Thu, 28 Nov 2019 09:53:33 +0100 Added tag 1.0.0 for changeset 32a9d1361d9e default tip
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Nov 2019 09:53:33 +0100] rev 3991
Added tag 1.0.0 for changeset 32a9d1361d9e
Thu, 28 Nov 2019 09:53:16 +0100 [pkg] Version 1.0.0 1.0.0
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Nov 2019 09:53:16 +0100] rev 3990
[pkg] Version 1.0.0
Thu, 28 Nov 2019 09:53:04 +0100 [doc] Update copyright date range
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Nov 2019 09:53:04 +0100] rev 3989
[doc] Update copyright date range
Thu, 28 Nov 2019 09:51:16 +0100 [doc] Mention Docker install, reorganize sections
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Nov 2019 09:51:16 +0100] rev 3988
[doc] Mention Docker install, reorganize sections
Thu, 28 Nov 2019 09:39:48 +0100 [doc] Remove more CW_MODE occurrences
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Nov 2019 09:39:48 +0100] rev 3987
[doc] Remove more CW_MODE occurrences
Wed, 06 Nov 2019 10:07:31 +0100 [doc] Remove CW_MODE from install instructions
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Nov 2019 10:07:31 +0100] rev 3986
[doc] Remove CW_MODE from install instructions This is not needed when working within a virtualenv.
Wed, 06 Nov 2019 10:03:30 +0100 [doc] Fix titles hierarchy
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Nov 2019 10:03:30 +0100] rev 3985
[doc] Fix titles hierarchy
Wed, 06 Nov 2019 09:54:07 +0100 Added tag 0.25.8 for changeset cf46a796cff6
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Nov 2019 09:54:07 +0100] rev 3984
Added tag 0.25.8 for changeset cf46a796cff6
Wed, 06 Nov 2019 09:54:03 +0100 [pkg] Version 0.25.8 0.25.8
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Nov 2019 09:54:03 +0100] rev 3983
[pkg] Version 0.25.8
Wed, 06 Nov 2019 09:53:45 +0100 [pkg] Drop reference to __pkginfo__.py in MANIFEST
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Nov 2019 09:53:45 +0100] rev 3982
[pkg] Drop reference to __pkginfo__.py in MANIFEST
Wed, 06 Nov 2019 09:52:59 +0100 [pkg] Include .conf in MANIFEST
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Nov 2019 09:52:59 +0100] rev 3981
[pkg] Include .conf in MANIFEST
Tue, 05 Nov 2019 17:06:14 +0100 Added tag 0.25.7 for changeset 12e22a78a4bf
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Nov 2019 17:06:14 +0100] rev 3980
Added tag 0.25.7 for changeset 12e22a78a4bf
Tue, 05 Nov 2019 17:06:08 +0100 [pkg] Version 0.25.7 0.25.7
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Nov 2019 17:06:08 +0100] rev 3979
[pkg] Version 0.25.7
Tue, 05 Nov 2019 17:03:43 +0100 [doc] Rewrite installation instructions
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Nov 2019 17:03:43 +0100] rev 3978
[doc] Rewrite installation instructions
Tue, 05 Nov 2019 16:51:23 +0100 [pkg] Keep cubicweb 3.26
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Nov 2019 16:51:23 +0100] rev 3977
[pkg] Keep cubicweb 3.26
Wed, 16 Oct 2019 12:09:14 +0200 Do not prevent deletion of Organization with an archival unit (XXX JIRA) draft
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 16 Oct 2019 12:09:14 +0200] rev 3976
Do not prevent deletion of Organization with an archival unit (XXX JIRA) It makes no sense to prevent deletion of an Organization when it's related to an OrganizationUnit because the latter has the "archival" role since we'll also delete the OrganizationUnit due to composite flag. In DontDeleteUnitIfArchival hook, we thus check if the organization is being also deleted before issuing the validation error.
Wed, 16 Oct 2019 12:08:55 +0200 Enable "show_composite" for deletion view of Organization draft
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 16 Oct 2019 12:08:55 +0200] rev 3975
Enable "show_composite" for deletion view of Organization It helps to see the tree of entities to be deleted.
Tue, 15 Oct 2019 16:49:13 +0200 Added tag 0.25.6 for changeset 2b766559e087
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 15 Oct 2019 16:49:13 +0200] rev 3974
Added tag 0.25.6 for changeset 2b766559e087
Tue, 15 Oct 2019 16:49:08 +0200 [pkg] Version 0.25.6 0.25.6
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 15 Oct 2019 16:49:08 +0200] rev 3973
[pkg] Version 0.25.6
Tue, 15 Oct 2019 16:44:47 +0200 Add a console.log() to make modal hidding work
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 15 Oct 2019 16:44:47 +0200] rev 3972
Add a console.log() to make modal hidding work cubicweb-relationwidget's javascript issues a "$modalDialog.modal('hide');" after the onValidate function finish but this happens to return directly and keep the gray backbround shown (the backdrop thing from bootstrap's modal). Apparently, this is because things happen too quickly. So by adding a console.log(), we slow things down and have the modal correctly hide.
Tue, 08 Oct 2019 14:10:09 +0200 Added tag 0.25.5 for changeset c7f6d6fb2b90
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 08 Oct 2019 14:10:09 +0200] rev 3970
Added tag 0.25.5 for changeset c7f6d6fb2b90
Tue, 08 Oct 2019 14:10:05 +0200 [pkg] Version 0.25.5 0.25.5
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 08 Oct 2019 14:10:05 +0200] rev 3969
[pkg] Version 0.25.5
Tue, 08 Oct 2019 14:09:12 +0200 Let EACSourceListItemView.entity_call() handle kwargs
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 08 Oct 2019 14:09:12 +0200] rev 3968
Let EACSourceListItemView.entity_call() handle kwargs Resolves https://sentry.logilab.fr/logilab/saemref/issues/506081/
Fri, 04 Oct 2019 15:12:02 +0200 Added tag 0.25.4 for changeset f348446ba080
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 15:12:02 +0200] rev 3965
Added tag 0.25.4 for changeset f348446ba080
Fri, 04 Oct 2019 15:11:33 +0200 [pkg] Version 0.25.4 0.25.4
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 15:11:33 +0200] rev 3964
[pkg] Version 0.25.4
Fri, 04 Oct 2019 15:11:12 +0200 Remove obsolete docker directory
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 15:11:12 +0200] rev 3963
Remove obsolete docker directory
Thu, 03 Oct 2019 16:27:45 +0200 Drop python2 support
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 16:27:45 +0200] rev 3962
Drop python2 support
Fri, 04 Oct 2019 15:05:56 +0200 [tox] Skip install for doc build
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 15:05:56 +0200] rev 3961
[tox] Skip install for doc build
Fri, 04 Oct 2019 15:03:45 +0200 [py3] Port our sphinx extension to python3
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 15:03:45 +0200] rev 3960
[py3] Port our sphinx extension to python3
Fri, 04 Oct 2019 11:24:52 +0200 Use new-style import of cubicweb-relationwidget
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 11:24:52 +0200] rev 3959
Use new-style import of cubicweb-relationwidget
Fri, 04 Oct 2019 11:21:10 +0200 Drop symlink of __pkginfo__.py
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 11:21:10 +0200] rev 3958
Drop symlink of __pkginfo__.py
Fri, 04 Oct 2019 10:40:46 +0200 Added tag 0.25.3 for changeset b4b83940efc0
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:40:46 +0200] rev 3956
Added tag 0.25.3 for changeset b4b83940efc0
Fri, 04 Oct 2019 10:40:02 +0200 [pkg] Version 0.25.3 0.25.3
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:40:02 +0200] rev 3955
[pkg] Version 0.25.3
Fri, 04 Oct 2019 10:39:44 +0200 Fix up abe752fe179a
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:39:44 +0200] rev 3954
Fix up abe752fe179a
Fri, 04 Oct 2019 10:38:53 +0200 Added tag 0.25.2 for changeset b390e0638bf9
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:38:53 +0200] rev 3953
Added tag 0.25.2 for changeset b390e0638bf9
Fri, 04 Oct 2019 10:38:51 +0200 [pkg] Version 0.25.2 0.25.2
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:38:51 +0200] rev 3952
[pkg] Version 0.25.2
Fri, 04 Oct 2019 10:38:40 +0200 [pkg] Specify encoding to open README
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:38:40 +0200] rev 3951
[pkg] Specify encoding to open README
Fri, 04 Oct 2019 10:33:21 +0200 Added tag 0.25.1 for changeset 8112f0760bd8
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:33:21 +0200] rev 3949
Added tag 0.25.1 for changeset 8112f0760bd8
Fri, 04 Oct 2019 10:33:01 +0200 [pkg] Version 0.25.1 0.25.1
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:33:01 +0200] rev 3948
[pkg] Version 0.25.1
Fri, 04 Oct 2019 10:18:03 +0200 [pkg] Specify encoding to open __pkginfo__.py
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 04 Oct 2019 10:18:03 +0200] rev 3947
[pkg] Specify encoding to open __pkginfo__.py
Thu, 03 Oct 2019 16:04:27 +0200 Added tag 0.25.0 for changeset 498e0de8c1e1
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 16:04:27 +0200] rev 3945
Added tag 0.25.0 for changeset 498e0de8c1e1
Thu, 03 Oct 2019 15:56:36 +0200 [pkg] Version 0.25.0 0.25.0
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 15:56:36 +0200] rev 3944
[pkg] Version 0.25.0
Thu, 19 Sep 2019 17:03:30 +0200 [tox] Add a py3 environment
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 19 Sep 2019 17:03:30 +0200] rev 3943
[tox] Add a py3 environment
Thu, 03 Oct 2019 14:21:32 +0200 [py3] Pass form parameters as "data" argument of http_publish test method
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 14:21:32 +0200] rev 3942
[py3] Pass form parameters as "data" argument of http_publish test method When giving a URL with a query string to http_publish() method it goes to admin_request_from_url() which wrongly parses the querystring as a dict of bytes (keys and values). Work around this cubicweb but by passing the query parameter as "data" argument of http_publish.
