ipaddr,
doc=_('Add the nameserver to DNS with this IP address'),
),
)
Improve FQDN handling in DNS and host plugins DNS and host plugin does not work well with domain names ending with dot. host plugin creates a record with two fqdn attributes when such hostname is created which then has to be manually fixed. DNS plugin handled zones with and without trailing dot as two distinct zones, which may lead to issues when both zones are created. This patch sanitizes approach to FQDNs in both DNS and host plugin. Hostnames are now always normalized to the form without trailing dot as this form did not work before and it would keep hostname form consistent without changes in our server/client enrollment process. As DNS zones always worked in both forms this patch rather makes sure that the plugin works with both forms of one zone and prevents creating 2 identical zones with just different format. https://fedorahosted.org/freeipa/ticket/2420
2012-02-28 02:13:13 -06:00
def pre_callback(self, ldap, dn, entry_attrs, attrs_list, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Port installer and host plugin to the new DNS plugin * move ipa dns-resolve to the new plugin * port the installer and the host plugin to the new interface * remove the old plugin
2011-01-12 14:02:05 -06:00
if not dns_container_exists(self.api.Backend.ldap2):
raise errors.NotFound(reason=_('DNS is not configured'))
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
entry_attrs['idnszoneactive'] = 'TRUE'
Fixes for the DNS plugin https://fedorahosted.org/freeipa/ticket/730
2011-01-07 03:29:16 -06:00
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
# Check nameserver has a forward record
Fixes for the DNS plugin https://fedorahosted.org/freeipa/ticket/730
2011-01-07 03:29:16 -06:00
nameserver = entry_attrs['idnssoamname']
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
Improve dnszone-add error message Check that NS address passed in dnszone-add is a domain name and not an IP address. Make this clear also the parameter help. https://fedorahosted.org/freeipa/ticket/1567
2011-08-04 02:59:01 -05:00
# NS record must contain domain name
if valid_ip(nameserver):
raise errors.ValidationError(name='name-server',
error=unicode(_("Nameserver address is not a fully qualified domain name")))
Fixes for the DNS plugin https://fedorahosted.org/freeipa/ticket/730
2011-01-07 03:29:16 -06:00
if nameserver[-1] != '.':
nameserver += '.'
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
Allow relative DNS name in NS validator Precallback validator was failing when a zone-relative name was used as a NS record (for example record "ns" in a zone "example.com"). However, this is valid in BIND and we should allow it as well. Imports in dns module had to be switched to absolute imports (available from Python 2.5) to deal with a conflict of IPA dns module and dnspython module. https://fedorahosted.org/freeipa/ticket/2630
2012-05-29 08:06:31 -05:00
if not 'ip_address' in options and not options['force']:
check_ns_rec_resolvable(keys[0], nameserver)
Fixes for the DNS plugin https://fedorahosted.org/freeipa/ticket/730
2011-01-07 03:29:16 -06:00
entry_attrs['nsrecord'] = nameserver
entry_attrs['idnssoamname'] = nameserver
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return dn
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
def post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
if 'ip_address' in options:
nameserver = entry_attrs['idnssoamname'][0][:-1] # ends with a dot
nsparts = nameserver.split('.')
add_forward_record('.'.join(nsparts[1:]),
nsparts[0],
options['ip_address'])
return dn
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnszone_add)
class dnszone_del(LDAPDelete):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Delete DNS zone (SOA record).')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Per-domain DNS record permissions IPA implements read/write permissions for DNS record or zones. Provided set of permissions and privileges can, however, only grant access to the whole DNS tree, which may not be appropriate. Administrators may miss more fine-grained permissions allowing them to delegate access per-zone. Create a new IPA auxiliary objectclass ipaDNSZone allowing a managedBy attribute for a DNS zone. This attribute will hold a group DN (in this case a permission) which allows its members to read or write in a zone. Member permissions in given zone will only have 2 limitations: 1) Members cannot delete the zone 2) Members cannot edit managedBy attribute Current DNS deny ACI used to enforce read access is removed so that DNS privileges are based on allow ACIs only, which is much more flexible approach as deny ACIs have always precedence and limit other extensions. Per-zone access is allowed in 3 generic ACIs placed in cn=dns,$SUFFIX so that no special ACIs has to be added to DNS zones itselves. 2 new commands have been added which allows an administrator to create the system permission allowing the per-zone access and fill a zone's managedBy attribute: * dnszone-add-permission: Add per-zone permission * dnszone-remove-permission: Remove per-zone permission https://fedorahosted.org/freeipa/ticket/2511
2012-06-27 06:10:10 -05:00
def post_callback(self, ldap, dn, *keys, **options):
try:
api.Command['permission_del'](self.obj.permission_name(keys[-1]),
force=True)
except errors.NotFound:
pass
return True
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnszone_del)
class dnszone_mod(LDAPUpdate):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Modify DNS zone (SOA record).')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnszone_mod)
class dnszone_find(LDAPSearch):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Search for DNS zones (SOA records).')
Validate name_from_ip parameter of dnszone. ticket 1627
2011-09-26 04:00:32 -05:00
def args_options_2_params(self, *args, **options):
Fix the procedure for getting default values of command parameters. The parameters used in default_from of other parameters are now properly validated before the default_from is called. ticket 1847
2012-01-16 08:21:50 -06:00
# FIXME: Check that name_from_ip is valid. This is necessary because
# custom validation rules, including _validate_ipnet, are not
# used when doing a search. Once we have a parameter type for
# IP network objects, this will no longer be necessary, as the
# parameter type will handle the validation itself (see
# <https://fedorahosted.org/freeipa/ticket/2266>).
Validate name_from_ip parameter of dnszone. ticket 1627
2011-09-26 04:00:32 -05:00
if 'name_from_ip' in options:
self.obj.params['name_from_ip'](unicode(options['name_from_ip']))
return super(dnszone_find, self).args_options_2_params(*args, **options)
Add ability to specify DNS reverse zone name by IP network address. In order for this to work, chaining of parameters through default_from is made possible. ticket 1474
2011-06-21 07:07:19 -05:00
def args_options_2_entry(self, *args, **options):
if 'name_from_ip' in options:
if 'idnsname' not in options:
options['idnsname'] = self.obj.params['idnsname'].get_default(**options)
del options['name_from_ip']
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return super(dnszone_find, self).args_options_2_entry(*args, **options)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Filter reverse zones in dnszone-find Implements a new option to filter out reverse zones. This patch also do some clean up in dns plugin - debug prints were accidentally left here in the last dns patch. https://fedorahosted.org/freeipa/ticket/1471
2011-07-13 07:01:17 -05:00
takes_options = LDAPSearch.takes_options + (
Flag('forward_only',
label=_('Forward zones only'),
cli_name='forward_only',
doc=_('Search for forward zones only'),
),
)
def pre_callback(self, ldap, filter, attrs_list, base_dn, scope, *args, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(base_dn, DN)
Filter reverse zones in dnszone-find Implements a new option to filter out reverse zones. This patch also do some clean up in dns plugin - debug prints were accidentally left here in the last dns patch. https://fedorahosted.org/freeipa/ticket/1471
2011-07-13 07:01:17 -05:00
if options.get('forward_only', False):
search_kw = {}
search_kw['idnsname'] = _valid_reverse_zones.keys()
rev_zone_filter = ldap.make_filter(search_kw, rules=ldap.MATCH_NONE, exact=False,
trailing_wildcard=False)
filter = ldap.combine_filters((rev_zone_filter, filter), rules=ldap.MATCH_ALL)
return (filter, base_dn, scope)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnszone_find)
class dnszone_show(LDAPRetrieve):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Display information about a DNS zone (SOA record).')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnszone_show)
class dnszone_disable(LDAPQuery):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Disable DNS Zone.')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
has_output = output.standard_value
msg_summary = _('Disabled DNS zone "%(value)s"')
def execute(self, *keys, **options):
ldap = self.obj.backend
dn = self.obj.get_dn(*keys, **options)
try:
ldap.update_entry(dn, {'idnszoneactive': 'FALSE'})
except errors.EmptyModlist:
pass
return dict(result=True, value=keys[-1])
api.register(dnszone_disable)
class dnszone_enable(LDAPQuery):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Enable DNS Zone.')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
has_output = output.standard_value
msg_summary = _('Enabled DNS zone "%(value)s"')
def execute(self, *keys, **options):
ldap = self.obj.backend
dn = self.obj.get_dn(*keys, **options)
try:
ldap.update_entry(dn, {'idnszoneactive': 'TRUE'})
except errors.EmptyModlist:
pass
return dict(result=True, value=keys[-1])
api.register(dnszone_enable)
Per-domain DNS record permissions IPA implements read/write permissions for DNS record or zones. Provided set of permissions and privileges can, however, only grant access to the whole DNS tree, which may not be appropriate. Administrators may miss more fine-grained permissions allowing them to delegate access per-zone. Create a new IPA auxiliary objectclass ipaDNSZone allowing a managedBy attribute for a DNS zone. This attribute will hold a group DN (in this case a permission) which allows its members to read or write in a zone. Member permissions in given zone will only have 2 limitations: 1) Members cannot delete the zone 2) Members cannot edit managedBy attribute Current DNS deny ACI used to enforce read access is removed so that DNS privileges are based on allow ACIs only, which is much more flexible approach as deny ACIs have always precedence and limit other extensions. Per-zone access is allowed in 3 generic ACIs placed in cn=dns,$SUFFIX so that no special ACIs has to be added to DNS zones itselves. 2 new commands have been added which allows an administrator to create the system permission allowing the per-zone access and fill a zone's managedBy attribute: * dnszone-add-permission: Add per-zone permission * dnszone-remove-permission: Remove per-zone permission https://fedorahosted.org/freeipa/ticket/2511
2012-06-27 06:10:10 -05:00
class dnszone_add_permission(LDAPQuery):
