Add missing relation types items in schema global variables Some of these global variables are used, in particular, to control schema views (web/views/schema.py). While they are currently not properly documented (see #3360868), their definition are currently incomplete (e.g. one could reasonably expect to find ``cw_schema`` in ``SYSTEM_RTYPES``). Closes #3486114.

[web] return 403 for Unauthorized, not 401 401 is reserved to HTTP authentication. Just because it's also called "unauthorized" doesn't mean it's the same as cubicweb's Unauthorized exception. Closes #3648809.

[test] update unittest_repoapi to 3.19 api (sic) [jcr: while at it, fix typos and add missing __main__ section]

[test] update unittest_utils to 3.19 api

[test] update unittest_rset to 3.19 api [jcr: use self.admin_access instead of creating a new one each time]

[test] update unittest_rqlrewrite to 3.19 api

[test] update unittest_req to 3.19 api [jcr: use self.admin_access instead of creating a new one each time]

[test] update unittest_predicates to 3.19 api [jcr: use self.admin_access instead of creating a new one each time]

[web/test] port unittest_reledit to RepoAccess API

[web/test] port unittest_magicsearch to RepoAccess API request created in setUp makes context manager usage painful, so create a context manager to do the setup for each test class.

[web/test] use real_error_handling context manager instead of open-coding it

[web/test] port unittest_idownloadable to RepoAccess API

[web/test] remove useless calls to self.request() from CORS tests

[web/test] port unittest_formfields to RepoAccess API

[web/test] port unittest_form to RepoAccess API

[web/test] port unittest_facet to RepoAccess API

[web/test] move unittest_controller to RepoAccess test API

[web/test] port unittest_breadcrumbs to 3.19 test api

[web/test] avoid deprecation warnings from find_entities()

[web/test] port unittest_application to new 3.19 API

[test] silence warnings for unittest_schema by upgrading to new 3.19 API

[test] silence warnings for unittest_cwctl by upgrading to new 3.19 API

[test] replace some assert{True,False} with more specific method

[test] silence warnings for unittest_entity by upgrading to new 3.19 API

Fix constraint sync during migration - restore constraints lost during merge in test schema. - use constraint_by_eid in BeforeDeleteCWConstraintHook as done in 3.17.14 for BeforeDeleteConstrainedByHook. Fixes handling of multiple constraints of the same type. - make sync_schema_props_perms() delete the CWConstraint entity instead of the constrained_by relation. In 3.19, the latter doesn't automatically result in the former just because the relation is composite. Simplify the constraint migration to delete all removed constraints and recreate new ones even if they share the same type; that optimization made the code more complicated for (AFAICT) no significant reason.

merge 3.18.4 into default server/test/unittest_migractions fails due to interaction between new constraint test (from 3.18) and composite deletion change (in 3.19).

Added tag cubicweb-version-3.18.4, cubicweb-debian-version-3.18.4-1, cubicweb-centos-version-3.18.4-1 for changeset 0176da9bc752

[test/data] Add missing 'add' permission on attribute Silences a warning introduced in 3.18.

[hooks/notification] Fix indentation Follow-up for changeset 1b8552265f3b "[hooks/notification] use a cnx not a session"

[devtools] deprecate CWTC.user()

minor cleanups

some copyright updates

[devtools] properly close open access on tearDown avoiding further warning about living sessions while the repo is turned off. The doc string says: The RepoAccess need to be closed to destroy the associated Session. TestCase usually take care of this aspect for the user. but that wasn't true until now.

[devtools] make comment (a bit) more readable

[devtools] avoid an internal deprecation warning websession and session are now one and the same.

[server/session] undeprecate session.anonymous_session It's not connection-specific.

[web/sessions] use session.sessionid instead of deprecated session.id

[repo] don't set cnxset when getting the session to close it and avoid warning about that. session._cnx.rollback won't do anything if it's not set and a newer connection is now created to run the session_close hook, so there is not need to set cnxset. Notice we have to take care about not trying to rollback the underlying cnx when the session is not 'session_handled' (i.e. bw compat mode). This case only occurs due to hackish things to keep bw compat in testlib.

