Aurelien Campeas <aurelien.campeas@pythonian.fr> [Fri, 26 Jun 2015 10:26:00 +0200] rev 11428
[hooks/test/synschema] remove an unneeded skipTest.
Related to #5557633.
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 19 Jul 2016 10:47:29 +0200] rev 11427
[test] Check retcode of subprocesses in devctl tests
And display decoded stdout in case of failure.
Sylvain Thénault <sylvain.thenault@logilab.fr> [Wed, 20 Jul 2016 17:58:49 +0200] rev 11426
merge 3.23
Denis Laxalde <denis.laxalde@logilab.fr> [Fri, 08 Jul 2016 13:40:50 +0200] rev 11425
[pkg] Bump version to 3.24.0.dev0
Denis Laxalde <denis.laxalde@logilab.fr> [Wed, 29 Jun 2016 10:35:22 +0200] rev 11424
[pkg] Drop NO_SETUPTOOLS option
setuptools is the defacto standard nowadays. The only place where
NO_SETUPTOOLS option was used is during packaging (RPM, Debian) and even in
this context, it seems that setuptools is the way to go
(see https://wiki.debian.org/Python/LibraryStyleGuide).
David Douard <david.douard@logilab.fr> [Tue, 12 Jul 2016 12:21:41 +0200] rev 11423
[debian] update cubicweb-dev.lintian-overrides
cubicweb/skeleton/debian/rules.tmpl is now cubicweb/skeleton/debian/rules
prevent lintian from complaining
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Jul 2016 13:27:19 +0200] rev 11422
[pkg] Properly export data files in setup.py and adjust "newcube" test
With the new package layout (everything under "cubicweb" package), the custom
install_lib rule which makes use of include_dirs defined in __pkginfo__.py did
not prepend the package name to source directories to be copied. Fixing this.
Also, in setup.py's export() function, the destination directories' path to be
created during source tree walk was wrong.
All this makes cubicweb/skeleton directory (which is not a package) properly
installed by setup.py.
The test in cubicweb/devtools/test/unittest_devctl.py wasn't properly
implemented because it used an installation of cubicweb in "develop" mode
which shadows such packaging issues. Also it used "python -m cubicweb" instead
of directly "cubicweb-ctl" and the former appears to fall back to using the
cubicweb package *from sources* instead of the installed one.
Now that this test runs against the installed version of cubicweb, fix
MANIFEST.in to include tox.ini files (cubicweb's and skeleton's) as this is
expected from the test.
Closes #14127941.
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Jul 2016 10:44:05 +0200] rev 11421
[pkg] Handle verbose option in setup.py's export function
Make debugging easier.
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Jul 2016 10:20:33 +0200] rev 11420
[pkg] Drop unused EMPTY_FILE global variable
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Jul 2016 13:00:38 +0200] rev 11419
[tox] Do not install package in "develop"
This might shadow so packaging issues.
And user can still pass --develop to tox.
Denis Laxalde <denis.laxalde@logilab.fr> [Tue, 05 Jul 2016 13:02:14 +0200] rev 11418
[pkg] Bump version to 3.23.2
Sylvain Thénault <sylvain.thenault@logilab.fr> [Fri, 08 Jul 2016 09:59:18 +0200] rev 11417
pep8 bits
Sylvain Thénault <sylvain.thenault@logilab.fr> [Fri, 08 Jul 2016 10:17:42 +0200] rev 11416
[schema] Add a method on yams constraints to compute its unique name
used to identify it in the DB backend.
Sylvain Thénault <sylvain.thenault@logilab.fr> [Fri, 08 Jul 2016 10:17:14 +0200] rev 11415
[migration] Test and fix case of addition of an attribute with some serialized constraint
It was crashing on attempt to remove an unexisting constraint. This code may be
removed safely as it's expected to be handled by the constraint removal hook.
Notice that adding a vocabulary on shortpara makes yams change the varchar max
size. This is arguable but not the point of this test, so simply adapt the max
size defined in the schema coherently so we don't have to bother with that at
this point.