__doc__ = _('Add a permission for per-zone access delegation.')
has_output = output.standard_value
msg_summary = _('Added system permission "%(value)s"')
def execute(self, *keys, **options):
ldap = self.obj.backend
dn = self.obj.get_dn(*keys, **options)
try:
(dn_, entry_attrs) = ldap.get_entry(dn, ['objectclass'])
except errors.NotFound:
self.obj.handle_not_found(*keys)
permission_name = self.obj.permission_name(keys[-1])
permission = api.Command['permission_add_noaci'](permission_name,
permissiontype=u'SYSTEM'
)['result']
update = {}
dnszone_ocs = entry_attrs.get('objectclass')
if dnszone_ocs:
dnszone_ocs.append('ipadnszone')
update['objectclass'] = list(set(dnszone_ocs))
update['managedby'] = [permission['dn']]
ldap.update_entry(dn, update)
return dict(
result=True,
value=permission_name,
)
api.register(dnszone_add_permission)
class dnszone_remove_permission(LDAPQuery):
__doc__ = _('Remove a permission for per-zone access delegation.')
has_output = output.standard_value
msg_summary = _('Removed system permission "%(value)s"')
def execute(self, *keys, **options):
ldap = self.obj.backend
dn = self.obj.get_dn(*keys, **options)
try:
ldap.update_entry(dn, {'managedby': None})
except errors.NotFound:
self.obj.handle_not_found(*keys)
except errors.EmptyModlist:
# managedBy attribute is clean, lets make sure there is also no
# dangling DNS zone permission
pass
permission_name = self.obj.permission_name(keys[-1])
api.Command['permission_del'](permission_name, force=True)
return dict(
result=True,
value=permission_name,
)
api.register(dnszone_remove_permission)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
class dnsrecord(LDAPObject):
"""
DNS record.
"""
parent_object = 'dnszone'
container_dn = api.env.container_dns
Fixed object_name and object_name_plural internationalization The object_name, object_name_plural and messages that use these attributes have been converted to support translation. The label attribute in the Param class has been modified to accept unicode string. Ticket #1435
2011-07-12 11:01:25 -05:00
object_name = _('DNS resource record')
object_name_plural = _('DNS resource records')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
object_class = ['top', 'idnsrecord']
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
default_attributes = ['idnsname'] + _record_attributes
Add rename option for DNS records This option will make renaming DNS records much easier. Add a unit test for this new functionality. https://fedorahosted.org/freeipa/ticket/2600
2012-05-29 08:58:36 -05:00
rdn_is_primary_key = True
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Fixed entity labels. The entity labels in the following locations have been fixed: - search facet title: plural - details facet title: singular - association facet title: singular - breadcrumb: plural - adder dialog title: singular - deleter dialog title: plural Some entity labels have been changed into the correct plural form. Unused file install/ui/test/data/i18n_messages.json has been removed. Ticket #1249 Ticket #1387
2011-06-24 11:39:48 -05:00
label = _('DNS Resource Records')
Fixed label capitalization The CSS text-transform sometimes produces incorrect capitalization, so the code has been modified to use translated labels that already contain the correct capitalization. Ticket #1424
2011-07-13 21:10:47 -05:00
label_singular = _('DNS Resource Record')
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
takes_params = (
Str('idnsname',
Improve hostname and domain name validation DNS plugin did not check DNS zone and DNS record validity and user was thus able to create domains like "foo bar" or other invalid DNS labels which would really confuse both user and bind-dyndb-ldap plugin. This patch at first consolidates hostname/domain name validators so that they use common functions and we don't have regular expressions and other checks defined in several places. These new cleaned validators are then used for zone/record name validation. https://fedorahosted.org/freeipa/ticket/2384
2012-02-28 02:05:01 -06:00
_dns_record_name_validator,
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
cli_name='name',
label=_('Record name'),
doc=_('Record name'),
primary_key=True,
),
Int('dnsttl?',
cli_name='ttl',
label=_('Time to live'),
doc=_('Time to live'),
),
StrEnum('dnsclass?',
cli_name='class',
label=_('Class'),
doc=_('DNS class'),
values=_record_classes,
),
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
) + _dns_record_options
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
structured_flag = Flag('structured',
label=_('Structured'),
doc=_('Parse all raw DNS records and return them in a structured way'),
)
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
def _nsrecord_pre_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
nsrecords = entry_attrs.get('nsrecord')
if options.get('force', False) or nsrecords is None:
return
Allow relative DNS name in NS validator Precallback validator was failing when a zone-relative name was used as a NS record (for example record "ns" in a zone "example.com"). However, this is valid in BIND and we should allow it as well. Imports in dns module had to be switched to absolute imports (available from Python 2.5) to deal with a conflict of IPA dns module and dnspython module. https://fedorahosted.org/freeipa/ticket/2630
2012-05-29 08:06:31 -05:00
for nsrecord in nsrecords:
check_ns_rec_resolvable(keys[0], nsrecord)
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
def _ptrrecord_pre_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
ptrrecords = entry_attrs.get('ptrrecord')
if ptrrecords is None:
return
zone = keys[-2]
if self.is_pkey_zone_record(*keys):
addr = u''
else:
addr = keys[-1]
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
zone_len = 0
for valid_zone in _valid_reverse_zones:
if zone.find(valid_zone) != -1:
zone = zone.replace(valid_zone,'')
zone_name = valid_zone
zone_len = _valid_reverse_zones[valid_zone]
if not zone_len:
allowed_zones = ', '.join(_valid_reverse_zones)
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
raise errors.ValidationError(name='ptrrecord',
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
error=unicode(_('Reverse zone for PTR record should be a sub-zone of one the following fully qualified domains: %s') % allowed_zones))
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
addr_len = len(addr.split('.')) if addr else 0
ip_addr_comp_count = addr_len + len(zone.split('.'))
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
if ip_addr_comp_count != zone_len:
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
raise errors.ValidationError(name='ptrrecord',
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
error=unicode(_('Reverse zone %(name)s requires exactly %(count)d IP address components, %(user_count)d given')
% dict(name=zone_name, count=zone_len, user_count=ip_addr_comp_count)))
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
def run_precallback_validators(self, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
ldap = self.api.Backend.ldap2
for rtype in entry_attrs:
rtype_cb = getattr(self, '_%s_pre_callback' % rtype, None)
if rtype_cb:
rtype_cb(ldap, dn, entry_attrs, *keys, **options)
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
Fix assorted bugs found by pylint
2011-01-25 11:46:26 -06:00
def is_pkey_zone_record(self, *keys):
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
idnsname = keys[-1]
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
if idnsname == str(_dns_zone_record) or idnsname == ('%s.' % keys[-2]):
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return True
return False
def get_dn(self, *keys, **options):
if self.is_pkey_zone_record(*keys):
Improve dnszone-add error message When a new DNS record is being added to DNS zone via command ipa dnsrecord-add ZONE @ and the target ZONE does not exist it returns ObjectclassViolation which may confuse users. Make sure that standard DNS Zone NotFound exception is returned. https://fedorahosted.org/freeipa/ticket/2270
2012-02-02 06:40:26 -06:00
dn = self.api.Object[self.parent_object].get_dn(*keys[:-1], **options)
# zone must exist
ldap = self.api.Backend.ldap2
try:
(dn_, zone) = ldap.get_entry(dn, [])
except errors.NotFound:
self.api.Object['dnszone'].handle_not_found(keys[-2])
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return self.api.Object[self.parent_object].get_dn(*keys[:-1], **options)
return super(dnsrecord, self).get_dn(*keys, **options)
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
def attr_to_cli(self, attr):
try:
cliname = attr[:-len('record')].upper()
except IndexError:
cliname = attr
return cliname
Improve DNS zone creation When a new DNS zone is being created a local hostname is set as a nameserver of the new zone. However, when the zone is created during ipa-replica-prepare, the the current master/replica doesn't have to be an IPA server with DNS support. This would lead to DNS zones with incorrect NS records as they wouldn't point to a valid name server. Now, a list of all master servers with DNS support is retrieved during DNS zone creation and added as NS records for a new DNS zone. https://fedorahosted.org/freeipa/ticket/1261
2011-06-01 07:51:06 -05:00
def get_dns_masters(self):
ldap = self.api.Backend.ldap2
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
base_dn = DN(('cn', 'masters'), ('cn', 'ipa'), ('cn', 'etc'), self.api.env.basedn)
Improve DNS zone creation When a new DNS zone is being created a local hostname is set as a nameserver of the new zone. However, when the zone is created during ipa-replica-prepare, the the current master/replica doesn't have to be an IPA server with DNS support. This would lead to DNS zones with incorrect NS records as they wouldn't point to a valid name server. Now, a list of all master servers with DNS support is retrieved during DNS zone creation and added as NS records for a new DNS zone. https://fedorahosted.org/freeipa/ticket/1261
2011-06-01 07:51:06 -05:00
ldap_filter = '(&(objectClass=ipaConfigObject)(cn=DNS))'
dns_masters = []
try:
entries = ldap.find_entries(filter=ldap_filter, base_dn=base_dn)[0]
for entry in entries:
master_dn = entry[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
assert isinstance(master_dn, DN)
try:
master = master_dn[1]['cn']
Improve DNS zone creation When a new DNS zone is being created a local hostname is set as a nameserver of the new zone. However, when the zone is created during ipa-replica-prepare, the the current master/replica doesn't have to be an IPA server with DNS support. This would lead to DNS zones with incorrect NS records as they wouldn't point to a valid name server. Now, a list of all master servers with DNS support is retrieved during DNS zone creation and added as NS records for a new DNS zone. https://fedorahosted.org/freeipa/ticket/1261
2011-06-01 07:51:06 -05:00
dns_masters.append(master)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
except (IndexError, KeyError):
pass
Improve DNS zone creation When a new DNS zone is being created a local hostname is set as a nameserver of the new zone. However, when the zone is created during ipa-replica-prepare, the the current master/replica doesn't have to be an IPA server with DNS support. This would lead to DNS zones with incorrect NS records as they wouldn't point to a valid name server. Now, a list of all master servers with DNS support is retrieved during DNS zone creation and added as NS records for a new DNS zone. https://fedorahosted.org/freeipa/ticket/1261
2011-06-01 07:51:06 -05:00
except errors.NotFound:
return []
return dns_masters
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def has_cli_options(self, options, no_option_msg, allow_empty_attrs=False):