Thu, 03 Oct 2019 12:07:03 +0200 [py3] Open file in binary mode in SKOS (LCSV) import tests
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 12:07:03 +0200] rev 3941
[py3] Open file in binary mode in SKOS (LCSV) import tests
Thu, 03 Oct 2019 11:59:02 +0200 [py3] Add a key function to sort a list of dicts in ListArchivalUnitsServiceTC test
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:59:02 +0200] rev 3940
[py3] Add a key function to sort a list of dicts in ListArchivalUnitsServiceTC test
Thu, 03 Oct 2019 11:52:48 +0200 [py3] Encode JSON strings before feeding a binary view
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:52:48 +0200] rev 3939
[py3] Encode JSON strings before feeding a binary view
Thu, 03 Oct 2019 11:55:49 +0200 [py3] Compare "PageInfo" value with bytes in "eac.import" tests
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:55:49 +0200] rev 3938
[py3] Compare "PageInfo" value with bytes in "eac.import" tests
Thu, 03 Oct 2019 11:47:45 +0200 [py3] Decode HTTP response body before processing in views tests
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:47:45 +0200] rev 3937
[py3] Decode HTTP response body before processing in views tests
Thu, 03 Oct 2019 11:47:16 +0200 [py3] Open file in binary mode in EAC import tests
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:47:16 +0200] rev 3936
[py3] Open file in binary mode in EAC import tests
Thu, 03 Oct 2019 11:38:13 +0200 [py3] Decode XML before doing string comparisons
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:38:13 +0200] rev 3935
[py3] Decode XML before doing string comparisons
Thu, 03 Oct 2019 11:29:26 +0200 [py3] Use unicode strings in value generator for ark
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:29:26 +0200] rev 3934
[py3] Use unicode strings in value generator for ark
Wed, 02 Oct 2019 17:00:07 +0200 [py3] Decode HTTP response's body before doing string comparisons
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 02 Oct 2019 17:00:07 +0200] rev 3917
[py3] Decode HTTP response's body before doing string comparisons
Wed, 02 Oct 2019 16:51:19 +0200 [test] Simplify HTTP response comparison in /authorityrecord tests
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 02 Oct 2019 16:51:19 +0200] rev 3916
[test] Simplify HTTP response comparison in /authorityrecord tests Drop check about lxml version (we use a recent version now). Use assertIn() to avoid discrepancies in message's tail (between python2 and python3, in particular).
Wed, 02 Oct 2019 16:45:11 +0200 [py3] Decode serialized XML before using assertIn()
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 02 Oct 2019 16:45:11 +0200] rev 3915
[py3] Decode serialized XML before using assertIn()
Thu, 03 Oct 2019 14:49:40 +0200 [py3] print() unicode string in EAC import ctl command
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 14:49:40 +0200] rev 3914
[py3] print() unicode string in EAC import ctl command
Wed, 02 Oct 2019 16:38:05 +0200 [py3] Replace sys.stdout with a StringIO in EAC import tests
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 02 Oct 2019 16:38:05 +0200] rev 3913
[py3] Replace sys.stdout with a StringIO in EAC import tests Otherwise, this does not work in python3. In fact, we already do so for SKOS import tests (same file), so this is now consistent.
Wed, 02 Oct 2019 16:36:32 +0200 [py3] Do not decode cmd line arguments
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 02 Oct 2019 16:36:32 +0200] rev 3912
[py3] Do not decode cmd line arguments Apparently, these are unicode strings in python3.
Wed, 02 Oct 2019 16:14:22 +0200 [py3] Modernize
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 02 Oct 2019 16:14:22 +0200] rev 3911
[py3] Modernize
Thu, 03 Oct 2019 11:19:29 +0200 [pkg] Require cubicweb-seda >= 0.17.2 for python3 support
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 11:19:29 +0200] rev 3910
[pkg] Require cubicweb-seda >= 0.17.2 for python3 support
Tue, 01 Oct 2019 15:42:40 +0200 [pkg] Require cubicweb-eac >= 0.8.4 for python3 support
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 01 Oct 2019 15:42:40 +0200] rev 3909
[pkg] Require cubicweb-eac >= 0.8.4 for python3 support
Fri, 27 Sep 2019 15:29:10 +0200 [pkg] Require cubicweb-skos >= 1.6.0
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 27 Sep 2019 15:29:10 +0200] rev 3891
[pkg] Require cubicweb-skos >= 1.6.0
Tue, 24 Sep 2019 16:03:15 +0200 [pkg] Require cubicweb-relationwidget >= 0.5.1
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 24 Sep 2019 16:03:15 +0200] rev 3890
[pkg] Require cubicweb-relationwidget >= 0.5.1 This version supports python3 and has packaging improvements.
Tue, 24 Sep 2019 15:57:45 +0200 [pkg] Require cubicweb-vtimeline >= 0.6.0
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 24 Sep 2019 15:57:45 +0200] rev 3889
[pkg] Require cubicweb-vtimeline >= 0.6.0
Tue, 24 Sep 2019 15:21:29 +0200 [pkg] Require cubicweb-oaipmh >= 0.6.0
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 24 Sep 2019 15:21:29 +0200] rev 3888
[pkg] Require cubicweb-oaipmh >= 0.6.0 This version switches to package layout and supports python3.
Thu, 03 Oct 2019 08:48:24 +0200 Let flake8 ignore "build" and "dist" directories
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 03 Oct 2019 08:48:24 +0200] rev 3887
Let flake8 ignore "build" and "dist" directories
Fri, 20 Sep 2019 14:37:21 +0200 Added tag 0.24.1 for changeset 2d238fc468e5
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 20 Sep 2019 14:37:21 +0200] rev 3871
Added tag 0.24.1 for changeset 2d238fc468e5
Fri, 20 Sep 2019 14:37:10 +0200 [pkg] Version 0.24.1 0.24.1
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 20 Sep 2019 14:37:10 +0200] rev 3870
[pkg] Version 0.24.1
Thu, 19 Sep 2019 16:59:49 +0200 [tox] Drop obsolete environment variable
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 19 Sep 2019 16:59:49 +0200] rev 3868
[tox] Drop obsolete environment variable
Wed, 08 May 2019 19:39:45 +0200 [pkg] Upgrade cubicweb-squareui
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 08 May 2019 19:39:45 +0200] rev 3867
[pkg] Upgrade cubicweb-squareui Adjust import to use the package name accordingly.