[repo] handle connection explicitly when calling session open/close hooks

[hooks/syncschema] work with connections not sessions

[server/sources] {before,after}_entity_insertion wants a cnx not a session

[hooks/syncsources] use a cnx not a session

[hooks/notification] use a cnx not a session

[hooks/metadata] use a cnx not a session

[hooks/bookmark] use a cnx not a session

[hooks/workflow] use a cnx not a session

[hooks/email] use a cnx not a session

[hooks/integrity] use a cnx not a session

[hooks/security] let's use a connection, not a session

[server/sources/native] deal with connections, not sessions

[server] eschema_eid needs a connection, not a session

[repository] operations get a connection instead of a session Left extid2eid unchanged because that function scares the shit out of me.

[server/hook] operations get a connection instead of a session

[repository] Use an internal connection in register_user

avoid to internally raise 3.19 warnings

[sources] ensure we have a cnxset in undoable_transactions

[server] move security/integrity hook management away from InternalSession.__init__ and to Repository.internal_{session,cnx} instead. Lets internal_cnx avoid deprecation warnings.

[migration] make 'session' object be a server-side Connection, not ClientConnection I think this better corresponds to the previous behaviour. However, that Connection does not have a cnxset, which should probably be changed (for both interactive_shell and cmd_process_script).

[sources] Skip problematic sources when starting shell instead of crashing. Closes #3670208 At least this gives an opportunity to fix the problem in using c-c shell, which is currently not possible because it crashes during initialization. Notice that since sources are copy based, it should not be a problem at all to have some disabled, beside for services such as authentication sources as ldapfeed.

[web] fix language negotiation Regression caused by commit 6fd0ac6506cb "[web-request] handle default language earlier". We would handle language negotiation at request init time, but overwrite the selected language with the site default when calling set_cnx.

[merge] bring the 3.17.14 fixes to 3.18

[test/utils] repair concat-resources test That was forgotten in 240a620b9cd3. Related to #3670503.

[pkg] bump cubicweb.spec

[dataimport] Correctly call rschema(rtype) in SqlGenObjectStore, closes #3694139

[pkg] prepare 3.18.4

[web] Disable 'concat-resources' by default (closes #3670503) It makes debugging harder and it can actually *break* working JS or CSS files. This is the first step towards complete removal.

[testlib] Drop override of assertItemsEqual method The override is not compatible with underlying (standard) method because it builds sets and rely on eid. Closes #3641111.

[rewrite] Fix crash when the main variable doesn't appear in the snippet's vargraph Also, the added test discover another bug once the first has been fixed (erroneous argument given to .rewrite recursive call). Closes #3661918

[ext/rest] Catch SystemExit raised by docutils docutils publisher may raise SystemExit which is not caught by Exception. Closes #3648763.

[entity] User-defined relation to skip for copy has precedence Otherwise permission problems can occur on rtypes not yet skipped Closes #3653459

Make EditController edit_entity method always return an eid In cases the entity was to be created or copied, eid was None hence the method returned None despite the dosctring says the method should always return an eid. Now if the eid variable is None, it is assigned to the newly created entity eid. Closes #3593029.

[migration] Improve update of in-memory schema during 3.18 CWAttribute.defaultval change We weren't updating the schema properly so a bunch of structures still had the old rdef with 'String' as object. That caused failures to add new attributes later in the migration as their default value would be a Binary object but the schema would expect a String. Closes #3683640

[migration] always rebuild infered relation This was skipped for some bad reason (see 12ad88615a12 which introduced the change). Fix for #231956 in Yams is necessary to allow this cset: during a migration, we want to always reinfer relations while allowing explicit redefinition of an infered relation later. Else, we may run the migration with missing parts of the schema (the one that should have been infered). Closes #3685463.

