Fri, 20 Nov 2015 18:23:17 +0100 [web] use a redirect for /data/foo requests that don't use the instance md5 hash
Julien Cristau <julien.cristau@logilab.fr> [Fri, 20 Nov 2015 18:23:17 +0100] rev 10851
[web] use a redirect for /data/foo requests that don't use the instance md5 hash The actual file contents can be cached for a long time. If the request doesn't use the hash, that's wrong. Using a redirect lets keep the caching for the actual file contents while not breaking requests missing the hash.
Fri, 20 Nov 2015 12:37:15 +0100 [web] fix typo in doc string
Julien Cristau <julien.cristau@logilab.fr> [Fri, 20 Nov 2015 12:37:15 +0100] rev 10850
[web] fix typo in doc string
Thu, 19 Nov 2015 18:55:10 +0100 [web] stop adding an Expires header with all responses
Rémi Cardona <remi.cardona@logilab.fr> [Thu, 19 Nov 2015 18:55:10 +0100] rev 10849
[web] stop adding an Expires header with all responses As per RFC 1945 (HTTP 1.0): If the date given is equal to or earlier than the value of the Date header, the recipient must not cache the enclosed entity. As per RFC 7234 (HTTP 1.1 Caching): If a response includes a Cache-Control field with the max-age directive (Section 5.2.2.8), a recipient MUST ignore the Expires field. Bottom line, the Expires header is now handled wherever the Cache-Control is handled: inside cache managers.
Fri, 20 Nov 2015 17:13:03 +0100 [web] stop looking at a request's Cache-Control header
Julien Cristau <julien.cristau@logilab.fr> [Fri, 20 Nov 2015 17:13:03 +0100] rev 10848
[web] stop looking at a request's Cache-Control header As per RFC 7234 (HTTP 1.1 Caching): The "Cache-Control" header field is used to specify directives for caches along the request/response chain. CubicWeb is not a cache.
Thu, 19 Nov 2015 17:50:48 +0100 [fti] add cw_fti_index_rql_limit method
Julien Cristau <julien.cristau@logilab.fr> [Thu, 19 Nov 2015 17:50:48 +0100] rev 10847
[fti] add cw_fti_index_rql_limit method Improves on and deprecate cw_fti_index_rql_queries: execute the rql directly, so that we don't need to compute the set of eids up-front, but can just keep track of the last seen entity. Use the new method in reindex_entities. Keep calling cw_fti_index_rql_queries if it has been redefined (e.g in cubes). Related to #3621392.
Thu, 19 Nov 2015 16:48:55 +0100 [fti] backout 166c6f7b1be4
Rémi Cardona <remi.cardona@logilab.fr> [Thu, 19 Nov 2015 16:48:55 +0100] rev 10846
[fti] backout 166c6f7b1be4 * breaks tests * see next commit for a better solution
Thu, 19 Nov 2015 17:16:19 +0100 [py3k] Fix unicode added in commit 2350424585f8
Rémi Cardona <remi.cardona@logilab.fr> [Thu, 19 Nov 2015 17:16:19 +0100] rev 10845
[py3k] Fix unicode added in commit 2350424585f8
Thu, 19 Nov 2015 15:56:22 +0100 [server/ldap] deal with unicode or str coming out of ldap
Julien Cristau <julien.cristau@logilab.fr> [Thu, 19 Nov 2015 15:56:22 +0100] rev 10844
[server/ldap] deal with unicode or str coming out of ldap It seems we can get either, I've no idea in which exact circumstances, so this should let us work with both.
Thu, 19 Nov 2015 14:36:54 +0100 [test] pull yapps from Logilab's github
Julien Cristau <julien.cristau@logilab.fr> [Thu, 19 Nov 2015 14:36:54 +0100] rev 10843
[test] pull yapps from Logilab's github Fixes a couple of failures, pull request is almost a year old (https://github.com/smurfix/yapps/pull/2).
Thu, 12 Nov 2015 15:20:45 +0100 [pkg] bump yams requirement to 0.41.1
Julien Cristau <julien.cristau@logilab.fr> [Thu, 12 Nov 2015 15:20:45 +0100] rev 10842
[pkg] bump yams requirement to 0.41.1 Needed by changeset c9321aedfbf9.