Thu, 02 May 2019 11:05:22 +0200 Added tag 0.24.0 for changeset 2714efdbeb99
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 02 May 2019 11:05:22 +0200] rev 3866
Added tag 0.24.0 for changeset 2714efdbeb99
Thu, 02 May 2019 10:46:27 +0200 [pkg] Require psycopg2 < 2.8 0.24.0
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 02 May 2019 10:46:27 +0200] rev 3865
[pkg] Require psycopg2 < 2.8 Because of new exception classes mapping in psycopg2 2.8 for which cubicweb is not compatible yet. See https://lists.cubicweb.org/pipermail/cubicweb-devel/2019-April/002602.html
Thu, 02 May 2019 09:44:42 +0200 [pkg] Version 0.24.0
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 02 May 2019 09:44:42 +0200] rev 3864
[pkg] Version 0.24.0
Thu, 02 May 2019 09:43:51 +0200 [pkg] Require cubicweb-seda>=0.17.0
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 02 May 2019 09:43:51 +0200] rev 3863
[pkg] Require cubicweb-seda>=0.17.0
Fri, 08 Mar 2019 15:50:50 +0100 Added tag 0.23.0 for changeset b3ccd053b770
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 15:50:50 +0100] rev 3861
Added tag 0.23.0 for changeset b3ccd053b770
Fri, 08 Mar 2019 15:26:28 +0100 [pkg] Version 0.23.0 0.23.0
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 15:26:28 +0100] rev 3860
[pkg] Version 0.23.0
Fri, 08 Mar 2019 15:22:51 +0100 [seda] Fix creation url from "noresult" view for SEDAArchiveUnit (CONSEJIRA-684)
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 15:22:51 +0100] rev 3859
[seda] Fix creation url from "noresult" view for SEDAArchiveUnit (CONSEJIRA-684) Creation of new SEDAArchiveUnit should be done with the URL parameter "unit_type=unit_content". But when build the creation URL in "noresult" view (class SAEMNoResultView, introduced in 330d7061dda4), we used the default "add/<etype>" URL path for any entity type. Fortunately, configuration of adding entity types is already present in AddEntityComponent. So we factor out this configuration in a module variable and reuse it in SAEMNoResultView.
Fri, 08 Mar 2019 14:54:48 +0100 Use schema.entities() instead of vreg["etypes"] in "noresult" view
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 14:54:48 +0100] rev 3857
Use schema.entities() instead of vreg["etypes"] in "noresult" view Apparently, the latter is missing entity types present in the former. Not sure why. One example is "SKOSSource" entity type in my instance (and our demo instance).
Fri, 08 Mar 2019 14:29:49 +0100 Added tag 0.22.4 for changeset 5809216369b3
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 14:29:49 +0100] rev 3856
Added tag 0.22.4 for changeset 5809216369b3
Fri, 08 Mar 2019 14:29:41 +0100 [pkg] Version 0.22.4 0.22.4
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 14:29:41 +0100] rev 3855
[pkg] Version 0.22.4
Fri, 08 Mar 2019 14:29:22 +0100 [pkg] Require cubicweb-seda>=0.16.0
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Mar 2019 14:29:22 +0100] rev 3854
[pkg] Require cubicweb-seda>=0.16.0 This version contains an updated version of file categories vocabulary.
Thu, 07 Mar 2019 18:15:12 +0100 Do not import "Catégories de fichier" SEDA vocabulary on db creation draft
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 07 Mar 2019 18:15:12 +0100] rev 3853
Do not import "Catégories de fichier" SEDA vocabulary on db creation We want this vocabulary to be managed as a SKOSSource. We thus modify the LSCV_FILES list that is used during postcreate (in cubicweb-seda) to initialize SEDA vocabularies and drop this item. Then, since we still need to have the "scheme_relation_type" set on the vocabulary, we make the relation appear in ConceptScheme edition form.
Thu, 07 Feb 2019 10:20:07 +0100 Added tag 0.22.3 for changeset c80aec49e16e
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 07 Feb 2019 10:20:07 +0100] rev 3852
Added tag 0.22.3 for changeset c80aec49e16e
Thu, 07 Feb 2019 10:19:58 +0100 [pkg] Version 0.22.3 0.22.3
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 07 Feb 2019 10:19:58 +0100] rev 3851
[pkg] Version 0.22.3
Wed, 06 Feb 2019 09:49:15 +0100 Change create_default_naa() into a "get or create" function
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Feb 2019 09:49:15 +0100] rev 3850
Change create_default_naa() into a "get or create" function For some reason, I now fail to db-create an instance because the default ARK NAA appears to be created twice. Trying to fix this here by making function create_default_naa() (renamed as ark_naa) either get or create the ARK NAA. Not sure why the issue I see does not show up in tests, but the change probably does not hurt.
Tue, 22 Jan 2019 09:15:08 +0100 Fix UnicodeDecodeError when creating default ARK NAA
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 22 Jan 2019 09:15:08 +0100] rev 3849
Fix UnicodeDecodeError when creating default ARK NAA If "who" contains non-ascii characters, just calling text_type() will fail and we need to specify the encoding.
Wed, 06 Feb 2019 09:45:11 +0100 Use ARK NAA from config in get_store() monkeypatch (CONSEJIRA-668)
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 06 Feb 2019 09:45:11 +0100] rev 3848
Use ARK NAA from config in get_store() monkeypatch (CONSEJIRA-668) Value 25651 corresponds to archival department at CD33, this is a left-over of a previously hard-coded value. Follow-up on 096406495866.
Thu, 20 Dec 2018 16:37:25 +0100 Added tag 0.22.2 for changeset 9b1ee0eafa85
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 20 Dec 2018 16:37:25 +0100] rev 3843
Added tag 0.22.2 for changeset 9b1ee0eafa85
Thu, 20 Dec 2018 16:23:43 +0100 [pkg] Version 0.22.2, requiring cubicweb-seda>=0.15.2 (CONSEJIRA-643) 0.22.2
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 20 Dec 2018 16:23:43 +0100] rev 3842
[pkg] Version 0.22.2, requiring cubicweb-seda>=0.15.2 (CONSEJIRA-643)
Thu, 29 Nov 2018 10:14:23 +0100 Added tag 0.22.1 for changeset 6839a20efe04
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 29 Nov 2018 10:14:23 +0100] rev 3841
Added tag 0.22.1 for changeset 6839a20efe04
Thu, 29 Nov 2018 09:17:00 +0100 [pkg] Version 0.22.1 0.22.1
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 29 Nov 2018 09:17:00 +0100] rev 3840
[pkg] Version 0.22.1
Thu, 29 Nov 2018 09:53:31 +0100 [pkg] Require cubicweb-seda>=0.15.1
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 29 Nov 2018 09:53:31 +0100] rev 3839
[pkg] Require cubicweb-seda>=0.15.1
Wed, 31 Oct 2018 14:48:42 +0100 Added tag 0.22.0 for changeset 037a9a9d4cf6
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 31 Oct 2018 14:48:42 +0100] rev 3836
Added tag 0.22.0 for changeset 037a9a9d4cf6
Wed, 31 Oct 2018 14:48:29 +0100 [pkg] Version 0.22.0 0.22.0
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 31 Oct 2018 14:48:29 +0100] rev 3835
[pkg] Version 0.22.0 Also moved 0.21.2 migration script since this version will never exist.
Wed, 31 Oct 2018 11:53:42 +0100 Use psycopg2-binary in dev requirements
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 31 Oct 2018 11:53:42 +0100] rev 3834
Use psycopg2-binary in dev requirements Getting rid of a UserWarning.
Tue, 30 Oct 2018 14:09:51 +0100 Add options to configure the default ARK NAA entity (CONSEJIRA-614)
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 30 Oct 2018 14:09:51 +0100] rev 3833
Add options to configure the default ARK NAA entity (CONSEJIRA-614) We need an initial ArkNameAssigningAuthority entity during db-init since we import concept schemes used as SEDA vocabularies and these need an ARK NAA in order their ARK to be generated. Previously, we used ADGIRONDE ARK NAA number as an hard-coded value in init_seda_scheme() but this is problematic for people wanting to use another authority. We thus introduce two configuration values to be set at instance creation and used during SEDA vocabularies import. Since I failed to set a value for these options during test setup that could be available early enough for "postcreate" script to see it, this script contains a "if config.mode == 'TEST'" that by-pass config reading and use values previously set in testutils.naa(); in testutils.naa() we no longer create an ArkNameAssigningAuthority entity but now always retrieve the same entity using the "what" value setup in postcreate. In automatic web tests, we need a special value generator to avoid getting into uniqueness conflict when an ArkNameAssigningAuthority with what=0 would be generated. We use a 5 digits integer as this is what the standards suggests.