Backout "[web/navigation] use add_onload instead of inline javascript href" This reverts changeset 170e1437948d, for at least two reasons: the PageNavigationSelect component was not updated for the new type of navigation links, and external cubes defining View.page_navigation_url exist in the wild and also got broken navigation links. This reopens #3501626 and closes #3677945.

[test] those are rql tests, not cubicweb (and they break with rql 0.31.5). Remove them.

[source/native] allow many eid creation per .create_eid call (closes #3526594) [jcr: move migration code to bootstrapmigration_repository.py to make sure we stop using the old sequence ASAP]

[pkg] bump the dependency on logilab.database This will allow use of an extended API. Related to #3526594.

[source/native] refactor eid generation code This set of related methods clutters the native source API. It will be better served with a dedicated eid genrator object. Related to #3526594. [jcr: allocate the eid generator in __init__]

Move entity cache from web.request.ConnectionCubicWebRequestBase to repoapi.ClientConnection ResultSet._build_entity relies on an entity cache to avoid going round in circles (stack overflow due to infinite recursion) e.g. in a postcreate script. Clear the cache at commit/rollback time to avoid unexpected side effects; and explicitly clear the cache in a few tests where we mix ORM and RQL using the same connection.

[server] Fix AttributeError in extid2eid Session._tx was renamed in cb87e831c183, use _cnx instead.

[book] Update documentation for new repoapi Quite a few things change in 3.19: - repoapi instead of dbapi - ClientConnection / Connection / Session rework - web authentication process - test APIs Closes #3638793

[utils] the json module is always available It's included since python 2.6, and we don't support earlier versions.

[fti] properly close the ProgressBar Ensures cubicweb-ctl db-rebuild-fti ends its output with a newline.

[serverctl] use repoapi for db-check, add-source, rebuild-fti commands One more step towards getting rid of dbapi.

[devtools] make get_repo_and_cnx return a repoapi ClientConnection And adjust callers accordingly (they need to use it as a context manager to open/close it properly).

[repoapi] Avoid deprecation warnings from session.id Apparently it's called session.sessionid now.

[server] some s/session/cnx/ For merely accessing the database we need a connection, not a session.

Move setting session.mtime from dbapi to web session manager clean_session was broken since the switch away from dbapi on the web side, since session.mtime wasn't being set anywhere.

[web] whitespace cleanup in http_headers.py

[web] implement cross origin resource sharing (CORS) (closes #2491768) Partial implementation that is enough to get started but leaves out some of the advanced features like caching and non-simple methods and headers.

[devtools] add a 'method' argument to RepoAccess.web_request so that one can easily forge a request with any HTTP method. Also a bunch of clone cleanups.

[web] Fix handling of some http headers The generator takes as argument a list of raw string values, and is supposed to return the parsed version. Calling str on that list makes no sense.

[web/request] deprecate user_callback Storing closures in session data considered harmful. Closes #3567793

[doc] undocument user_callback Related to #3567793

Added tag cubicweb-centos-version-3.17.14-1, cubicweb-version-3.17.14, cubicweb-debian-version-3.17.14-1 for changeset fa00fc251d57

[migractions] Better handle removal of RQLConstraint in sync_schema In rdef synchronisation, filter unmodified constraints before processing the sets of old and new constraints. Add a new `constraint_by_eid` method on RelationDefinitionSchema, which eliminates the ambiguity of `constraint_by_type` usage when several constraints of the same type are involved in the same migration. Closes #3633644.

[test] when the config change, we should cleanup sys.path and sys.modules. Closes #3634164 Without this, modules imported within instance home will be messed up.

[wsgi] make sure request.content is available for consumption The wsgi.input stream is not seekable, so make a copy either to memory or disk so it's still available to the user even after we've parsed it (for POST form processing). Closes #3554996

[facet] create a RangeRQLPathFacet (closes #2852512) This facet is a mix between a RQLPathFacet and a RangeFacet, allowing to use the FacetRangeWidget on the RQLPathFacet target attribute

[request] Make sure set_message() actually displays its given message (closes #3003425) The default automatic form sets a standard submit message as session data. However in some cases, such as the cancellation of a form edition, another message is intended to be printed on-screen. This patch makes sure that set_message() really scrubs all previous messages (and ids) from the request object and from the session data.

