Julien Cristau <julien.cristau@logilab.fr> [Mon, 18 May 2015 16:36:26 +0200] rev 10438
[devtools] make test db name for postgresql process-specific
In case we run two tests concurrently on the same cluster, better not
use the same db name for both.
Julien Cristau <julien.cristau@logilab.fr> [Mon, 18 May 2015 16:30:32 +0200] rev 10437
[devtools] extract functions to start/stop a postgresql cluster
Julien Cristau <julien.cristau@logilab.fr> [Mon, 18 May 2015 20:00:06 +0200] rev 10436
[test] don't leave NULL columns around when making an attribute required
It doesn't matter on sqlite (it doesn't do ALTER COLUMN), but when
running this test on postgresql it fails to add the 'NOT NULL'
constraint otherwise.
Rémi Cardona <remi.cardona@logilab.fr> [Tue, 23 Jun 2015 10:02:41 +0200] rev 10435
[server/session] Roll back in connection exit
This was changed in commit 8b35a898b334 "[server] remove cnxset
tracking, it is now unneeded".
When exiting a CubicWeb Connection, the database connection is put back
into the common pool. At that point anything done during the lifetime
of the CubicWeb Connection needs to be committed or rolled back.
Do a proper rollback to finish the transaction, as was happening before
the above-mentionned commit.
This showed up on testing with postgresql, when a DB schema modification
was blocked by an old connection that was still "in transaction".
Alain Leufroy <alain.leufroy@logilab.fr> [Wed, 20 May 2015 16:13:07 +0200] rev 10434
[web/views/formrenderer] do not use `cubicweb:target` attribute on form (closes #5534074)
The `cubicweb:target` is a flag indicating that the form data (with
file input) shall be posted inside an iframe - this is a well known
"ajax-like" workaround to post files with browsers that do not support
`FormData` (a.k.a IE<10).
The `cubicweb:target` was introduced when CW used the "xhtml strict"
doctype. Now that CW uses the "html5" doctype, this namespaced
attribute is no longer necessary and the iframe can be generated
directly.
Before this patch, CW inserts the `cubicweb:target` attribute in the
form DOM element (server side) and `setFormsTarget()` updates the
DOM with a new `<iframe>` element (client side).
Now, CW inserts the `<iframe>` DOM element directly (server side),
making `setFormsTarget` useless.
Adrien Di Mascio <Adrien.DiMascio@logilab.fr> [Fri, 25 Jul 2014 17:10:05 +0200] rev 10433
[js] lint : make it explicit that typeof is an operator, not a function.
Also, use === / !== except for null / undefined comparison
Rémi Cardona <remi.cardona@logilab.fr> [Thu, 21 May 2015 16:20:04 +0200] rev 10432
[devtools] qunit: show helpful data when an assertion fails
Related to #5533333.
Rémi Cardona <remi.cardona@logilab.fr> [Wed, 20 May 2015 17:25:40 +0200] rev 10431
[web/test] Use shorter $() instead of jQuery()
Consistency FTW.
Related to #5533333.
Rémi Cardona <remi.cardona@logilab.fr> [Wed, 20 May 2015 16:17:28 +0200] rev 10430
[devtools] qunit: use new async testing APIs
http://qunitjs.com/cookbook/#asynchronous-callbacks
QUnit keeps track of all the assert.async() objects created inside the
test functions and expects all done() functions to be called. Failure to
do so will result in the test being failed.
Unlike .start and .stop which were internal APIs, assert.async() is
stricter and fails tests if assert methods are used *after* all done()
functions are called (see "test callback execution order").
Related to #5533333.
Rémi Cardona <remi.cardona@logilab.fr> [Wed, 20 May 2015 15:45:43 +0200] rev 10429
[devtools] qunit: use new assert APIs
test functions now accept an "assert" argument which exposes QUnit's
assert methods: assert.{expect,equal,deepEqual,ok}
Related to #5533333.