[test] fix failure introduced by yams'changeset 644:f9c79356b758

[testlib] add convenience assertion method to check __message

[test] fix failure introduced by 1719eab69551: the regexp should be fixed, not the error message

[web test] fix test failures introduced by 7431:cab99ccdb774

backport stable changes

[ui message] fix regression introduced by 7429:ef5165fa99e0, leading to 'None' message displayed in the ui

backport stable changes

[json controller] fix .exception call msg format string (missing one slot)

[test] fix tests broken by 7427:5338d895b891

merge default heads

[ui messages, xss] Start migration towards use of _msgid instead of __message (prone to XSS injection) closes #1698245

backport stable

[ui messages] make application message component works when request has no cnx set and support for explicit message given through render argument

backport stable

[web session] fix session handling so we get a chance to have for instance the 'forgotpwd' feature working on a site where anonymous are not allowed fix several pbs: * we need a session id and a session cookie anyway, else subsequent http queries are unrelated * this imply some changes in the session attribution workflow for session without a cnx * some views/selectors must be fixed for cases where session has no cnx On the way, avoid unnecessary Redirect on successful login. closes #750543

[ajax js/css] modconcat fix: load code exactly once (auc, fcayre) * introduces cw.ajax namespace (currently hosts only recent modconcat functionality) * handle load-one/load-many-at-once edge case * properly handle already-loaded js resource * avoid .appendTo to trigger script evaluation since from jquery 1.5 this results in uncached ajax calls

[test] update test broken by previous commit

[entity fetch_attrs] remove ambiguous relations on different etypes from fetched attrs or it may produce wrong related results; closes #1700896

typo : s/undoed/undone/

[cwvreg] inline function dump_class() Grepping the entire code base (cubicweb+all the known cubes) shows that inlining this function is harmless: no one is using it.

[js/ajax] undeprecate (async)remoteExec. closes #1381205

[rql2sql] closes #1697862: ORDERBY SUM(C) cancel GROUPBY by wrongly adding C to grouped variables

backport stable

Added tag cubicweb-debian-version-3.12.5-1 for changeset 6dfe78a0797c

Added tag cubicweb-version-3.12.5 for changeset ae33197db1f0

3.12.5

i18nupdate

[views/table (facetform)] introduced in facetbox, was missing in tableview form filter

[web.facet] self.w -> w

[web/facet|views/table] bump a bit some heights, split table view methods for easier subclassing

backport stable

cleanup and update some copyright

[migractions] fix change_attribute_type

[migractions] dropped_constraints should accept only droprequired without cstrtype

[form rendering] some refactoring to ease FormRenderer overriding

[deprecation] drop remaining warnings

backport stable

[session] new methods on session to ease access to security/hooks control context managers

[workflow] new 'fire_transition_if_possible' method on IWorkflowable to fire a transition if it's possible

[workflow] new 'incontext-state' view for IWorkflowable entities

[facets/table] vertically group facets by two when possible in order to have a more compact horizontal layout in table filter form

[workflow] add method on IWorkflowable to set initial state of an entity (closes #1686357)

[web test] fix test broken by making cw_source editable by managers

[multi-sources] support for moving an entity from an external source (closes #343818) Original need is to move a user from a ldap source to the system source so we can delete it from ldap without loosing information into the cubicweb instance. We can't wait for the user to be deleted from the ldap since it will be too late then to get back user attributes, so it has to be a manual operation to operate before actual deletion. This makes sense for other sources as well. So the idea is to make the "Any cw_source CWSource" relation editable by managers, and to watch changes of it. We then check the move is possible (ie from an external source to the system source) and do necessary stuff (essentially changing source information and copying data into the system source). Remaining pb is that we don't want the moved entity to be reimported later. To distinguish this state, the trick is to change the associated record in the 'entities' system table with eid=-eid while leaving other fields unchanged, and to add a new record with eid=eid, source='system'. External source will then have consider case where `extid2eid` return a negative eid as 'this entity was known but has been moved, ignore it'. Notice no ui is provided yet, it has currently to be done in a c-c shell.

[repository api] definitly kill usage of word 'pool' to refer to connections set used by a session Also document session's data storage and some other internals. Hopefuly things will get clearer. Closes #1684860: vocabulary confusion in repository code: notion of 'pool'

backport stable

[rql2sql] fix in 7394:d11d88bd08e8 was actually not needed

[rset] close #1683703: rset.get_entity crash on rql query with subquery and aggregat

[rql2sql] fix generated sql for eid comparison. closes #1638695

fix some deprecation warnings

[querier test] test groupby function call / column number

backport stable

[plot views] don't paginate

backport oldstable

[url rewriting] fix nasty bug in build_rset inner function: the same dictionary is used by all queries, hence triggering potentially important bugs on concurrent requests (this dictionary goes back and forth the repository)

[hooks selection optimization] prune hooks when multiple entities are concerned by a hm.call_hooks() (closes: #1672022) the idea is to make a first pass over all the hooks in the registry and to mark put some of them in a disabled list. The disabled hooks are the one which: * are disabled at the session level * have a match_rtype or an is_instance selector which does not match the rtype / etype of the relations / entities for which we are calling the hooks. This works because the repository calls the hooks grouped by rtype or by etype when using the entities or eids_to_from keyword arguments Only hooks with a simple selector or an AndSelector of simple selectors (is_instance and match_rtype) are considered for disabling.

[repo test] datafeed parser's process method should now take raise_on_error argument

[datafeed] propagate raise_on_error to parser's process method

[http] Make a better __repr__ for Status responce exception. This help debug.

[schema] fix imprecise error message