Mon, 29 Oct 2018 15:54:26 +0100 [pkg] Use version of cubicweb-eac version 0.8.1
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 29 Oct 2018 15:54:26 +0100] rev 3831
[pkg] Use version of cubicweb-eac version 0.8.1
Tue, 30 Oct 2018 12:28:11 +0100 Add options to configure the default ARK NAA entity (CONSEJIRA-614) draft
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 30 Oct 2018 12:28:11 +0100] rev 3824
Add options to configure the default ARK NAA entity (CONSEJIRA-614) We need an initial ArkNameAssigningAuthority entity during db-init since we import concept schemes used as SEDA vocabularies and these need an ARK NAA in order their ARK to be generated. Previously, we used ADGIRONDE ARK NAA number as an hard-coded value in init_seda_scheme() but this is problematic for people wanting to use another authority.
Tue, 30 Oct 2018 10:49:38 +0100 [test] Use cw_clear_all_caches() instead of deprecated clear_all_caches()
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 30 Oct 2018 10:49:38 +0100] rev 3823
[test] Use cw_clear_all_caches() instead of deprecated clear_all_caches()
Tue, 30 Oct 2018 10:48:22 +0100 [test] Use assertFalse instead of deprecated failIf
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 30 Oct 2018 10:48:22 +0100] rev 3822
[test] Use assertFalse instead of deprecated failIf
Mon, 29 Oct 2018 15:54:42 +0100 [pkg] Require new version of cubicweb-seda
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 29 Oct 2018 15:54:42 +0100] rev 3816
[pkg] Require new version of cubicweb-seda
Mon, 29 Oct 2018 12:25:29 +0100 flake8 fixes
Philippe Pepiot <philippe.pepiot@logilab.fr> [Mon, 29 Oct 2018 12:25:29 +0100] rev 3814
flake8 fixes Fix new flake8 3.6.0 checks, mostly line break after binary operator and local variable 'exc' is assigned to but never used.
Fri, 19 Oct 2018 15:01:03 +0200 Allow direct deletion of SEDAArchiveUnit entities (CONSEJIRA-591)
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 19 Oct 2018 15:01:03 +0200] rev 3813
Allow direct deletion of SEDAArchiveUnit entities (CONSEJIRA-591) This is a companion change of the previous one. Here we allow direct deletion of a SEDAArchiveUnit entity, even without going through deletion of the "seda_archive_unit" relation first.
Fri, 19 Oct 2018 15:00:07 +0200 Allow deletion of "seda_archive_unit" relation (CONSEJIRA-591)
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 19 Oct 2018 15:00:07 +0200] rev 3812
Allow deletion of "seda_archive_unit" relation (CONSEJIRA-591) Per CONSEJIRA-591, we now want to allow deletion of a SEDAArchiveUnit entity even when related to a SEDAArchiveTransfer in public workflow state (whereas we used to want to disallow this). The first step for this is to allow deletion of the "seda_archive_unit" relation. As permission of this relation is defined in cubicweb-seda, we override it in post-build callback. In the new test, we first check that deletion of the relation works. Then we check that deletion of the unlinked entity (SEDAArchiveUnit) is also possible. This should be true as it is no longer related to an archive transfer but due to an integrity bug in cubicweb-seda which makes the "container" relation out-of-sync, this does not work. So we require a new cubicweb-seda version with a fix for this issue and the check passes.
Fri, 19 Oct 2018 09:15:31 +0200 [migration] Fix "container" relation table and schema
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 19 Oct 2018 09:15:31 +0200] rev 3807
[migration] Fix "container" relation table and schema It has been noticed on running instances that the "container" relation table (named container_relation) contains rows with their eid_from column corresponding to entities of type Activity which is not supposed to be subject of this relation. This comes from changeset 03590369788c in which this entity type was added to the entity types to be skipped when building the compound graph of SEDAArchiveTransfer entity type (this graph is then used to programmatically create relation definitions in cubicweb-seda). Yet, at that time, not database migration was done, thus leading to the inconsistent state we have now. (One noticeable and annoying issue is that some entities cannot be deleted due to foreign keys from these spurious rows). So we fix the database by dropping the relation definitions from schema and then deleting rows of the container_relation table. A manual deletion, in SQL, is needed because deletion of relation definition would not trigger data deletion (as RQL would apparently ignore rows not matching know subject entity type).
Tue, 09 Oct 2018 13:42:42 +0200 Added tag 0.21.1 for changeset 3640a1bd46e0
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 09 Oct 2018 13:42:42 +0200] rev 3806
Added tag 0.21.1 for changeset 3640a1bd46e0
Tue, 09 Oct 2018 13:42:37 +0200 [pkg] Version 0.21.1 0.21.1
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 09 Oct 2018 13:42:37 +0200] rev 3805
[pkg] Version 0.21.1
Tue, 09 Oct 2018 12:03:49 +0200 Clear broader_concept relation before setting when reparenting a Concept
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 09 Oct 2018 12:03:49 +0200] rev 3804
Clear broader_concept relation before setting when reparenting a Concept Before this change, calling JQTree adapter's reparent() method on a concept under another concept would add another broader_concept concept relation instead of replacing the existing one. This is because .cw_set() *does not* replace relations but adds them. So we must always clear the broader_concept relation (calling .cw_set(broader_concept=None)) and reset it (in case the target parent is another concept). Adding non-regression tests. Follow-up on f37b9abae3a9.
Tue, 09 Oct 2018 10:09:43 +0200 Added tag 0.21.0 for changeset 2577e6427d63
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 09 Oct 2018 10:09:43 +0200] rev 3803
Added tag 0.21.0 for changeset 2577e6427d63
Tue, 09 Oct 2018 09:35:46 +0200 [pkg] Version 0.21.0 0.21.0
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 09 Oct 2018 09:35:46 +0200] rev 3802
[pkg] Version 0.21.0
Tue, 09 Oct 2018 09:35:14 +0200 [pkg] Use now released version of eac, skos and seda cubes
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 09 Oct 2018 09:35:14 +0200] rev 3801
[pkg] Use now released version of eac, skos and seda cubes
Mon, 24 Sep 2018 15:44:38 +0200 Use 'nth-child' selector instead of classes to display striped lists
Frank Bessou <frank.bessou@logilab.fr> [Mon, 24 Sep 2018 15:44:38 +0200] rev 3800
Use 'nth-child' selector instead of classes to display striped lists Currently, the markup generated when rendering a list looks like: <ul> <li class='odd'>..</li> <li class='even'>..</li> <li class='odd'>..</li> <li class='even'>..</li> </ul> ... where 'odd' and 'even' classes are used to display a different background on cells with an event index. When the list is re-ordered dynamically (by drag and drop for example) the stripes order can become inconsistent: <ul> <li class='odd'>..</li> <li class='odd'>..</li> <li class='even'>..</li> <li class='even'>..</li> </ul> This is fixed by using the 'nth-child' selector
Tue, 25 Sep 2018 11:31:00 +0200 Use a jQuery widget to make some lists sortable
Frank Bessou <frank.bessou@logilab.fr> [Tue, 25 Sep 2018 11:31:00 +0200] rev 3799
Use a jQuery widget to make some lists sortable This jQuery widget calls a new ajax controller which uses ISortable properties to sort elements in database. We add a 'collection' property to Sortable adapter which returns a list the list of entities which are in the same collection (related to a common parent through the same relation). We make RelatedEntitiesListView use a listvid property to select which view must be used to display a given list. The SortableListView adds js corresponding to sorting and adds the needed attributes on the generated markup.
Tue, 25 Sep 2018 11:39:19 +0200 Display sortable entities in a sorted view
Frank Bessou <frank.bessou@logilab.fr> [Tue, 25 Sep 2018 11:39:19 +0200] rev 3796
Display sortable entities in a sorted view We introduce a related_rset method in RelatedEntitiesListView this allows to apply post-processes on the rset its subclasses. The SortableListViewMixin provide an implementation for the related_rset method. It sorts the rset according to the 'index' relation. This mixin is used to display functions and mandates on authority records.
Tue, 25 Sep 2018 11:48:12 +0200 Initialize index of sortable related entities on creation
Frank Bessou <frank.bessou@logilab.fr> [Tue, 25 Sep 2018 11:48:12 +0200] rev 3795
Initialize index of sortable related entities on creation We will want to display newly created entities at the end of the list so we use their eid as index. We introduce an ISortable adapter which we make selectable for mandates and functions. This adapater will be used to ease the manipulation of the sortable things and is used here to select the hook.
Wed, 19 Sep 2018 12:28:50 +0200 Add an "index" attribute to Mandate and AgentFunction
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 19 Sep 2018 12:28:50 +0200] rev 3778
Add an "index" attribute to Mandate and AgentFunction We want to be able to sort mandates and functions of related to a given autority record. So introduce an index to keep track of respective position. We hide this attribute from primary view and autoform since it is meant to be managed by a hook or a dedicated UI logic (to be implemented in forthcoming patches). CONSEJIRA-578, CONSEJIRA-584.