Add rename option for DNS records This option will make renaming DNS records much easier. Add a unit test for this new functionality. https://fedorahosted.org/freeipa/ticket/2600
2012-05-29 08:58:36 -05:00
if any(k in options for k in ('setattr', 'addattr', 'delattr', 'rename')):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
has_options = False
for attr in options.keys():
if attr in self.params and not self.params[attr].primary_key:
if options[attr] or allow_empty_attrs:
has_options = True
break
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if not has_options:
raise errors.OptionError(no_option_msg)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def get_record_entry_attrs(self, entry_attrs):
return dict((attr, val) for attr,val in entry_attrs.iteritems() \
if attr in self.params and not self.params[attr].primary_key)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def postprocess_record(self, record, **options):
if options.get('structured', False):
for attr in record.keys():
try:
param = self.params[attr]
except KeyError:
continue
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
if not isinstance(param, DNSRecord):
continue
parts_params = param.get_parts()
for dnsvalue in record[attr]:
dnsentry = {
u'dnstype' : unicode(param.rrtype),
u'dnsdata' : dnsvalue
}
values = param._get_part_values(dnsvalue)
Fix TXT record parsing TXT record validation fails to parse the record if it contains spaces. Standard DNS part parser uses a space to divide record parts. A special parser thus need to be implemented for this RR type. https://fedorahosted.org/freeipa/ticket/2306
2012-02-03 03:01:27 -06:00
if values is None:
continue
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
for val_id, val in enumerate(values):
if val is not None:
dnsentry[parts_params[val_id].name] = val
record.setdefault('dnsrecords', []).append(dnsentry)
del record[attr]
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
def get_rrparam_from_part(self, part_name):
"""
Get an instance of DNSRecord parameter that has part_name as its part.
If such parameter is not found, None is returned
:param part_name Part parameter name
"""
try:
param = self.params[part_name]
if not any(flag in param.flags for flag in \
('dnsrecord_part', 'dnsrecord_extra')):
return None
# All DNS record part or extra parameters contain a name of its
# parent RR parameter in its hint attribute
rrparam = self.params[param.hint]
except (KeyError, AttributeError):
return None
return rrparam
def iterate_rrparams_by_parts(self, kw, skip_extra=False):
"""
Iterates through all DNSRecord instances that has at least one of its
parts or extra options in given dictionary. It returns the DNSRecord
instance only for the first occurence of part/extra option.
:param kw Dictionary with DNS record parts or extra options
:param skip_extra Skip DNS record extra options, yield only DNS records
with a real record part
"""
processed = []
for opt in kw:
rrparam = self.get_rrparam_from_part(opt)
if rrparam is None:
continue
if skip_extra and 'dnsrecord_extra' in self.params[opt].flags:
continue
if rrparam.name not in processed:
Fix dnsrecord_add interactive mode dnsrecord_add interactive mode did not work correctly when more than one DNS record part was entered as command line option. It asked for remaining options more than once. This patch fixes this situation and also adds tests to cover this use case properly. https://fedorahosted.org/freeipa/ticket/2641
2012-04-16 04:00:00 -05:00
processed.append(rrparam.name)
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
yield rrparam
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
def check_record_type_collisions(self, old_entry, entry_attrs):
# Test that only allowed combination of record types was created
attrs = set(attr for attr in entry_attrs if attr in _record_attributes
and entry_attrs[attr])
attrs.update(attr for attr in old_entry if attr not in entry_attrs)
try:
attrs.remove('cnamerecord')
except KeyError:
rec_has_cname = False
else:
rec_has_cname = True
# CNAME and PTR record combination is allowed
attrs.discard('ptrrecord')
rec_has_other_types = True if attrs else False
if rec_has_cname and rec_has_other_types:
raise errors.ValidationError(name='cnamerecord',
error=_('CNAME record is not allowed to coexist with any other '
'records except PTR'))
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
api.register(dnsrecord)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class dnsrecord_add(LDAPCreate):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Add new DNS resource record.')
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
no_option_msg = 'No options to add a specific record provided.\n' \
"Command help may be consulted for all supported record types."
Add an address for a nameserver when a new zone is created during install https://fedorahosted.org/freeipa/ticket/881
2011-01-31 11:05:07 -06:00
takes_options = LDAPCreate.takes_options + (
Flag('force',
Need force option in DNS zone adder dialog When adding a new DNS zone in the WebUI, IPA server will verify whether the nameserver is in DNS. Sometimes it is necessary to skip the verification. This patch adds a --force option already available in CLI which can skip this the verification. https://fedorahosted.org/freeipa/ticket/1105
2011-04-01 10:09:07 -05:00
label=_('Force'),
Add an address for a nameserver when a new zone is created during install https://fedorahosted.org/freeipa/ticket/881
2011-01-31 11:05:07 -06:00
flags=['no_option', 'no_output'],
doc=_('force NS record creation even if its hostname is not in DNS'),
),
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
dnsrecord.structured_flag,
Add an address for a nameserver when a new zone is created during install https://fedorahosted.org/freeipa/ticket/881
2011-01-31 11:05:07 -06:00
)
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
def args_options_2_entry(self, *keys, **options):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
self.obj.has_cli_options(options, self.no_option_msg)
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
return super(dnsrecord_add, self).args_options_2_entry(*keys, **options)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
def interactive_prompt_callback(self, kw):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
try:
self.obj.has_cli_options(kw, self.no_option_msg)
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
# Some DNS records were entered, do not use full interactive help
# We should still ask user for required parts of DNS parts he is
# trying to add in the same way we do for standard LDAP parameters
#
# Do not ask for required parts when any "extra" option is used,
# it can be used to fill all required params by itself
new_kw = {}
for rrparam in self.obj.iterate_rrparams_by_parts(kw, skip_extra=True):
user_options = rrparam.prompt_missing_parts(self.Backend, kw,
prompt_optional=False)
new_kw.update(user_options)
kw.update(new_kw)
return
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
except errors.OptionError:
pass
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
# check zone type
if kw['idnsname'] == _dns_zone_record:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
common_types = u', '.join(_zone_top_record_types)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
elif zone_is_reverse(kw['dnszoneidnsname']):
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
common_types = u', '.join(_rev_top_record_types)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
else:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
common_types = u', '.join(_top_record_types)
self.Backend.textui.print_plain(_(u'Please choose a type of DNS resource record to be added'))
self.Backend.textui.print_plain(_(u'The most common types for this type of zone are: %s\n') %\
common_types)
ok = False
while not ok:
rrtype = self.Backend.textui.prompt(_(u'DNS resource record type'))
if rrtype is None:
return
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
try:
name = '%srecord' % rrtype.lower()
param = self.params[name]
if not isinstance(param, DNSRecord):
raise ValueError()
Improve dnsrecord-add interactive mode When an invalid record type is entered during dnsrecord-add interactive mode, user is provided with a list of allowed values (record types). However, the provided list contains also unsupported record types (APL, DHCID, etc.) and any attempt to add such records would end with error. This patch limits the list to supported record types only. https://fedorahosted.org/freeipa/ticket/2378
2012-02-28 01:48:08 -06:00
if not param.supported:
raise ValueError()
except (KeyError, ValueError):
all_types = u', '.join(_dns_supported_record_types)
self.Backend.textui.print_plain(_(u'Invalid or unsupported type. Allowed values are: %s') % all_types)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
continue
ok = True
user_options = param.prompt_parts(self.Backend)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
kw.update(user_options)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def pre_callback(self, ldap, dn, entry_attrs, attrs_list, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
precallback_attrs = []
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
processed_attrs = []
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
for option in options:
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
try:
param = self.params[option]
except KeyError:
continue
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
rrparam = self.obj.get_rrparam_from_part(option)
if rrparam is None:
continue
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
if 'dnsrecord_part' in param.flags:
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
if rrparam.name in processed_attrs:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
# this record was already entered
continue
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
if rrparam.name in entry_attrs:
# this record is entered both via parts and raw records
raise errors.ValidationError(name=param.cli_name or param.name,
error=_('Raw value of a DNS record was already set by "%(name)s" option') \
% dict(name=rrparam.cli_name or rrparam.name))
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
parts = rrparam.get_parts_from_kw(options)
dnsvalue = [rrparam._convert_scalar(parts)]
entry_attrs[rrparam.name] = dnsvalue
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
processed_attrs.append(rrparam.name)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
continue
if 'dnsrecord_extra' in param.flags:
# do not run precallback for unset flags
if isinstance(param, Flag) and not options[option]:
continue
# extra option is passed, run per-type pre_callback for given RR type
Improve dnsrecord interactive help Add 2 new features to DNS record interactive help to increase its usability and also make its behavior more consistent with standard parameter interactive help: 1) Ask for missing DNS parts When a required part of a newly added DNS record was missing, we just returned a ValidationError. Now, the interactive help rather asks for all missing required parts of all DNS records that were being added by its parts. 2) Let user amend invalid part When an interactive help asked for a DNS record part value and user enters an invalid value, the entire interactive help exits with an error. This may upset a user if he already entered several correct DNS record part values. Now, the help rather tells user what's wrong and give him an opportunity to amend the value. https://fedorahosted.org/freeipa/ticket/2386
2012-02-28 08:04:05 -06:00
precallback_attrs.append(rrparam.name)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
# Run pre_callback validators
self.obj.run_precallback_validators(dn, entry_attrs, *keys, **options)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