[pkg] prepare 3.17.14

[hooks/syncschema] do not crash on double removal of a constraint Some complicated migrations may entail a double constraint removal. It is not yet known whether this is a genuine cw bug or the result of a complicated migration story, but we should nevertheless not crash. Rather, emit a CRITICAL warning and proceed. Closes #3595309.

[doc/book] Fix typo in import

[server/session] fix a couple typos in doc strings

[server/migractions] finish migration to repoapi objects Changeset 241b1232ed7f (Use repoapi instead of dbapi for cwctl shell, upgrade and db-init) only did half of the job. It left the migration handler with both a session (cubicweb.server.session.Session) and a cnx (cubicweb.repoapi.ClientConnection) attribute with different ideas of what Connection they were talking to. With this change, we: - make the caller responsible of disabling security on the Connection if it wants to - turn mih.session into an actual attribute, set on __init__ - same for cnx (the client connection) - drop the "lazy connection" logic, and establish the connection up-front - always go through the connection instead of the session when we need to talk to the db

[server] use the ClientConnection directly now that it has more methods available Gets rid of ugly self.cnx._cnx.

[repoapi] add security/hook control and system_sql Just redirect these methods to the repo-side Connection.

[predicates] fix thinko in doc strings Replace "e.g." with more appropriate "i.e.". These are not examples.

[web/basecomponents] remove EtypeRestrictionComponent Closes #3119951

[ldapparser, book] document additional error causes

[js] remove leftover dead code in loadxhtml form.callback support was removed in changeset 68cde7431c2c, but some signs of it remained.

[web] Avoid using an empty dict as default parameter of 'CubicWebRequestBase'. Also update the related docstring of the __init__ method.

[schema] fix composite deletion handling Do not delete component entities if the composite is not deleted in the same transaction. Deletion semantics are thus: if the composite is deleted, the components are deleted. Closes #3463112.

[schema] add composite handling helpers on EntitySchema (related to #3463112) These will help write predicates, hooks, etc.

[server] Handle unique constraint violations under recent sqlite The error message changed from "columns foo, bar, baz are not unique" to "UNIQUE constraint failed: table.foo, table.bar, table.baz". Closes #3564510

[i18n] update de: 1065 translated messages, 1 fuzzy translation, 259 untranslated messages. en: 656 translated messages, 669 untranslated messages. es: 1325 translated messages. fr: 1314 translated messages, 11 untranslated messages.

[i18n] es.po updated

merge 3.18.x in 3.19 branch

3.18 is stable

[schema] explicitly declare 'add' permissions for a couple attributes

Added tag cubicweb-version-3.17.13, cubicweb-debian-version-3.17.13-1, cubicweb-centos-version-3.17.13-1 for changeset 09b4ebb9b0f1

Added tag cubicweb-version-3.18.3, cubicweb-debian-version-3.18.3-1, cubicweb-centos-version-3.18.3-1 for changeset afd21fea201a

Update translations

[pkg] prepare 3.18.3

merge 3.17.13

[ldapfeed] fix encode error during initial user import Closes #3539196.

[web/data] Ignore disabled widgets in cw.utils.formContents() (closes #3544492) 17.12.1 Disabled controls : http://www.w3.org/TR/REC-html40/interact/forms.html#h-17.12.1 When set, the disabled attribute has the following effects on an element: * Disabled controls do not receive focus. * Disabled controls are skipped in tabbing navigation. * Disabled controls cannot be successful. The third one is the important one. 17.13.2 Successful controls : http://www.w3.org/TR/REC-html40/interact/forms.html#successful-controls A successful control is "valid" for submission. Every successful control has its control name paired with its current value as part of the submitted form data set. Bottom line, disable widgets should not be part of the names and values lists returned by formContents().

[etwist] Fix an empty request content after a Twisted processing (closes #3546795). The content of a POST request could be empty when the Twisted server is used.