Fri, 28 Sep 2018 15:35:24 +0200 Use existing popovers in ConceptScheme jquery tree view
Frank Bessou <frank.bessou@logilab.fr> [Fri, 28 Sep 2018 15:35:24 +0200] rev 3777
Use existing popovers in ConceptScheme jquery tree view To make popovers work in the dynamically loaded tree, we add a script that adds popovers on all elements with a "data-toggle='popover'" attribute (this is done in a new PopoverJQTreeView). We adapt the ConceptPopoverView to make it work within a JQTree and we add the "data-toggle" attribute on the label anchors.
Thu, 27 Sep 2018 12:04:48 +0200 Use JQTree to display Concepts in ConceptScheme tab view
Frank Bessou <frank.bessou@logilab.fr> [Thu, 27 Sep 2018 12:04:48 +0200] rev 3776
Use JQTree to display Concepts in ConceptScheme tab view We extend the ConceptSchemeConceptsTab to use the jqtree.tree view and we disable the base class. The render_tree method is added in an unreleased version of cubicweb-skos.
Mon, 01 Oct 2018 15:33:32 +0200 Make ConceptSchemes and Concepts adaptable to IJQTree
Frank Bessou <frank.bessou@logilab.fr> [Mon, 01 Oct 2018 15:33:32 +0200] rev 3775
Make ConceptSchemes and Concepts adaptable to IJQTree We implement needed adapters so that a ConceptScheme can be rendered as a JQTree: IJQTreeAdapter and ITreeBaseAdapter. Since Concept is already adaptable to IContained we make it adaptable to ITreeBase by extending the IContainedToITreeBase adapter. We have to provide js callback to JQTree adapters which tells if it is possible the move the Concept to a given node. Since a Concept can always be moved we make this callback return always true. We use a currently unrelease version of cubicweb-seda to avoid having a "seda profile tree" contextual component displayed on the concept scheme page.
Mon, 01 Oct 2018 15:28:57 +0200 Use "import cubicweb_skos" instead of "import cubes.skos"
Frank Bessou <frank.bessou@logilab.fr> [Mon, 01 Oct 2018 15:28:57 +0200] rev 3774
Use "import cubicweb_skos" instead of "import cubes.skos" This change is actually done in unreleased versions of cube eac and cube seda.
Fri, 28 Sep 2018 10:31:07 +0200 [skos] Add back vreg.register_all() call in views's registration callback
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 28 Sep 2018 10:31:07 +0200] rev 3770
[skos] Add back vreg.register_all() call in views's registration callback This got dropped by mistake in 1efe603bff6c.
Tue, 25 Sep 2018 11:54:16 +0200 [eac] Display AuthorityRecord's name entries as a list with authorized forms first
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 25 Sep 2018 11:54:16 +0200] rev 3761
[eac] Display AuthorityRecord's name entries as a list with authorized forms first We use a "list" view to display the name_entry_for relation on AuthorityRecord and sort items by putting those with "form_variant == 'authorized'" first as requested in CONSEJIRA-580.
Wed, 19 Sep 2018 11:40:09 +0200 Avoid 404 on /<etype> when there's no entity yet (CONSEJIRA-511)
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 19 Sep 2018 11:40:09 +0200] rev 3754
Avoid 404 on /<etype> when there's no entity yet (CONSEJIRA-511) The default behavior of CubicWeb is to raise a NotFound exception when the result set is empty on a /<etype> route. This is not very user friendly (not even sure this is semantically correct). So we override this behavior by selecting the "noresult" view in such cases through RestPathEvaluator.set_vid_for_rset() (where the NotFound exception is raised). Since we added a custom version of this view in the previous changeset, the resulting page will contain an invite to create the first entity of <etype>.
Wed, 19 Sep 2018 11:20:42 +0200 Add a custom "noresult" view with an invite to create the first entity (CONSEJIRA-511)
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 19 Sep 2018 11:20:42 +0200] rev 3753
Add a custom "noresult" view with an invite to create the first entity (CONSEJIRA-511) We add a custom "noresult" view and replace the default one that, instead of displaying "no result matching query", will display a more specific message when the query concerns a single entity type (for instance, the /sedalib route corresponds to 'Any X WHERE X is SEDAArchiveUnit' RQL query). This message is complemented by an invite to create the first entity if permission is granted. This helps bootstrapping the instance after db-creation.
Thu, 05 Jul 2018 08:21:04 +0200 Added tag 0.20.7 for changeset 6c65689c7ed7
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 05 Jul 2018 08:21:04 +0200] rev 3748
Added tag 0.20.7 for changeset 6c65689c7ed7
Wed, 04 Jul 2018 17:44:41 +0200 [pkg] Version 0.20.7 0.20.7
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 04 Jul 2018 17:44:41 +0200] rev 3747
[pkg] Version 0.20.7
Wed, 04 Jul 2018 17:42:29 +0200 Backed out changeset 1ac675a21bc1
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 04 Jul 2018 17:42:29 +0200] rev 3746
Backed out changeset 1ac675a21bc1 This breaks NCName constraint on xml:id attribute. See CONSEJIRA-510.
Thu, 28 Jun 2018 16:39:30 +0200 Added tag 0.20.6 for changeset a32468e9e013
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Jun 2018 16:39:30 +0200] rev 3745
Added tag 0.20.6 for changeset a32468e9e013
Thu, 28 Jun 2018 16:39:22 +0200 [pkg] Version 0.20.6 0.20.6
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Jun 2018 16:39:22 +0200] rev 3744
[pkg] Version 0.20.6
Thu, 28 Jun 2018 16:38:55 +0200 Fix selector of SAEMXmlIdAdapter
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 28 Jun 2018 16:38:55 +0200] rev 3743
Fix selector of SAEMXmlIdAdapter The implementation introduced in 1ac675a21bc1 was blatantly wrong as not all entity types has an 'ark' attribute.
Tue, 26 Jun 2018 14:55:19 +0200 Added tag 0.20.5 for changeset f8f33ed3f3cf
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 26 Jun 2018 14:55:19 +0200] rev 3742
Added tag 0.20.5 for changeset f8f33ed3f3cf
Tue, 26 Jun 2018 14:55:12 +0200 [pkg] Version 0.20.5 0.20.5
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 26 Jun 2018 14:55:12 +0200] rev 3741
[pkg] Version 0.20.5
Tue, 26 Jun 2018 14:22:37 +0200 [pkg] Depends on cubicweb-seda >= 0.13.3
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 26 Jun 2018 14:22:37 +0200] rev 3740
[pkg] Depends on cubicweb-seda >= 0.13.3
Tue, 26 Jun 2018 09:43:55 +0200 [seda] Restrict targets of "originating agency" relation to published AutorityRecord
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 26 Jun 2018 09:43:55 +0200] rev 3739
[seda] Restrict targets of "originating agency" relation to published AutorityRecord We add an RQL constraint on "seda_originating_agency_to" (class name originating_agency_originating_agency_to in cubicweb-seda) to restrict possible target AutorityRecord entities with a workflow state "published". We need to do this with a full RQL constraint (although a vocabulary constraint might be appropriate from a business point of view) because the relation widget is used in archive unit form to render this relation and the latter does not support vocabulary constraints. CONSEJIRA-507.
Mon, 25 Jun 2018 10:30:04 +0200 [seda] Use ARK identifier as xml:id attribute in SEDA profiles export
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 25 Jun 2018 10:30:04 +0200] rev 3738
[seda] Use ARK identifier as xml:id attribute in SEDA profiles export We add a version of IXmlId adapter for entity type supporting the "ark" attribute and use this identifier as xml:id attribute in SEDA profiles export. This concerns the "<xsd:element name="Contains">" element present with a profile contains archive units in particular, hence the added unit tests. CONSEJIRA-510.