# run precallback also for all new RR type attributes in entry_attrs
for attr in entry_attrs:
try:
param = self.params[attr]
except KeyError:
continue
if not isinstance(param, DNSRecord):
continue
precallback_attrs.append(attr)
precallback_attrs = list(set(precallback_attrs))
for attr in precallback_attrs:
# run per-type
try:
param = self.params[attr]
except KeyError:
continue
param.dnsrecord_add_pre_callback(ldap, dn, entry_attrs, attrs_list, *keys, **options)
# Store all new attrs so that DNSRecord post callback is called for
# new attributes only and not for all attributes in the LDAP entry
setattr(context, 'dnsrecord_precallback_attrs', precallback_attrs)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
# We always want to retrieve all DNS record attributes to test for
# record type collisions (#2601)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
try:
(dn_, old_entry) = ldap.get_entry(
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
dn, _record_attributes,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
normalize=self.obj.normalize_dn)
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
except errors.NotFound:
pass
else:
for attr in entry_attrs:
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if attr not in _record_attributes:
continue
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
if entry_attrs[attr] is None:
entry_attrs[attr] = []
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if not isinstance(entry_attrs[attr], (tuple, list)):
vals = [entry_attrs[attr]]
else:
vals = list(entry_attrs[attr])
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
entry_attrs[attr] = list(set(old_entry.get(attr, []) + vals))
self.obj.check_record_type_collisions(old_entry, entry_attrs)
Enforce that all NS records are resolvable Bind cannot load a zone if any of its name server records is not resolvable. https://fedorahosted.org/freeipa/ticket/838
2011-01-27 10:16:22 -06:00
return dn
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
def exc_callback(self, keys, options, exc, call_func, *call_args, **call_kwargs):
if call_func.func_name == 'add_entry':
if isinstance(exc, errors.DuplicateEntry):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
# A new record is being added to existing LDAP DNS object
# Update can be safely run as old record values has been
# already merged in pre_callback
ldap = self.obj.backend
dn = call_args[0]
entry_attrs = self.obj.get_record_entry_attrs(call_args[1])
ldap.update_entry(dn, entry_attrs, **call_kwargs)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return
raise exc
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
for attr in getattr(context, 'dnsrecord_precallback_attrs', []):
param = self.params[attr]
param.dnsrecord_add_post_callback(ldap, dn, entry_attrs, *keys, **options)
Return correct record name in DNS plugin When dnsrecord-add or dnsrecord-mod commands are used on a root zone record (it has a special name "@"), a zone name is returned instead of a special name "@". This confuses DNS part of Web UI which is then not able to manipulate records in the root zone when these commands are used. This patch fixes these 2 commands to return correct value when a root zone is modified. https://fedorahosted.org/freeipa/ticket/2627 https://fedorahosted.org/freeipa/ticket/2628
2012-04-12 02:44:50 -05:00
if self.obj.is_pkey_zone_record(*keys):
entry_attrs[self.obj.primary_key.name] = [_dns_zone_record]
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
self.obj.postprocess_record(entry_attrs, **options)
return dn
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnsrecord_add)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class dnsrecord_mod(LDAPUpdate):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Modify a DNS resource record.')
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
no_option_msg = 'No options to modify a specific record provided.'
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
takes_options = LDAPUpdate.takes_options + (
dnsrecord.structured_flag,
)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def args_options_2_entry(self, *keys, **options):
self.obj.has_cli_options(options, self.no_option_msg, True)
return super(dnsrecord_mod, self).args_options_2_entry(*keys, **options)
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
def pre_callback(self, ldap, dn, entry_attrs, attrs_list, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add rename option for DNS records This option will make renaming DNS records much easier. Add a unit test for this new functionality. https://fedorahosted.org/freeipa/ticket/2600
2012-05-29 08:58:36 -05:00
if options.get('rename') and self.obj.is_pkey_zone_record(*keys):
# zone rename is not allowed
raise errors.ValidationError(name='rename',
error=_('DNS zone root record cannot be renamed'))
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
# check if any attr should be updated using structured instead of replaced
# format is recordname : (old_value, new_parts)
updated_attrs = {}
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
for param in self.obj.iterate_rrparams_by_parts(options, skip_extra=True):
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
parts = param.get_parts_from_kw(options, raise_on_none=False)
if parts is None:
# old-style modification
continue
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
old_value = entry_attrs.get(param.name)
if not old_value:
raise errors.RequirementError(name=param.name)
if isinstance(old_value, (tuple, list)):
if len(old_value) > 1:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
raise errors.ValidationError(name=param.name,
error=_('DNS records can be only updated one at a time'))
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
old_value = old_value[0]
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Harden raw record processing in DNS plugin There were cases where DNS plugin was too tolerant in a raw DNS record option (--<rrtype-rec) processing. It let people specify DNS record parts options in dnsrecord-mod operations for some record without specifying the record that should be updated. It also ignored DNS record parts in dnsrecord-add operation when the raw DNS record value was already set via --<rrtype>-rec option. This patch hardens the processing and returns error in both described cases to make the processes clearer and more robust. All these use cases were also covered by new unit tests. https://fedorahosted.org/freeipa/ticket/2551
2012-03-22 11:35:39 -05:00
updated_attrs[param.name] = (old_value, parts)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Fix precallback validators in DNS plugin DNS plugin contains several RR type record validators run in pre_callback which cannot be used as standard param validator as it needs more data and resources that standard validators provide. However, the precallback validators are not run for DNS records created by new structured options and thus an invalid value may slip in. This patch moves the execution of these precallback validators _after_ the processing of structured DNS options. It also cleans them up a little and makes them more robust. https://fedorahosted.org/freeipa/ticket/2550
2012-03-21 07:25:42 -05:00
# Run pre_callback validators
self.obj.run_precallback_validators(dn, entry_attrs, *keys, **options)
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
# current entry is needed in case of per-dns-record-part updates and
# for record type collision check
try:
(dn_, old_entry) = ldap.get_entry(dn, _record_attributes,
normalize=self.obj.normalize_dn)
except errors.NotFound:
if updated_attrs:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
self.obj.handle_not_found(*keys)
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
if updated_attrs:
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
for attr in updated_attrs:
param = self.params[attr]
old_dnsvalue, new_parts = updated_attrs[attr]
if old_dnsvalue not in old_entry.get(attr, []):
Improve dns error message Improve AttrValueNotFound exception error message raised in the DNS module when a deleted (modified) attribute value is not found in a record. In order to be consistent with previous DNS module implementation this error message should include an attribute label instead of an attribute name. https://fedorahosted.org/freeipa/ticket/2377
2012-02-21 03:37:08 -06:00
attr_name = unicode(param.label or param.name)
raise errors.AttrValueNotFound(attr=attr_name,
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
value=old_dnsvalue)
old_entry[attr].remove(old_dnsvalue)
old_parts = param._get_part_values(old_dnsvalue)
modified_parts = tuple(part if part is not None else old_parts[part_id] \
for part_id,part in enumerate(new_parts))
new_dnsvalue = [param._convert_scalar(modified_parts)]
entry_attrs[attr] = list(set(old_entry[attr] + new_dnsvalue))
Enforce CNAME constrains for DNS commands RFC 1912 states that no record (besides PTR) is allowed to coexist with any other record type. When BIND detects this situation, it refuses to load such records. Enforce the constrain for dnsrecord-mod and dnsrecord-add commands. https://fedorahosted.org/freeipa/ticket/2601
2012-07-13 05:23:38 -05:00
self.obj.check_record_type_collisions(old_entry, entry_attrs)
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
return dn
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def execute(self, *keys, **options):
result = super(dnsrecord_mod, self).execute(*keys, **options)
# remove if empty
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
if not self.obj.is_pkey_zone_record(*keys):
Add rename option for DNS records This option will make renaming DNS records much easier. Add a unit test for this new functionality. https://fedorahosted.org/freeipa/ticket/2600
2012-05-29 08:58:36 -05:00
rename = options.get('rename')
if rename is not None:
keys = keys[:-1] + (rename,)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
dn = self.obj.get_dn(*keys, **options)
ldap = self.obj.backend
(dn_, old_entry) = ldap.get_entry(
dn, _record_attributes,
normalize=self.obj.normalize_dn)
del_all = True
for attr in old_entry:
if old_entry[attr]:
del_all = False
break
if del_all:
return self.obj.methods.delentry(*keys)
return result
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Return correct record name in DNS plugin When dnsrecord-add or dnsrecord-mod commands are used on a root zone record (it has a special name "@"), a zone name is returned instead of a special name "@". This confuses DNS part of Web UI which is then not able to manipulate records in the root zone when these commands are used. This patch fixes these 2 commands to return correct value when a root zone is modified. https://fedorahosted.org/freeipa/ticket/2627 https://fedorahosted.org/freeipa/ticket/2628
2012-04-12 02:44:50 -05:00
if self.obj.is_pkey_zone_record(*keys):
entry_attrs[self.obj.primary_key.name] = [_dns_zone_record]
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
self.obj.postprocess_record(entry_attrs, **options)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
return dn
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def interactive_prompt_callback(self, kw):
try:
self.obj.has_cli_options(kw, self.no_option_msg, True)
except errors.OptionError:
pass
else:
# some record type entered, skip this helper
return
# get DNS record first so that the NotFound exception is raised
# before the helper would start
dns_record = api.Command['dnsrecord_show'](kw['dnszoneidnsname'], kw['idnsname'])['result']
rec_types = [rec_type for rec_type in dns_record if rec_type in _record_attributes]
self.Backend.textui.print_plain(_("No option to modify specific record provided."))