[web/data] remove images that aren't used anywhere

[web/data] Add missing images from jquery-ui 1.10.3 Closes #3175933

[web/navigation] use add_onload instead of inline javascript href This way our javascript code isn't thrown out by the html cleaner e.g. when using the rql rest directive and a table view. To make things simpler, we now always use ajax URLs for navigation, even when we would previously have used regular links. Closes #3501626

[web/data] fix treeview regression (closes #3526466) Changeset 68cde7431c2c "[js] remove 3.9 bw compat (where apparently unused)" removed the use of form.callback from loadxhtml, which treeview relied on. Update to add a callback on the loadxhtml return value instead.

backout 8f3e963501e2, which is not ready yet

[pkg] also bump it there

[pkg] prepare 3.17.13

[navigation] use add_onload instead of inline javascript href This way our javascript code isn't thrown out by the html cleaner e.g. when using the rql rest directive and a table view. Closes #3501626

[uilib] allow canvas tags in the html cleaner Used by the iprogress cube. Closes #3524254.

[ajax] use a custom tag to handle dynamically loaded js Using <pre class="script"> makes it trivial for a malicious user to inject arbitrary javascript into a html or rest text element (because it looks innocent to the html sanitizer). Using a custom tag we can be sure that it actually comes from our code and not from untrusted user data. IE ignores custom tags, though, so we put it in its own namespace. https://extranet.logilab.fr/1530578

[dataimport] fix comment

[hooks/security, devtools/fill] silence yams 0.38.0 warnings

Drop 3.13 incomplete backward compat in edit controller. It is very old and broken (there is another non-backward-copmpatible usage of `_cw_entity_fields`), better to remove it instead of fixing. Closes #3515223.

[devtools] force locale to C for postgresql test clusters Avoids initdb failure with e.g. iso8859-X locale and utf-8 encoding.

[rql2sql] remove special behaviour of symmetric relation vs DISTINCT 0542a85fe667 replacing SQL OR by hooks for symmetric relations allows that. This involve a single test value change for a border case: when querying a symmetric relation without specifying the subject nor the object, you may get some duplicated result. IMO this is fine to let the user explicitly use DISTINCT or not and to remove the dedicated handling we had which didn't let any choice. Related to #3259713

Typo in comments and error messages

Fix typo in a setup.py comment

[dataimport, migration] more fixes in the spirit of a6c32edabc8d: [entity, metadata] huuum, use resolvable url as cwuri... And fix existing ones. Closes #3390388

[doc] Fix typo in devrepo/fti

[doc/3.19] Clarify repoapi.get_repository usage Followup for 9a62c52d167e.

[server] use a connection instead of a session for user authentication

Use repoapi instead of dbapi for cwctl shell, upgrade and db-init Hopefully nobody uses dbapi-specific API in their migration scripts. I guess we'll find out.

[server/repo] use internal_cnx in init_sources_from_database

[server/session] avoid spurious warnings from Session.close Add internal non-noisy _rollback method.

[server] rename session to cnx in querier and plan They don't actually need a session, so let's make that clear.

[repo] Use a connection instead of a session for repo.stats()

[repo] Use a connection instead of a session for repo.gc_stats()

[repo] Use a connection instead of a session for repo.get_versions()

[webconfig] get_repository moved to repoapi

[schemaserial] Replace 'cursor' with 'cnx' We don't have cursors in rql, since execute is synchronous.

[book] basic documentation for LDAPfeed

[testlib] Use session.sessionid instead of deprecated session.id

Add data/pgdb to hgignore patterns PostgresTestDataBaseHandler uses it as datadir.

[test] make unittest_postgres pass Use PostgresApptestConfiguration which starts a postgresql cluster locally for the duration of the test.