Thu, 24 May 2018 11:25:55 +0200 Added tag 0.20.4 for changeset 1ac1206655da
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 24 May 2018 11:25:55 +0200] rev 3733
Added tag 0.20.4 for changeset 1ac1206655da
Thu, 24 May 2018 10:51:42 +0200 [pkg] Version 0.20.4 0.20.4
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 24 May 2018 10:51:42 +0200] rev 3732
[pkg] Version 0.20.4
Fri, 18 May 2018 12:17:48 +0200 [seda] Do not use "container" relation when looking for archive unit parent
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 18 May 2018 12:17:48 +0200] rev 3730
[seda] Do not use "container" relation when looking for archive unit parent We rework archive_unit_parent() function introduced in 2f47a5deb560 to avoid relying on "container" relation to find a parent to a SEDAArchiveUnit entity through which we can generate/locate ARK identifiers. Instead, we use the business SEDA model and walk relations until we get a parent archive unit or archive transfer. The function is also recursive now, which should make it able to handle nested archive units. We add a regression test about cloning a SEDAArchiveTransfer with nested archive units; this test failed before changes in archive_unit_parent function. Follow up on CONSEJIRA-491.
Tue, 15 May 2018 16:56:37 +0200 Added tag 0.20.3 for changeset 9aa796dc6ac5
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 15 May 2018 16:56:37 +0200] rev 3729
Added tag 0.20.3 for changeset 9aa796dc6ac5
Tue, 15 May 2018 16:56:13 +0200 [pkg] Version 0.20.3 0.20.3
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 15 May 2018 16:56:13 +0200] rev 3728
[pkg] Version 0.20.3
Mon, 14 May 2018 16:36:57 +0200 [seda] Handle all target types for seda_archive_unit relation in archive unit ARK attribution
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 14 May 2018 16:36:57 +0200] rev 3727
[seda] Handle all target types for seda_archive_unit relation in archive unit ARK attribution Relation seda_archive_unit may have SEDAArchiveTransfer and SEDASeqAltArchiveUnitArchiveUnitRefIdManagement entity types as object. Previously, we only handled the first case which occurs when a SEDAArchiveUnit is related directly to a SEDAArchiveTransfer. However, when relating a SEDAArchiveUnit to another SEDAArchiveUnit our previous assumption (that the target of seda_archive_unit relation has an "ark_naa" relation) breaks, thus making it impossible to insert such a relation. We thus check the entity type of target of this relation in ARK locator and generator for SEDAArchiveUnit. In case a SEDASeqAltArchiveUnitArchiveUnitRefIdManagement entity type is obtained, we use the "container" relation to obtain the archive transfer and its ARK NAA. There might be a "business" way to obtain this information but, given the complexity of the data model, I could not find it. The consequence of this *fix* is that ARK identifiers of nested SEDAArchiveUnit are now always qualified with respect to their parent. This does not change if the parent is a SEDAArchiveTransfer but this changes if it is another SEDAArchiveUnit since, previously, the nested entity would simply have had an unqualified ARK: hopefully, it makes sense from a business perspective... CONSEJIRA-491.
Mon, 14 May 2018 16:29:15 +0200 [seda] Hide "ark_naa" relation for "nested" SEDAArchiveUnit entities
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 14 May 2018 16:29:15 +0200] rev 3725
[seda] Hide "ark_naa" relation for "nested" SEDAArchiveUnit entities When a SEDAArchiveUnit is related to another SEDAArchiveUnit or to a SEDAArchiveTransfer through "seda_archive_unit" relation, it usually inherits its ARK naa from its parent. So we hide the relation from entity primary view and autoform in these cases. This is done through derivation of base uicfg tags; in the case of autoform, we need a special predicate to parse the __linkto form parameter and detect if we are creating an "nested" entity or not. CONSEJIRA-491
Mon, 14 May 2018 15:47:23 +0200 Fix a typo in i18n
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 14 May 2018 15:47:23 +0200] rev 3724
Fix a typo in i18n
Mon, 14 May 2018 15:28:22 +0200 [seda] Make "ark_naa" relation from SEDAArchiveUnit non-required
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 14 May 2018 15:28:22 +0200] rev 3723
[seda] Make "ark_naa" relation from SEDAArchiveUnit non-required This should be required only when the entity is created alone. When it's related to a SEDAArchiveTransfer or another SEDAArchiveUnit, the ARK naa of its parent should be used. In schema.py, we add a new class for "ark_naa" relation definition with SEDAArchiveUnit as subject. Since the "description" attribute is common with mandatory_ark_naa class, we factor it out in "_ark_naa" base class. CONSEJIRA-491
Tue, 17 Apr 2018 11:42:01 +0200 Added tag 0.20.2 for changeset 1aaf1ac9f38c
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 17 Apr 2018 11:42:01 +0200] rev 3721
Added tag 0.20.2 for changeset 1aaf1ac9f38c
Tue, 17 Apr 2018 11:41:33 +0200 [pkg] Version 0.20.2 0.20.2
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 17 Apr 2018 11:41:33 +0200] rev 3720
[pkg] Version 0.20.2
Tue, 17 Apr 2018 11:33:02 +0200 Add missing migration from 0.20.0
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 17 Apr 2018 11:33:02 +0200] rev 3719
Add missing migration from 0.20.0 Followup on ac4b89c371aa. CONSEJIRA-447.
Tue, 17 Apr 2018 09:11:18 +0200 [seda] Consider all relations before validating deprecation of a ConceptScheme
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 17 Apr 2018 09:11:18 +0200] rev 3718
[seda] Consider all relations before validating deprecation of a ConceptScheme Followup on a65281f047d1 in which not all possible relations were considered. I scanned all relation definitions in cubicweb-seda's schema that points to/from ConceptScheme. Previous test is still valid although it only consider one type of relation. CONSEJIRA-447
Mon, 16 Apr 2018 10:06:38 +0200 Added tag 0.20.1 for changeset 8c865a76b8d3
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 16 Apr 2018 10:06:38 +0200] rev 3717
Added tag 0.20.1 for changeset 8c865a76b8d3
Mon, 16 Apr 2018 10:06:03 +0200 [pkg] Version 0.20.1 0.20.1
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 16 Apr 2018 10:06:03 +0200] rev 3716
[pkg] Version 0.20.1
Mon, 16 Apr 2018 09:39:43 +0200 Hide transferring agency identifier field from archive unit description in non-simplified profiles
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 16 Apr 2018 09:39:43 +0200] rev 3715
Hide transferring agency identifier field from archive unit description in non-simplified profiles Followup on d9f0ae5b83c6. CONSEJIRA-449
Fri, 13 Apr 2018 16:47:09 +0200 Added tag 0.20.0 for changeset c3b5cdf3410f
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 13 Apr 2018 16:47:09 +0200] rev 3714
Added tag 0.20.0 for changeset c3b5cdf3410f
Fri, 13 Apr 2018 16:45:12 +0200 [pkg] Version 0.20.0 0.20.0
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 13 Apr 2018 16:45:12 +0200] rev 3713
[pkg] Version 0.20.0
Fri, 13 Apr 2018 14:55:28 +0200 Allow non-simplified profile mode to be selected upon SEDAArchiveTransfer creation
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 13 Apr 2018 14:55:28 +0200] rev 3712
Allow non-simplified profile mode to be selected upon SEDAArchiveTransfer creation CONSEJIRA-449
Fri, 13 Apr 2018 10:54:40 +0200 [test] Set archival role to "deposit" for the other OU in security tests
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 13 Apr 2018 10:54:40 +0200] rev 3710
[test] Set archival role to "deposit" for the other OU in security tests There's a 'S archival_role R, R name "deposit"' RQL constraint on "use_profile" relation which we are trying to insert in test_create_update_organizationunit_in_own_organization. The previous data did not fulfill this constraint. This is not a problem in most cases because we should get into an Unauthorized error first but it appears that the validation check might occur before permissions check sometimes (it occurred in CI environment, when running all tests; I could not reproduce locally).
Fri, 13 Apr 2018 14:00:00 +0200 Prevent deprecation of a concept scheme used by a relation type
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 13 Apr 2018 14:00:00 +0200] rev 3709
Prevent deprecation of a concept scheme used by a relation type ConceptScheme entities may used to qualify a relation type through the "scheme_relation_type" relation in order to restrict Concept entities as possible targets of that relation. We must not be able to deprecate a ConceptScheme used in such cases. This is implemented in a hook because it's not possible to express this in workflow permissions because group permissions would be checked first and, if they succeed, RQL conditions will not be checked. Also, this is arguably more an integrity check than a permission check, so a hook is fine. CONSEJIRA-447
Thu, 12 Apr 2018 15:32:21 +0200 [schema] Allow to delete/update a concept scheme in "deprecated" state
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 12 Apr 2018 15:32:21 +0200] rev 3708
[schema] Allow to delete/update a concept scheme in "deprecated" state CONSEJIRA-447.
Thu, 12 Apr 2018 15:24:09 +0200 Let publication_permissions() accept a list of states to configure permissions
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 12 Apr 2018 15:24:09 +0200] rev 3707
Let publication_permissions() accept a list of states to configure permissions We thus turn RQL expression to use IN, hence changes in permissions test.