# ask user for records to be removed
self.Backend.textui.print_plain(_(u'Current DNS record contents:\n'))
record_params = []
for attr in dns_record:
try:
param = self.params[attr]
except KeyError:
continue
if not isinstance(param, DNSRecord):
continue
record_params.append(param)
rec_type_content = u', '.join(dns_record[param.name])
self.Backend.textui.print_plain(u'%s: %s' % (param.label, rec_type_content))
self.Backend.textui.print_plain(u'')
# ask what records to remove
for param in record_params:
rec_values = list(dns_record[param.name])
for rec_value in dns_record[param.name]:
rec_values.remove(rec_value)
mod_value = self.Backend.textui.prompt_yesno(
_("Modify %(name)s '%(value)s'?") % dict(name=param.label, value=rec_value), default=False)
if mod_value is True:
user_options = param.prompt_parts(self.Backend, mod_dnsvalue=rec_value)
kw[param.name] = [rec_value]
kw.update(user_options)
if rec_values:
self.Backend.textui.print_plain(ngettext(
u'%(count)d %(type)s record skipped. Only one value per DNS record type can be modified at one time.',
u'%(count)d %(type)s records skipped. Only one value per DNS record type can be modified at one time.',
0) % dict(count=len(rec_values), type=param.rrtype))
break
Add DNS record modification command The DNS record plugin does not support modification of a record. One can only add A type addresses to a DNS record or remove the current ones. To actually change a DNS record value it has to be removed and then added with a desired value. This patch adds a new DNS plugin command "dnsrecord-mod" which enables user to: - modify a DNS record value (note than DNS record can hold multiple values and those will be overwritten) - remove a DNS record when an empty value is passed New tests for this new command have been added to the CLI test suite. https://fedorahosted.org/freeipa/ticket/1137
2011-07-12 02:11:00 -05:00
api.register(dnsrecord_mod)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
class dnsrecord_delentry(LDAPDelete):
"""
Delete DNS record entry.
"""
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
msg_summary = _('Deleted record "%(value)s"')
Rename INTERNAL to NO_CLI for commands we hide from the cli. Also make i18n_messages and json_metadata NO_CLI. ticket 821
2011-01-20 14:07:43 -06:00
NO_CLI = True
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnsrecord_delentry)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class dnsrecord_del(LDAPUpdate):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Delete DNS resource record.')
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
no_option_msg = _('Neither --del-all nor options to delete a specific record provided.\n'\
"Command help may be consulted for all supported record types.")
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
takes_options = (
Flag('del_all',
default=False,
label=_('Delete all associated records'),
),
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
dnsrecord.structured_flag,
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def get_options(self):
for option in super(dnsrecord_del, self).get_options():
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
if any(flag in option.flags for flag in \
('dnsrecord_part', 'dnsrecord_extra',)):
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
continue
Add rename option for DNS records This option will make renaming DNS records much easier. Add a unit test for this new functionality. https://fedorahosted.org/freeipa/ticket/2600
2012-05-29 08:58:36 -05:00
elif option.name in ('rename', ):
# options only valid for dnsrecord-mod
continue
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
elif isinstance(option, DNSRecord):
yield option.clone(option_group=None)
continue
yield option
Avoid deleting DNS zone when a context is reused When dnsrecord-del pre_callback detects that the record does not contain any records, it sets a flag to connection context and deletes the record object later. However, when more dnsrecord-del commands share the same context (and this is the case of "ipa-replica-manage del $MASTER" DNS cleanup), it may reuse a positive flag from previous dnsrecord-del command and delete the root DNS zone record and thus effectively delete the zone. This patch makes sure that this flag is always initialized to a sane value in dnsrecord-del pre_callback to make sure that the DNS zone is not deleted. It also fixes pre_callback function definition to prevent adding attrs_list to "keys" parameter and thus confuse developers. https://fedorahosted.org/freeipa/ticket/2503
2012-03-23 04:29:30 -05:00
def pre_callback(self, ldap, dn, entry_attrs, attrs_list, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
try:
(dn_, old_entry) = ldap.get_entry(
dn, _record_attributes,
normalize=self.obj.normalize_dn)
except errors.NotFound:
self.obj.handle_not_found(*keys)
for attr in entry_attrs.keys():
if attr not in _record_attributes:
continue
if not isinstance(entry_attrs[attr], (tuple, list)):
vals = [entry_attrs[attr]]
else:
vals = entry_attrs[attr]
for val in vals:
try:
old_entry[attr].remove(val)
except (KeyError, ValueError):
Improve dns error message Improve AttrValueNotFound exception error message raised in the DNS module when a deleted (modified) attribute value is not found in a record. In order to be consistent with previous DNS module implementation this error message should include an attribute label instead of an attribute name. https://fedorahosted.org/freeipa/ticket/2377
2012-02-21 03:37:08 -06:00
try:
param = self.params[attr]
attr_name = unicode(param.label or param.name)
except:
attr_name = attr
raise errors.AttrValueNotFound(attr=attr_name,
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
value=val)
entry_attrs[attr] = list(set(old_entry[attr]))
Avoid deleting DNS zone when a context is reused When dnsrecord-del pre_callback detects that the record does not contain any records, it sets a flag to connection context and deletes the record object later. However, when more dnsrecord-del commands share the same context (and this is the case of "ipa-replica-manage del $MASTER" DNS cleanup), it may reuse a positive flag from previous dnsrecord-del command and delete the root DNS zone record and thus effectively delete the zone. This patch makes sure that this flag is always initialized to a sane value in dnsrecord-del pre_callback to make sure that the DNS zone is not deleted. It also fixes pre_callback function definition to prevent adding attrs_list to "keys" parameter and thus confuse developers. https://fedorahosted.org/freeipa/ticket/2503
2012-03-23 04:29:30 -05:00
del_all = False
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if not self.obj.is_pkey_zone_record(*keys):
Avoid deleting DNS zone when a context is reused When dnsrecord-del pre_callback detects that the record does not contain any records, it sets a flag to connection context and deletes the record object later. However, when more dnsrecord-del commands share the same context (and this is the case of "ipa-replica-manage del $MASTER" DNS cleanup), it may reuse a positive flag from previous dnsrecord-del command and delete the root DNS zone record and thus effectively delete the zone. This patch makes sure that this flag is always initialized to a sane value in dnsrecord-del pre_callback to make sure that the DNS zone is not deleted. It also fixes pre_callback function definition to prevent adding attrs_list to "keys" parameter and thus confuse developers. https://fedorahosted.org/freeipa/ticket/2503
2012-03-23 04:29:30 -05:00
record_found = False
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
for attr in old_entry:
if old_entry[attr]:
Avoid deleting DNS zone when a context is reused When dnsrecord-del pre_callback detects that the record does not contain any records, it sets a flag to connection context and deletes the record object later. However, when more dnsrecord-del commands share the same context (and this is the case of "ipa-replica-manage del $MASTER" DNS cleanup), it may reuse a positive flag from previous dnsrecord-del command and delete the root DNS zone record and thus effectively delete the zone. This patch makes sure that this flag is always initialized to a sane value in dnsrecord-del pre_callback to make sure that the DNS zone is not deleted. It also fixes pre_callback function definition to prevent adding attrs_list to "keys" parameter and thus confuse developers. https://fedorahosted.org/freeipa/ticket/2503
2012-03-23 04:29:30 -05:00
record_found = True
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
break
Avoid deleting DNS zone when a context is reused When dnsrecord-del pre_callback detects that the record does not contain any records, it sets a flag to connection context and deletes the record object later. However, when more dnsrecord-del commands share the same context (and this is the case of "ipa-replica-manage del $MASTER" DNS cleanup), it may reuse a positive flag from previous dnsrecord-del command and delete the root DNS zone record and thus effectively delete the zone. This patch makes sure that this flag is always initialized to a sane value in dnsrecord-del pre_callback to make sure that the DNS zone is not deleted. It also fixes pre_callback function definition to prevent adding attrs_list to "keys" parameter and thus confuse developers. https://fedorahosted.org/freeipa/ticket/2503
2012-03-23 04:29:30 -05:00
del_all = not record_found
# set del_all flag in context
# when the flag is enabled, the entire DNS record object is deleted
# in a post callback
setattr(context, 'del_all', del_all)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return dn
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
def execute(self, *keys, **options):
if options.get('del_all', False):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
if self.obj.is_pkey_zone_record(*keys):
raise errors.ValidationError(
name='del_all',
error=_('Zone record \'%s\' cannot be deleted') \
% _dns_zone_record
)
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
return self.obj.methods.delentry(*keys)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
result = super(dnsrecord_del, self).execute(*keys, **options)
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
Avoid deleting DNS zone when a context is reused When dnsrecord-del pre_callback detects that the record does not contain any records, it sets a flag to connection context and deletes the record object later. However, when more dnsrecord-del commands share the same context (and this is the case of "ipa-replica-manage del $MASTER" DNS cleanup), it may reuse a positive flag from previous dnsrecord-del command and delete the root DNS zone record and thus effectively delete the zone. This patch makes sure that this flag is always initialized to a sane value in dnsrecord-del pre_callback to make sure that the DNS zone is not deleted. It also fixes pre_callback function definition to prevent adding attrs_list to "keys" parameter and thus confuse developers. https://fedorahosted.org/freeipa/ticket/2503
2012-03-23 04:29:30 -05:00
if getattr(context, 'del_all', False) and not \
self.obj.is_pkey_zone_record(*keys):
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
return self.obj.methods.delentry(*keys)
return result
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
def post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
if self.obj.is_pkey_zone_record(*keys):
entry_attrs[self.obj.primary_key.name] = [_dns_zone_record]
self.obj.postprocess_record(entry_attrs, **options)
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
return dn
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
def args_options_2_entry(self, *keys, **options):
self.obj.has_cli_options(options, self.no_option_msg)
return super(dnsrecord_del, self).args_options_2_entry(*keys, **options)
Delete the whole DNS record with no parameters Error out when deleting a nonexistent DNS record Also fixes the DNS unit tests. https://fedorahosted.org/freeipa/ticket/816 https://fedorahosted.org/freeipa/ticket/829
2011-01-20 06:54:14 -06:00
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
def interactive_prompt_callback(self, kw):
if kw.get('del_all', False):
return
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
try:
self.obj.has_cli_options(kw, self.no_option_msg)
except errors.OptionError:
pass
else:
# some record type entered, skip this helper
return
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
# get DNS record first so that the NotFound exception is raised
# before the helper would start
dns_record = api.Command['dnsrecord_show'](kw['dnszoneidnsname'], kw['idnsname'])['result']
rec_types = [rec_type for rec_type in dns_record if rec_type in _record_attributes]
self.Backend.textui.print_plain(_("No option to delete specific record provided."))