[devtools] make PostgresTestDataBaseHandler start its own postgresql Don't depend on postgresql already running on the system, or on a static config. To use this, set the configcls attribute of your CubicWebTC-derived test class to PostgresApptestConfiguration. Caveats: - this won't work with several tests running concurrently from the same directory, since each would start its own cluster with a shared data directory and stomp on each other's toes - you need initdb/pg_ctl in $PATH, e.g /usr/lib/postgresql/$version/bin on Debian/Ubuntu systems. Closes #3489631

[web/staticcontrollers] create a unique temporary file for concat handling Predictable names means different processes can race against each other. Let's avoid that and use mkstemp to get a real unique temporary file, and atomically rename it when we're done. Closes #3524182

[server/repo] Use session.sessionid instead of session.id Silences deprecation warning.

[server/session] use _free_cnxset instead of free_cnxset to avoid warnings No need to make DeprecationWarning noise for something the user doesn't have any control over.

[server] Use internal_cnx instead of internal_session in init_repository

[server/session] session.id → session.sessionid The former causes deprecation warnings.

English typography Get rid of extra space before `!' and `?'

[repo] restore python 2.6 compatibility Broken in 12dfce15c8ea.

[server] use internal_cnx instead of internal_session in deserialize_schema

[server/schemaserial] rename cursor argument to cnx cursor is a sql concept it has no meaning for rql.

[repo] Use a connection instead of a session in Repository.properties()

[server/sources] replace references to dbapi with repoapi

[doc/3.19] Make repoapi description match code

[serverctl] remove broken schema-sync command As far as I can tell the migration handler hasn't had a cmd_synchronize_schema method since cubicweb 3.2.0.

[server/session] Fix docstring typo for InternalSession

[book] typo

[server] fix TypeErrors With multisource gone, cnxset.reconnect() doesn't take a source argument. Broken since d62e13eba033.

[notification] use an InternalManager object for notifications if we have an (email, lang) tuple The non-uniform handling of CWUser vs (email, lang) tuple has caused too many bugs; this change mostly unifies the two code paths. Closes #3381521

[req] if a user has a None property, fall back to the default This can happen for InternalManager; the normal CWUser implementation of property_value already has a similar fallback. Related to #3381521

[server] Allow setting an InternalManager's preferred language Related to #3381521

merge 3.18.2 into 3.19 branch

[migration/3.18] protect against attributes in db missing from fsschema Happens if a previous migration forgot a drop_attribute. Closes #3471609

[doc] Update copyright date to 2014 (closes #3463136)

[repoapi] Fix typos in doc strings and comments

[cwxml parser] minor pylint fixes

[devtools] make CubicWebTC.login as admin undo previous logins Fixes things when login is not used as a context manager.

[devtools] fix self.session._cnx vs self.cnx._cnx confusion after self.login Make sure the client connection and the session refer to the same server connection.

[devtools] _orig_cnx does not exist any more (missing change)

[c-c schema] option to show single etype (closes #3404831)

[multi-sources-removal] Drop entities.source column Since we remove federated multi-sources support, the same value ('system') is always stored in this column and so could be removed. Also, cleanup repository caches and a few api where the very same useless information could be removed. Closes #2919300, at last [jcr: restore sanity check of etype against schema in checkintegrity.has_eid]

[repo] pylint cleanup, mainly of imports, with a bit of style

[multi-sources-removal] Drop no more used federated ('true') multi-sources related code Related to #2919300 (almost done!)

[multi-sources-removal] Turn ConnectionsSet into simpler ConnectionWrapper also, SqliteConnectionWrapper becomes a subclass. This is allowed since now ConnectionsSet responsability has been reduced to handling the system source only. This changeset also: * drops useless cnxset_set / check_connection api * deprecates former .source / .connection / container api but it renames neither the session's cnxset attribute, nor the related repository method. Related to #2919300

[multi-sources-removal] Drop cnxset_freed from source interface Though a hack is still needed when using sqlite Related to #2919300

[cnxset] add backwards compat for cnxset.source_cnxs It's used at least in a vcsfile migration script.