Wed, 11 Apr 2018 17:42:04 +0200 Fix __select__ of QualifiedARKIdentifierGenerator subclasses
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 17:42:04 +0200] rev 3691
Fix __select__ of QualifiedARKIdentifierGenerator subclasses I.e. use the direct parent class, instead of the mixin. This is equivalent since QualifiedARKIdentifierGenerator has the same __select__ than the mixin.
Wed, 11 Apr 2018 16:29:46 +0200 [seda] Configure SEDAArchiveUnit form for qualified ARK generation
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 16:29:46 +0200] rev 3690
[seda] Configure SEDAArchiveUnit form for qualified ARK generation CONSEJIRA-431
Wed, 11 Apr 2018 12:22:52 +0200 [seda] Generate a qualified ARK identifier for archive unit related to a transfer
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 12:22:52 +0200] rev 3689
[seda] Generate a qualified ARK identifier for archive unit related to a transfer When a SEDAArchiveUnit is created or related to an archive transfer we generated an ARK identifier for this entity that is qualified with respect to the transfer's ARK identifier. This is implemented through a subclass of QualifiedARKIdentifierGenerator that is selected for SEDAArchiveUnit entity type. This is similar to how Concept entities get their ARK from their related ConceptScheme. Note that we had to fix an "ark" attribute for the SEDAArchiveUnit created in automatic views tests (test_saem_ref.py) because otherwise, an ark would have been generated based on parent SEDAArchiveTransfer entity which itself has a generated ark that does not match what's expected to produce a qualified identifier (namely, it has a random NAAN, which corresponds to no existing NAA and its NAME part would have no prefix/suffix). CONSEJIRA-431
Wed, 11 Apr 2018 15:50:10 +0200 [seda] Add an "ark" attribute to SEDAArchiveUnit
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 15:50:10 +0200] rev 3688
[seda] Add an "ark" attribute to SEDAArchiveUnit We add an "ark" attribute and an "ark_naa" relation to SEDAArchiveUnit. A couple of things has to be implemented in order things to keep working: * we need a IArkNAALocator adapter for this entity type; as for other entity types that can live under a parent entity (like Concept vs. ConceptScheme), we need to handle retrieval of the ARK NAA entity either directly or using the parent (first). * we need to make sure that the "ark" attribute is not copied, thus the custom SEDAArchiveUnit entity class: this is completely identical to SEDAArchiveTransfer defined just above in entities/seda.py. * we need a custom version of testutils.create_archive_unit() which would retrieve the "ark_naa" relation either directly or from the parent, if it is set. Tests for ARK generation added in unittest_hooks.py. CONSEJIRA-431
Wed, 11 Apr 2018 14:17:14 +0200 Set AssignARKHook's category to "metadata"
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 14:17:14 +0200] rev 3685
Set AssignARKHook's category to "metadata" The reason for this is that the IClonable adapter will disable all hooks but that in "metadata" category. So when AssignARKHook is not in this category, any entity that is cloned but needs an "ark" attribute (which is obviously not copied as it must be unique) while not have an "ark" assigned. This will become needed in forthcoming changesets in which and "ark" attribute will be added to SEDAArchiveUnit entity type and which is involved in clone operation of their respective archive transfer.
Wed, 11 Apr 2018 12:06:02 +0200 [ark] Extract a direct_naa_what function out of DirectArkNAALocator adapter
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 12:06:02 +0200] rev 3681
[ark] Extract a direct_naa_what function out of DirectArkNAALocator adapter
Wed, 11 Apr 2018 10:15:07 +0200 [pkg] Bump cubicweb-oaipmh to 0.5.0
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 11 Apr 2018 10:15:07 +0200] rev 3679
[pkg] Bump cubicweb-oaipmh to 0.5.0 This version includes a configuration option "list-max-size" to control how many items list-like OAI-PMH requests return. CONSEJIRA-448
Tue, 27 Mar 2018 11:03:14 +0200 Added tag 0.19.2 for changeset 9694b883989e
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 27 Mar 2018 11:03:14 +0200] rev 3678
Added tag 0.19.2 for changeset 9694b883989e
Tue, 27 Mar 2018 10:34:55 +0200 [pkg] Version 0.19.2 0.19.2
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 27 Mar 2018 10:34:55 +0200] rev 3677
[pkg] Version 0.19.2
Tue, 27 Mar 2018 10:34:32 +0200 [seda] Drop fixed rng:value tag from TransferringAgencyArchiveIdentifier tag
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 27 Mar 2018 10:34:32 +0200] rev 3676
[seda] Drop fixed rng:value tag from TransferringAgencyArchiveIdentifier tag We actually want to allow users of the profile to insert any value in this TransferringAgencyArchiveIdentifier tag and the previous "empty" rng:value tag was actually preventing this. See discussion in CONSEJIRA-445.
Fri, 23 Mar 2018 09:55:59 +0100 Added tag 0.19.1 for changeset 7c009a528009
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 23 Mar 2018 09:55:59 +0100] rev 3675
Added tag 0.19.1 for changeset 7c009a528009
Fri, 23 Mar 2018 09:55:41 +0100 [pkg] Version 0.19.1 0.19.1
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 23 Mar 2018 09:55:41 +0100] rev 3674
[pkg] Version 0.19.1
Thu, 22 Mar 2018 15:16:09 +0100 [seda] Hide seda_transferring_agency_archive_unit_identifier from archiveunit form
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Mar 2018 15:16:09 +0100] rev 3673
[seda] Hide seda_transferring_agency_archive_unit_identifier from archiveunit form Followup on 1493be198e6e, which I thought was enough but is apparently not working for the edition form. The trick here is that we cannot rely on a plain "autoform_section" uicfg rule because SAEM is using the simplified version of SEDA profiles for which there is a dedicated (derived) uicfg rtag in cubicweb_seda/views/simplified.py. So we add a new tag to hide this seda_transferring_agency_archive_unit_identifier relation using this "simplified_afs" relation tag object. CONSEJIRA-445
Wed, 21 Mar 2018 17:11:32 +0100 Added tag 0.19.0 for changeset 533191804fc2
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Mar 2018 17:11:32 +0100] rev 3672
Added tag 0.19.0 for changeset 533191804fc2
Wed, 21 Mar 2018 17:11:07 +0100 [pkg] Version 0.19.0 0.19.0
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Mar 2018 17:11:07 +0100] rev 3671
[pkg] Version 0.19.0
Wed, 21 Mar 2018 14:10:25 +0100 [seda] Unlink deprecated profiles from OrganizationUnit using them
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Mar 2018 14:10:25 +0100] rev 3670
[seda] Unlink deprecated profiles from OrganizationUnit using them We only want "published" profiles to be linked to an OrganizationUnit. There's already an RQL constraint on "use_profile" relation for that, but upon change of the SEDA archive transfer workflow state, this constraint is not fulfilled anymore. So add a hook that drops the relation when state changes to "deprecated". (CONSEJIRA-427)
Wed, 21 Mar 2018 14:07:23 +0100 [test] Give a more specific name to test method in SEDAArchiveTransferHooksTC
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Mar 2018 14:07:23 +0100] rev 3669
[test] Give a more specific name to test method in SEDAArchiveTransferHooksTC We're going to add another method, a more generic will help distinguish their respective purpose.
Wed, 21 Mar 2018 11:44:58 +0100 [seda] Insert an empty TransferringAgencyArchiveIdentifier element in rng 0.2 and XSD exports
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Mar 2018 11:44:58 +0100] rev 3666
[seda] Insert an empty TransferringAgencyArchiveIdentifier element in rng 0.2 and XSD exports We insert an empty tag named TransferringAgencyArchiveIdentifier in RNG 0.2 and XSD exports of SEDA profiles as long as they have an archive unit. The tag has a fixed "value" and default values for other attributes. This is meant for downstream consumption by GED-SAS and Asalae which would insert an actual identifier in this element. We implement this as a monkeypatch just like any other customization of SEDA export mechanism in SAEM-Ref (apparently, this is to avoid duplication induced by having to implement twice the same logic for SEDA 1 and SEDA 0.2). (CONSEJIRA-445)
Wed, 21 Mar 2018 11:28:16 +0100 [seda] Hide seda_transferring_agency_archive_unit_identifier from archive unit form/view
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Mar 2018 11:28:16 +0100] rev 3665
[seda] Hide seda_transferring_agency_archive_unit_identifier from archive unit form/view We want the value of "identifiant pour le service versant" to be hidden from archive unit view and form so that it can later automatically inserted with an empty value in SEDA exports (RNG 0.2 and XSD). (CONSEJIRA-445)
Mon, 19 Mar 2018 09:44:34 +0100 Drop redundant uifcg rule about "simplified_profile"
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 19 Mar 2018 09:44:34 +0100] rev 3664
Drop redundant uifcg rule about "simplified_profile" The same rule appears two lines below.