user_del_all = self.Backend.textui.prompt_yesno(_("Delete all?"), default=False)
if user_del_all is True:
kw['del_all'] = True
return
# ask user for records to be removed
self.Backend.textui.print_plain(_(u'Current DNS record contents:\n'))
present_params = []
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
for attr in dns_record:
try:
param = self.params[attr]
except KeyError:
continue
if not isinstance(param, DNSRecord):
continue
present_params.append(param)
rec_type_content = u', '.join(dns_record[param.name])
self.Backend.textui.print_plain(u'%s: %s' % (param.label, rec_type_content))
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
self.Backend.textui.print_plain(u'')
# ask what records to remove
for param in present_params:
deleted_values = []
for rec_value in dns_record[param.name]:
user_del_value = self.Backend.textui.prompt_yesno(
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
_("Delete %(name)s '%(value)s'?") \
% dict(name=param.label, value=rec_value), default=False)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
if user_del_value is True:
deleted_values.append(rec_value)
if deleted_values:
Fix dnsrecord-del interactive mode User was not able to select multiple values in dnsrecord-del interactive mode. https://fedorahosted.org/freeipa/ticket/2581
2012-03-27 06:02:35 -05:00
kw[param.name] = tuple(deleted_values)
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnsrecord_del)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class dnsrecord_show(LDAPRetrieve):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Display DNS resource.')
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
takes_options = LDAPRetrieve.takes_options + (
dnsrecord.structured_flag,
)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
def post_callback(self, ldap, dn, entry_attrs, *keys, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(dn, DN)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
if self.obj.is_pkey_zone_record(*keys):
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
entry_attrs[self.obj.primary_key.name] = [_dns_zone_record]
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
self.obj.postprocess_record(entry_attrs, **options)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return dn
api.register(dnsrecord_show)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
class dnsrecord_find(LDAPSearch):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Search for DNS resources.')
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
takes_options = LDAPSearch.takes_options + (
dnsrecord.structured_flag,
)
Add reverse DNS record when forward is created Adding reverse DNS record may be a time consuming task, especially for IPv6 addresses. Having a way to automatically create a reverse record when a forward record is created could speed up the process. host-add command already has this possibility. This patch takes advantage of the new per-type API and adds new options for A/AAAA record types: --a-create-reverse and --aaaa-create-reverse. These commands can be used to automatically create reverse records for new A/AAAA addresses (both forward and reverse zones need to be managed by FreeIPA server): ipa dnsrecord-add example.com foo --a-rec=10.0.0.1 --a-create-reverse This command would add a new A record to record foo in zone example.com and a PTR record to appropriate reverse zone for IP address 10.0.0.1 (for example PTR record 1 in zone 0.0.10.in-addr.arpa. pointing to foo.example.com.). Few modification were done to new DNS API to support this feature: - Refactor --ip-address option handling from host-add and place it to dns.py to be used by both modules - Add support for "extra" per-type options - Hide DNS record part options in dnsrecord_find command as they have no effect for this command https://fedorahosted.org/freeipa/ticket/2009
2012-02-24 07:15:24 -06:00
def get_options(self):
for option in super(dnsrecord_find, self).get_options():
if any(flag in option.flags for flag in \
('dnsrecord_part', 'dnsrecord_extra',)):
continue
elif isinstance(option, DNSRecord):
yield option.clone(option_group=None)
continue
yield option
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
def pre_callback(self, ldap, filter, attrs_list, base_dn, scope, *args, **options):
Use DN objects instead of strings * Convert every string specifying a DN into a DN object * Every place a dn was manipulated in some fashion it was replaced by the use of DN operators * Add new DNParam parameter type for parameters which are DN's * DN objects are used 100% of the time throughout the entire data pipeline whenever something is logically a dn. * Many classes now enforce DN usage for their attributes which are dn's. This is implmented via ipautil.dn_attribute_property(). The only permitted types for a class attribute specified to be a DN are either None or a DN object. * Require that every place a dn is used it must be a DN object. This translates into lot of:: assert isinstance(dn, DN) sprinkled through out the code. Maintaining these asserts is valuable to preserve DN type enforcement. The asserts can be disabled in production. The goal of 100% DN usage 100% of the time has been realized, these asserts are meant to preserve that. The asserts also proved valuable in detecting functions which did not obey their function signatures, such as the baseldap pre and post callbacks. * Moved ipalib.dn to ipapython.dn because DN class is shared with all components, not just the server which uses ipalib. * All API's now accept DN's natively, no need to convert to str (or unicode). * Removed ipalib.encoder and encode/decode decorators. Type conversion is now explicitly performed in each IPASimpleLDAPObject method which emulates a ldap.SimpleLDAPObject method. * Entity & Entry classes now utilize DN's * Removed __getattr__ in Entity & Entity clases. There were two problems with it. It presented synthetic Python object attributes based on the current LDAP data it contained. There is no way to validate synthetic attributes using code checkers, you can't search the code to find LDAP attribute accesses (because synthetic attriutes look like Python attributes instead of LDAP data) and error handling is circumscribed. Secondly __getattr__ was hiding Python internal methods which broke class semantics. * Replace use of methods inherited from ldap.SimpleLDAPObject via IPAdmin class with IPAdmin methods. Directly using inherited methods was causing us to bypass IPA logic. Mostly this meant replacing the use of search_s() with getEntry() or getList(). Similarly direct access of the LDAP data in classes using IPAdmin were replaced with calls to getValue() or getValues(). * Objects returned by ldap2.find_entries() are now compatible with either the python-ldap access methodology or the Entity/Entry access methodology. * All ldap operations now funnel through the common IPASimpleLDAPObject giving us a single location where we interface to python-ldap and perform conversions. * The above 4 modifications means we've greatly reduced the proliferation of multiple inconsistent ways to perform LDAP operations. We are well on the way to having a single API in IPA for doing LDAP (a long range goal). * All certificate subject bases are now DN's * DN objects were enhanced thusly: - find, rfind, index, rindex, replace and insert methods were added - AVA, RDN and DN classes were refactored in immutable and mutable variants, the mutable variants are EditableAVA, EditableRDN and EditableDN. By default we use the immutable variants preserving important semantics. To edit a DN cast it to an EditableDN and cast it back to DN when done editing. These issues are fully described in other documentation. - first_key_match was removed - DN equalty comparison permits comparison to a basestring * Fixed ldapupdate to work with DN's. This work included: - Enhance test_updates.py to do more checking after applying update. Add test for update_from_dict(). Convert code to use unittest classes. - Consolidated duplicate code. - Moved code which should have been in the class into the class. - Fix the handling of the 'deleteentry' update action. It's no longer necessary to supply fake attributes to make it work. Detect case where subsequent update applies a change to entry previously marked for deletetion. General clean-up and simplification of the 'deleteentry' logic. - Rewrote a couple of functions to be clearer and more Pythonic. - Added documentation on the data structure being used. - Simplfy the use of update_from_dict() * Removed all usage of get_schema() which was being called prior to accessing the .schema attribute of an object. If a class is using internal lazy loading as an optimization it's not right to require users of the interface to be aware of internal optimization's. schema is now a property and when the schema property is accessed it calls a private internal method to perform the lazy loading. * Added SchemaCache class to cache the schema's from individual servers. This was done because of the observation we talk to different LDAP servers, each of which may have it's own schema. Previously we globally cached the schema from the first server we connected to and returned that schema in all contexts. The cache includes controls to invalidate it thus forcing a schema refresh. * Schema caching is now senstive to the run time context. During install and upgrade the schema can change leading to errors due to out-of-date cached schema. The schema cache is refreshed in these contexts. * We are aware of the LDAP syntax of all LDAP attributes. Every attribute returned from an LDAP operation is passed through a central table look-up based on it's LDAP syntax. The table key is the LDAP syntax it's value is a Python callable that returns a Python object matching the LDAP syntax. There are a handful of LDAP attributes whose syntax is historically incorrect (e.g. DistguishedNames that are defined as DirectoryStrings). The table driven conversion mechanism is augmented with a table of hard coded exceptions. Currently only the following conversions occur via the table: - dn's are converted to DN objects - binary objects are converted to Python str objects (IPA convention). - everything else is converted to unicode using UTF-8 decoding (IPA convention). However, now that the table driven conversion mechanism is in place it would be trivial to do things such as converting attributes which have LDAP integer syntax into a Python integer, etc. * Expected values in the unit tests which are a DN no longer need to use lambda expressions to promote the returned value to a DN for equality comparison. The return value is automatically promoted to a DN. The lambda expressions have been removed making the code much simpler and easier to read. * Add class level logging to a number of classes which did not support logging, less need for use of root_logger. * Remove ipaserver/conn.py, it was unused. * Consolidated duplicate code wherever it was found. * Fixed many places that used string concatenation to form a new string rather than string formatting operators. This is necessary because string formatting converts it's arguments to a string prior to building the result string. You can't concatenate a string and a non-string. * Simplify logic in rename_managed plugin. Use DN operators to edit dn's. * The live version of ipa-ldap-updater did not generate a log file. The offline version did, now both do. https://fedorahosted.org/freeipa/ticket/1670 https://fedorahosted.org/freeipa/ticket/1671 https://fedorahosted.org/freeipa/ticket/1672 https://fedorahosted.org/freeipa/ticket/1673 https://fedorahosted.org/freeipa/ticket/1674 https://fedorahosted.org/freeipa/ticket/1392 https://fedorahosted.org/freeipa/ticket/2872