[multi-sources-removal] Simplify ConnectionsSet internal structures and public methods since it now handles a connection to the system source only Related to #2919300 [jcr: adjust 3.17.11 migration, fix a number of bugs in new ConnectionsSet implementation, fix source.{open,close}_source_connections]

[ldap] simplify connection handling since we deleted ldapuser source, we don't need anymore the get_connection and ConnectionWrapper stuff (that was used to put the ldap connection into the cnxset). Also, we should simply let connection errors propagate so this is properly reported to import logs.

[ldap] merge cw.server.ldaputils back into ldapfeed source we don't need separated modules since ldapuser source has been removed

[config] cleanup/refactor server sources file values handling * kill former `sources` method, misnamed since we've only the system source configuration in the sources file (and some default admin account information) * introduce new system_source_config and default_admin_config properties to access to the two information in this file Then use one or the other when appropriate

[repoapi] drop get_option_value's foreid argument repoapi was introduced in 3.19, so no need to introduce legacy deprecated stuff at the same time.

[multi-sources-removal] Drop foreid argument of repo.get_option_value It doesn't need it anymore. Related to #2919300

[c-c restore] drop useless restore-all option, and related systemonly argument

[multi-sources-removal] Kill repo.sources there is only the system source in there now! Related to #2919300

[multi-sources-removal] Drop no more necessary scleanup (source cleanup) argument Related to #2919300

[multi-sources-removal] drop source location search on glob_*[entity|relation] methods source is now always the system source Related to #2919300 [jcr: fix wrong rebase in symmetric relation removal]

[server test] simplify data-schemaserial test instance's schema

[multi-sources-removal] Drop (dont_)cross relation related code That was only needed for now deleted 'true' multisources. Related to #2919300 [jcr: also remove it from server/test/data-schemaserial/schema.py]

[multi-sources-removal] Simplify repo.delete_info_multi arguments, uri is no more used Related to #2919300

[multi-sources-removal] Drop deleted_entities system table and entities.mtime column since they were only used by the entities_modified_since api of the repository which has been dropped. Along with them, the multi-sources-etypes configuration variable and some sql queries at modification/deletion time of entities. Bon vent ! Related to #2919300

[multi-sources-removal] drop repository entities_modified_since public api actually only used by pyrorql/zmqrql sources which have been dropped. This will allow to drop a bunch of system meta-data (eg deleted_entities) Related to #2919300

[multi-sources-removal] Drop the "true" multi-sources planner Only keep copy-based sources. This will soon allow for greater code simplification. Related to #2919300

[multi-sources-removal] Drop cw.server.sources.extlite no more used for a while, beside an utility class Related to #2919300

[multi-sources-removal] Move cw.server.sources.ConnectionWrapper where it belongs It's only used by the ldap utilities. Related to #2919300

[multi-sources-removal] Drop pyrorql and zmqrql sources After a few years experementing "true" multi-sources, we're now moving to "copy-based" source à la datafeed. As pyro and zmq sources have no more known customers and the related code is in the way of future refactoring of cubicweb's core, we decided to drop support for those sources without backward compatibility. If you're still using a zmqrql or pyrorql source and you want to upgrade, ask support to move it to datafeed using a pre-3.19 version first. Related to #2919300 (first step)

[web/formfields] Make CompoundField inherit needs_multipart from its subfields If there's a FileField within a CompoundField, the latter will not have the enctype multipart set, even though the FileField needs it. CompoundField should inspect its children fields to determine whether it needs multipart. Closes #3406930.

[wsgi/handler] fix api change from 3.18

[req] deprecate get_cache (closes #3181499)

Added tag cubicweb-version-3.18.2, cubicweb-debian-version-3.18.2-1, cubicweb-centos-version-3.18.2-1 for changeset 6880674c1a26

merge 3.17.12 into 3.18 branch

[editcontroller] fix cardinality computation (closes #3120495)

Added tag cubicweb-version-3.17.12, cubicweb-debian-version-3.17.12-1, cubicweb-centos-version-3.17.12-1 for changeset 838d58a30f7e

[debian] move yams max version to Depends And fix a syntax error and wrong package name in the process.