Thu, 12 Nov 2015 10:53:50 +0100 [entities] import range from six.moves
Julien Cristau <julien.cristau@logilab.fr> [Thu, 12 Nov 2015 10:53:50 +0100] rev 10841
[entities] import range from six.moves
Thu, 12 Nov 2015 10:52:28 +0100 merge head published by accident
Julien Cristau <julien.cristau@logilab.fr> [Thu, 12 Nov 2015 10:52:28 +0100] rev 10840
merge head published by accident
Thu, 06 Mar 2014 15:55:33 +0100 [fti] Improve big table reindexation
Christophe de Vienne <christophe@unlish.com> [Thu, 06 Mar 2014 15:55:33 +0100] rev 10839
[fti] Improve big table reindexation * Slice the reindexing in batches of 1000 entities. * Make the output more verbose. Closes #3621392
Mon, 09 Nov 2015 11:11:31 +0100 [tox] Reference requirements.txt for hooks environment
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 09 Nov 2015 11:11:31 +0100] rev 10838
[tox] Reference requirements.txt for hooks environment The file existed but was not referenced.
Wed, 25 Mar 2015 14:20:56 +0100 [devtools/testlib] Remove assertions about direct usage of CubicWebTC/AutomaticWebTest
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 25 Mar 2015 14:20:56 +0100] rev 10837
[devtools/testlib] Remove assertions about direct usage of CubicWebTC/AutomaticWebTest The point is to be able to use a test runner other than pytest (e.g. nose2, nosetests or py.test) to run CubicWeb test suite. The issue is that most other test loaders inspects members of all base classes of test case classes to determine test methods and it hits this `config` property (actually executing it), which breaks the assertion. So just return None instead of failing. Closes #7817147.
Fri, 06 Nov 2015 17:02:09 +0100 Add ldap3 to server test requirements
Julien Cristau <julien.cristau@logilab.fr> [Fri, 06 Nov 2015 17:02:09 +0100] rev 10836
Add ldap3 to server test requirements
Fri, 06 Nov 2015 16:44:23 +0100 [pkg] bump version number so tox won't use cw from pypi
Julien Cristau <julien.cristau@logilab.fr> [Fri, 06 Nov 2015 16:44:23 +0100] rev 10835
[pkg] bump version number so tox won't use cw from pypi
Tue, 28 Jul 2015 10:54:24 +0200 [schema] add CubicWebRelationDefinitionSchema
Julien Cristau <julien.cristau@logilab.fr> [Tue, 28 Jul 2015 10:54:24 +0200] rev 10834
[schema] add CubicWebRelationDefinitionSchema Stop monkeypatching yams, inherit from RelationDefinitionSchema instead.
Wed, 22 Apr 2015 11:39:13 +0200 [schema] Let RQLConstraint inherit from BaseRQLConstraint instead of RQLVocabularyConstraint
Denis Laxalde <denis.laxalde@logilab.fr>, Rémi Cardona <remi.cardona@logilab.fr> [Wed, 22 Apr 2015 11:39:13 +0200] rev 10833
[schema] Let RQLConstraint inherit from BaseRQLConstraint instead of RQLVocabularyConstraint There is apparently no reason for RQLConstraint to inherit from RQLVocabularyConstraint as it gets its `repo_check` method from `RepoEnforcedRQLConstraintMixIn` and RQLVocabularyConstraint (which inherits from BaseRQLConstraint) actually only overrides that method. So simplify inheritance by removing a useless level of indirection. Adjust cw_unrelated_rql accordingly, where it was that RQLConstraint inherited from RQLVocabularyConstraint.
Tue, 08 Sep 2015 06:48:48 +0200 [server] unormalize source logger's name to avoid encoding errors
Sylvain Thénault <sylvain.thenault@logilab.fr> [Tue, 08 Sep 2015 06:48:48 +0200] rev 10832
[server] unormalize source logger's name to avoid encoding errors
Tue, 22 Sep 2015 14:20:53 +0200 [entities] parameters should be given to fire_transition, else they are simply ignored
Sylvain Thénault <sylvain.thenault@logilab.fr> [Tue, 22 Sep 2015 14:20:53 +0200] rev 10831
[entities] parameters should be given to fire_transition, else they are simply ignored
Mon, 12 Oct 2015 09:53:51 +0200 [entity] set _cw_adapters_cache in __init__ rather than on first use
Julien Cristau <julien.cristau@logilab.fr> [Mon, 12 Oct 2015 09:53:51 +0200] rev 10830
[entity] set _cw_adapters_cache in __init__ rather than on first use
Mon, 19 Oct 2015 23:23:46 +0000 fix typos and indentation in docstrings/docs
Jérôme Roy <jerome.roy@logilab.fr> [Mon, 19 Oct 2015 23:23:46 +0000] rev 10829
fix typos and indentation in docstrings/docs
Wed, 16 Sep 2015 11:31:36 +0200 session.id → session.sessionid
Julien Cristau <julien.cristau@logilab.fr> [Wed, 16 Sep 2015 11:31:36 +0200] rev 10828
session.id → session.sessionid
(0) -10000 -3000 -1000 -300 -100 -50 -24 +24 +50 +100 +300 +1000 tip