Thu, 15 Mar 2018 11:31:53 +0100 [doc] Explain ARK identifiers structure and generation mechanism
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 15 Mar 2018 11:31:53 +0100] rev 3663
[doc] Explain ARK identifiers structure and generation mechanism
Thu, 15 Mar 2018 10:54:20 +0100 [doc] Explain how component entities have their ARK generated
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 15 Mar 2018 10:54:20 +0100] rev 3662
[doc] Explain how component entities have their ARK generated
Thu, 15 Mar 2018 10:53:18 +0100 [doc] Rewrap last part of ark documentation to 80 characters
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 15 Mar 2018 10:53:18 +0100] rev 3660
[doc] Rewrap last part of ark documentation to 80 characters
Thu, 15 Mar 2018 10:37:49 +0100 [doc] Update ARK identifiers format doc to reflect recent changes
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 15 Mar 2018 10:37:49 +0100] rev 3659
[doc] Update ARK identifiers format doc to reflect recent changes We only drop no-longer accurate information. More changes will come in next changesets.
Mon, 26 Feb 2018 11:20:43 +0100 i18n update
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 26 Feb 2018 11:20:43 +0100] rev 3658
i18n update
Mon, 26 Feb 2018 11:08:10 +0100 Added tag 0.18.0 for changeset 9d5ff318ba97
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 26 Feb 2018 11:08:10 +0100] rev 3657
Added tag 0.18.0 for changeset 9d5ff318ba97
Mon, 26 Feb 2018 11:07:44 +0100 [pkg] Version 0.18.0 0.18.0
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 26 Feb 2018 11:07:44 +0100] rev 3656
[pkg] Version 0.18.0
Mon, 26 Feb 2018 10:12:12 +0100 [pkg] Require cubicweb >= 3.26.1
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 26 Feb 2018 10:12:12 +0100] rev 3655
[pkg] Require cubicweb >= 3.26.1
Mon, 26 Feb 2018 09:49:48 +0100 [ark] Avoid infinite loops by settings a maximum number of iterations limit
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 26 Feb 2018 09:49:48 +0100] rev 3654
[ark] Avoid infinite loops by settings a maximum number of iterations limit Arbitrarily set this limit to 20, without further usage knowledge. We don't catch the error on application side for now, waiting for an actual use case.
Mon, 26 Feb 2018 10:29:28 +0100 [ark] Move creation of "ark" table in precreate script
Denis Laxalde <denis.laxalde@logilab.fr> [Mon, 26 Feb 2018 10:29:28 +0100] rev 3653
[ark] Move creation of "ark" table in precreate script SQL scripts in schema directory get executed before tables being created so this seems better and thus it might get hard to handle migration for them. Thus we only let functions in the "ark" script and move table creation in precreate migration script. We cannot do this in postcreate because cubicweb-seda's postcreate script will get executed before saem_ref's one and it will trigger SEDA schemes import, which needs the "ark" to exist.
Thu, 22 Feb 2018 17:05:43 +0100 [pkg] Version 0.18.0a2
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 17:05:43 +0100] rev 3650
[pkg] Version 0.18.0a2
Thu, 22 Feb 2018 17:05:08 +0100 Create ARK NAA for ADGIRONDE before importing SEDA schemes
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 17:05:08 +0100] rev 3649
Create ARK NAA for ADGIRONDE before importing SEDA schemes We cannot call generate_ark() as this will insert a record in "ark" table and which will conflict with the one that'll get generated by hooks.
Thu, 22 Feb 2018 16:33:07 +0100 [pkg] Version 0.18.0a1
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 16:33:07 +0100] rev 3648
[pkg] Version 0.18.0a1
Wed, 21 Feb 2018 17:58:23 +0100 [test] Check that we create qualified ARK identifiers in views tests
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Feb 2018 17:58:23 +0100] rev 3647
[test] Check that we create qualified ARK identifiers in views tests Currently, this fails for test_ark_concept_creation_linkto (because of __linkto, I guess) but works for other ones.
Thu, 22 Feb 2018 16:00:23 +0100 [ark] Configure Concept autoform so that ARK identifier generation works
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 16:00:23 +0100] rev 3646
[ark] Configure Concept autoform so that ARK identifier generation works We need "in_scheme" relation to appear in entity.cw_edited when going through ARK generation hook. Hence we tag the relation as being displayed in "attributes" section. However, as there's no point in editing the relation because a concept is only created related to a concept scheme, we use a hidden widget to make hidden to the end user (but still present in posted form).
Thu, 22 Feb 2018 15:30:01 +0100 [ark] Get rid of usage of transaction data to handle in_scheme relation
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 15:30:01 +0100] rev 3645
[ark] Get rid of usage of transaction data to handle in_scheme relation This mechanism (a hack) got introduced in 8d3da9757a5d (which we partly revert here) to handle creation of ARK identifiers for Concept entities being created but not yet related to a ConceptScheme (through in_scheme relation). This is no longer need since 7b0c7dc35542, in which we make "in_scheme" relation inlined. In particular, we get rid of editcontroller hack and thus drop respective test.
Thu, 22 Feb 2018 15:03:01 +0100 Drop hack around cubicweb-seda's dataimport.EXTID2EID_CACHE
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 15:03:01 +0100] rev 3644
Drop hack around cubicweb-seda's dataimport.EXTID2EID_CACHE In cubicweb-seda 0.13.2 (which itself relies on a recent cubicweb-skos), we dropped a hack about dataimport (see changeset 960c8c66cfc2 in cubicweb-seda); this lead to the disappearance of EXTID2EID_CACHE global in dataimport module. Accordingly, we need to drop the its usage from saem_ref.
Thu, 22 Feb 2018 15:46:41 +0100 Let "ark" value generator return unicode string
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 15:46:41 +0100] rev 3643
Let "ark" value generator return unicode string Followup on ce883b461791.
Thu, 22 Feb 2018 16:19:55 +0100 [ark] Insert "external" ARK identifiers into ark table
Denis Laxalde <denis.laxalde@logilab.fr> [Thu, 22 Feb 2018 16:19:55 +0100] rev 3642
[ark] Insert "external" ARK identifiers into ark table When calling set_ark_and_cwuri() (from a hook upon entity addition or from a dataimport metadata generator), we now insert a record into "ark" table when an "ark" value is specified in entity.cw_edited (meaning the ARK identifier is externally given). Afterwards, the value returned by insert_ark() is used to update cw_edited dict to ensure consistency with parsed value (in particular, users may specify the "ark:/" scheme, which we do not store). When an external ARK identifier is given, we only check it is correctly formatted but don't check for our internal constraints (presence of a prefix and control character, etc.). By inserting external ARK identifiers in this "ark" table, we can now ensure uniqueness of values across entity types, thus solving a major integrity bug. Notice the change in test_ccplugin.py, where an external ark "000/123" was specified for inserted concept scheme; this ark is now transformed into "0/123" because postgres "ark" table has an integer column for "naan". Henceforth, we use the *real* ark when calling ctl import command.
Wed, 21 Feb 2018 16:40:28 +0100 [ark] Add an insert_ark helper function
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Feb 2018 16:40:28 +0100] rev 3633
[ark] Add an insert_ark helper function
Wed, 21 Feb 2018 11:42:38 +0100 Insert valid "ark" attributes in all tests
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Feb 2018 11:42:38 +0100] rev 3622
Insert valid "ark" attributes in all tests See previous changeset for why.
Wed, 21 Feb 2018 11:41:39 +0100 Add a custom value generator for automatic web tests for "ark" attribute
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 21 Feb 2018 11:41:39 +0100] rev 3621
Add a custom value generator for automatic web tests for "ark" attribute We will soon require that inserted "ark" attribute match a regexp pattern. In test_saem_ref.py's automatic web tests, ark attribute values are automatically generated and will most of the times not match this regexp. So we need to implement a custom "value generator" for "ark" attribute so that it matches expected regular expression. (How does this work? Apparently, the metaclass of ValueGenerator make is automagically used by AutomaticWebTest subclasses defined in the same module.)
(0) -3000 tip