2012-05-13 06:36:35 -05:00
assert isinstance(base_dn, DN)
Refactor dnsrecord processing Current DNS record processing architecture has many flaws, including custom execute() methods which does not take advantage of base LDAP commands or nonstandard and confusing DNS record option processing. This patch refactors DNS record processing with the following improvements: * Every DNS record has now own Parameter type. Each DNS record consists from one or more "parts" which are also Parameters. This architecture will enable much easier implementation of future per-DNS-type API. * Validation is now not written as a separate function for every parameter but is delegated to DNS record parts. * Normalization is also delegated to DNS record parts. * Since standard LDAP base commands execute method is now used, dnsrecord-add and dnsrecord-mod correctly supports --setattr and --addattr options. * In order to prevent confusion unsupported DNS record types are now hidden. They are still present in the plugin so that old clients receive proper validation error. The patch also contains several fixes: * Fix domain-name validation and normalization- allow domain names that are not fully qualified. For example --cname-rec=bar is a valid domain-name for bind which will translate it then as bar.<owning-domain>. This change implies, that fully qualified domain names must end with '.'. * Do not let user accidentally remove entire zone with command "ipa dnsrecord-del @ --del-all". * Fix --ttl and --class option processing in dnsrecord-add and dnsrecord-mod. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-06 08:12:41 -06:00
# include zone record (root entry) in the search
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
return (filter, base_dn, ldap.SCOPE_SUBTREE)
def post_callback(self, ldap, entries, truncated, *args, **options):
if entries:
zone_obj = self.api.Object[self.obj.parent_object]
zone_dn = zone_obj.get_dn(args[0])
if entries[0][0] == zone_dn:
Improve interactive mode for DNS plugin Interactive mode for commands manipulating with DNS records (dnsrecord-add, dnsrecord-del) is not usable. This patch enhances the server framework with new callback for interactive mode, which can be used by commands to inject their own interactive handling. The callback is then used to improve aforementioned commands' interactive mode. https://fedorahosted.org/freeipa/ticket/1018
2011-05-26 02:55:53 -05:00
entries[0][1][zone_obj.primary_key.name] = [_dns_zone_record]
Create per-type DNS API Use new structured DNSRecord parameters to generate per-type API for all supported DNS RR types. This should help significantly the end-user with manipulating complex DNS record type (MX, LOC, etc.). All enhancements are integrated to current DNS record commands: 1) dnsrecord-add - Records can be either entered as a raw value (e.g. --mx-rec= "1 srv1.example.com" for MX record) or per-part: --mx-preference=1 --mx-exchanger=srv1.example.com - CLI interactive help behavior was changed. It will ask for a record type and then ask for all DNS record part values (e.g. MX Preference value, MX Exchanger value). 2) dnsrecord-mod - This command can now operate in 2 modes. When only a raw DNS record is entered (e.g. --mx-rec="1 srv1.example.com") it operates in standard mode and replaces any previous mxrecord value with the --mx-rec value. When any structured parameter (e.g. --mx-preference) is passed it modifies just the specified parts of one mxrecord value referred by --mx-rec: --mx-rec="1 srv1.example.com" --mx-preference=2 - New interactive help has been implemented. It will ask for a record to be modified (in the same manner as dnsrecord-del) and then let user change DNS record part(s) for chosen records. 3) All dnsrecord-* commands have now --structured option - When this option is passed, instead of displaying raw DNS values all DNS records are parsed and displayed per-part. Example: $ ipa dnsrecord-show example.com @ --structured Record name: @ Records: Record type: MX Record data: 0 server1.example.com. MX Preference: 0 MX Exchanger: server1.example.com. Record type: NS Record data: ns1.example.com. NS Hostname: ns1.example.com. All API changes are compatible with clients without this patch. https://fedorahosted.org/freeipa/ticket/2082
2012-01-12 02:37:36 -06:00
for entry in entries:
self.obj.postprocess_record(entry[1], **options)
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
Enforce sizelimit in permission-find, post_callback returns truncated We actually perform two searches in permission-find. The first looks for matches within the permission object itself. The second looks at matches in the underlying aci. We need to break out in two places. The first is if we find enough matches in the permission itself. The second when we are appending matches from acis. The post_callback() definition needed to be modified to return the truncated value so a plugin author can modify that value. https://fedorahosted.org/freeipa/ticket/2322
2012-05-23 10:00:24 -05:00
return truncated
Add new version of DNS plugin: complete rework with baseldap + unit tests. Ticket #36 Ticket #450
2010-11-08 21:34:14 -06:00
api.register(dnsrecord_find)
Port installer and host plugin to the new DNS plugin * move ipa dns-resolve to the new plugin * port the installer and the host plugin to the new interface * remove the old plugin
2011-01-12 14:02:05 -06:00
class dns_resolve(Command):
ticket 1669 - improve i18n docstring extraction This patch reverts the use of pygettext for i18n string extraction. It was originally introduced because the help documentation for commands are in the class docstring and module docstring. Docstrings are a Python construct whereby any string which immediately follows a class declaration, function/method declaration or appears first in a module is taken to be the documentation for that object. Python automatically assigns that string to the __doc__ variable associated with the object. Explicitly assigning to the __doc__ variable is equivalent and permitted. We mark strings in the source for i18n translation by embedding them in _() or ngettext(). Specialized extraction tools (e.g. xgettext) scan the source code looking for strings with those markers and extracts the string for inclusion in a translation catalog. It was mistakingly assumed one could not mark for translation Python docstrings. Since some docstrings are vital for our command help system some method had to be devised to extract docstrings for the translation catalog. pygettext has the ability to locate and extract docstrings and it was introduced to acquire the documentation for our commands located in module and class docstrings. However pygettext was too large a hammer for this task, it lacked any fined grained ability to extract only the docstrings we were interested in. In practice it extracted EVERY docstring in each file it was presented with. This caused a large number strings to be extracted for translation which had no reason to be translated, the string might have been internal code documentation never meant to be seen by users. Often the superfluous docstrings were long, complex and likely difficult to translate. This placed an unnecessary burden on our volunteer translators. Instead what is needed is some method to extract only those strings intended for translation. We already have such a mechanism and it is already widely used, namely wrapping strings intended for translation in calls to _() or _negettext(), i.e. marking a string for i18n translation. Thus the solution to the docstring translation problem is to mark the docstrings exactly as we have been doing, it only requires that instead of a bare Python docstring we instead assign the marked string to the __doc__ variable. Using the hypothetical class foo as an example. class foo(Command): ''' The foo command takes out the garbage. ''' Would become: class foo(Command): __doc__ = _('The foo command takes out the garbage.') But which docstrings need to be marked for translation? The makeapi tool knows how to iterate over every command in our public API. It was extended to validate every command's documentation and report if any documentation is missing or not marked for translation. That information was then used to identify each docstring in the code which needed to be transformed. In summary what this patch does is: * Remove the use of pygettext (modification to install/po/Makefile.in) * Replace every docstring with an explicit assignment to __doc__ where the rhs of the assignment is an i18n marking function. * Single line docstrings appearing in multi-line string literals (e.g. ''' or """) were replaced with single line string literals because the multi-line literals were introducing unnecessary whitespace and newlines in the string extracted for translation. For example: ''' The foo command takes out the garbage. ''' Would appear in the translation catalog as: "\n The foo command takes out the garbage.\n " The superfluous whitespace and newlines are confusing to translators and requires us to strip leading and trailing whitespace from the translation at run time. * Import statements were moved from below the docstring to above it. This was necessary because the i18n markers are imported functions and must be available before the the doc is parsed. Technically only the import of the i18n markers had to appear before the doc but stylistically it's better to keep all the imports together. * It was observed during the docstring editing process that the command documentation was inconsistent with respect to the use of periods to terminate a sentence. Some doc had a trailing period, others didn't. Consistency was enforced by adding a period to end of every docstring if one was missing.
2011-08-24 21:48:30 -05:00
__doc__ = _('Resolve a host name in DNS.')
Port installer and host plugin to the new DNS plugin * move ipa dns-resolve to the new plugin * port the installer and the host plugin to the new interface * remove the old plugin
2011-01-12 14:02:05 -06:00
has_output = output.standard_value
msg_summary = _('Found \'%(value)s\'')
takes_args = (
Str('hostname',
label=_('Hostname'),
),
)
def execute(self, *args, **options):
query=args[0]
if query.find(api.env.domain) == -1 and query.find('.') == -1:
query = '%s.%s.' % (query, api.env.domain)
if query[-1] != '.':
query = query + '.'
Replace DNS client based on acutil with python-dns IPA client and server tool set used authconfig acutil module to for client DNS operations. This is not optimal DNS interface for several reasons: - does not provide native Python object oriented interface but but rather C-like interface based on functions and structures which is not easy to use and extend - acutil is not meant to be used by third parties besides authconfig and thus can break without notice Replace the acutil with python-dns package which has a feature rich interface for dealing with all different aspects of DNS including DNSSEC. The main target of this patch is to replace all uses of acutil DNS library with a use python-dns. In most cases, even though the larger parts of the code are changed, the actual functionality is changed only in the following cases: - redundant DNS checks were removed from verify_fqdn function in installutils to make the whole DNS check simpler and less error-prone. Logging was improves for the remaining checks - improved logging for ipa-client-install DNS discovery https://fedorahosted.org/freeipa/ticket/2730 https://fedorahosted.org/freeipa/ticket/1837
2012-05-11 07:38:09 -05:00
if not is_host_resolvable(query):
Fix translatable strings in ipalib plugins. Needed for xgettext/pygettext processing.
2011-02-23 15:47:49 -06:00
raise errors.NotFound(
reason=_('Host \'%(host)s\' not found') % {'host': query}
)
Port installer and host plugin to the new DNS plugin * move ipa dns-resolve to the new plugin * port the installer and the host plugin to the new interface * remove the old plugin
2011-01-12 14:02:05 -06:00
return dict(result=True, value=query)
api.register(dns_resolve)
Provide API to check if IPA DNS is enabled on some server Fixes: https://fedorahosted.org/freeipa/ticket/600
2011-01-17 11:28:24 -06:00
class dns_is_enabled(Command):
"""
Checks if any of the servers has the DNS service enabled.
"""
Rename INTERNAL to NO_CLI for commands we hide from the cli. Also make i18n_messages and json_metadata NO_CLI. ticket 821
2011-01-20 14:07:43 -06:00
NO_CLI = True
Provide API to check if IPA DNS is enabled on some server Fixes: https://fedorahosted.org/freeipa/ticket/600
2011-01-17 11:28:24 -06:00
has_output = output.standard_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
base_dn = DN(('cn', 'masters'), ('cn', 'ipa'), ('cn', 'etc'), api.env.basedn)
Provide API to check if IPA DNS is enabled on some server Fixes: https://fedorahosted.org/freeipa/ticket/600
2011-01-17 11:28:24 -06:00
filter = '(&(objectClass=ipaConfigObject)(cn=DNS))'
def execute(self, *args, **options):
ldap = self.api.Backend.ldap2
dns_enabled = False
try:
Fix dns_is_enabled command
2011-01-20 14:42:50 -06:00
ent = ldap.find_entries(filter=self.filter, base_dn=self.base_dn)
if len(ent):
Provide API to check if IPA DNS is enabled on some server Fixes: https://fedorahosted.org/freeipa/ticket/600
2011-01-17 11:28:24 -06:00
dns_enabled = True
except Exception, e:
pass
return dict(result=dns_enabled, value=u'')
api.register(dns_is_enabled)
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
class dnsconfig(LDAPObject):
"""
DNS global configuration object
"""
object_name = _('DNS configuration options')
Add missing global options in dnsconfig Add a support for new global options in bind-dyndb-ldap, that is: * idnsforwardpolicy: Default policy for conditional forwarding * idnsallowsyncptr: Allow globaly PTR synchronization for dynamic updates * idnszonerefresh: Default interval between regular polls of the name server for new DNS zones https://fedorahosted.org/freeipa/ticket/2439
2012-03-07 08:54:38 -06:00
default_attributes = [
'idnsforwardpolicy', 'idnsforwarders', 'idnsallowsyncptr',
'idnszonerefresh'
]
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
label = _('DNS Global Configuration')
label_singular = _('DNS Global Configuration')
takes_params = (
Str('idnsforwarders*',
Allow port numbers for idnsForwarders Let user enter custom ports for zone conditional forwarders or global forwarders in dnsconfig. Ports can be specified in a standard BIND format: IP_ADDRESS [port PORT] https://fedorahosted.org/freeipa/ticket/2462
2012-03-20 09:35:54 -05:00
_validate_bind_forwarder,
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
cli_name='forwarder',
label=_('Global forwarders'),
Allow port numbers for idnsForwarders Let user enter custom ports for zone conditional forwarders or global forwarders in dnsconfig. Ports can be specified in a standard BIND format: IP_ADDRESS [port PORT] https://fedorahosted.org/freeipa/ticket/2462
2012-03-20 09:35:54 -05:00
doc=_('A list of global forwarders. A custom port can be specified ' \
'for each forwarder using a standard format "IP_ADDRESS port PORT"'),
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
csv=True,
),
Add missing global options in dnsconfig Add a support for new global options in bind-dyndb-ldap, that is: * idnsforwardpolicy: Default policy for conditional forwarding * idnsallowsyncptr: Allow globaly PTR synchronization for dynamic updates * idnszonerefresh: Default interval between regular polls of the name server for new DNS zones https://fedorahosted.org/freeipa/ticket/2439
2012-03-07 08:54:38 -06:00
StrEnum('idnsforwardpolicy?',
cli_name='forward_policy',
label=_('Forward policy'),
values=(u'only', u'first',),
),
Bool('idnsallowsyncptr?',
cli_name='allow_sync_ptr',
label=_('Allow PTR sync'),
doc=_('Allow synchronization of forward (A, AAAA) and reverse (PTR) records'),
),
Int('idnszonerefresh?',
cli_name='zone_refresh',
label=_('Zone refresh interval'),
doc=_('An interval between regular polls of the name server for new DNS zones'),
minvalue=0,
),
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
)
def get_dn(self, *keys, **kwargs):
return api.env.container_dns
def get_dnsconfig(self, ldap):
(dn, entry) = ldap.get_entry(self.get_dn(), None,
normalize=self.normalize_dn)
return entry
Improve user awareness about dnsconfig Global DNS configuration is a nice tool to maintain a common DNS settings stored in LDAP which are then used for all enrolled IPA servers. However, the settings stored in LDAP override local settings in named.conf on DNS servers. This patch adds more information about global DNS configuration options in install scripts and DNS module help. https://fedorahosted.org/freeipa/ticket/2525
2012-03-15 07:51:59 -05:00
def postprocess_result(self, result):
if not any(param in result['result'] for param in self.params):
result['summary'] = unicode(_('Global DNS configuration is empty'))
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
api.register(dnsconfig)
class dnsconfig_mod(LDAPUpdate):
__doc__ = _('Modify global DNS configuration.')
Improve user awareness about dnsconfig Global DNS configuration is a nice tool to maintain a common DNS settings stored in LDAP which are then used for all enrolled IPA servers. However, the settings stored in LDAP override local settings in named.conf on DNS servers. This patch adds more information about global DNS configuration options in install scripts and DNS module help. https://fedorahosted.org/freeipa/ticket/2525
2012-03-15 07:51:59 -05:00
def execute(self, *keys, **options):
result = super(dnsconfig_mod, self).execute(*keys, **options)
self.obj.postprocess_result(result)
return result
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
api.register(dnsconfig_mod)
class dnsconfig_show(LDAPRetrieve):
__doc__ = _('Show the current global DNS configuration.')
Improve user awareness about dnsconfig Global DNS configuration is a nice tool to maintain a common DNS settings stored in LDAP which are then used for all enrolled IPA servers. However, the settings stored in LDAP override local settings in named.conf on DNS servers. This patch adds more information about global DNS configuration options in install scripts and DNS module help. https://fedorahosted.org/freeipa/ticket/2525
2012-03-15 07:51:59 -05:00
def execute(self, *keys, **options):
result = super(dnsconfig_show, self).execute(*keys, **options)
self.obj.postprocess_result(result)
return result
Global DNS options Implement API for DNS global options supported in bind-dyndb-ldap. Currently, global DNS option overrides any relevant option in named.conf. Thus they are not filled by default they are left as a possibility for a user. Bool encoding had to be fixed so that Bool LDAP attribute can also be deleted and not just set to True or False. https://fedorahosted.org/freeipa/ticket/2216
2012-02-10 05:54:49 -06:00
api.register(dnsconfig_show)
Reference in New Issue Copy Permalink