The failure was introduced in dd1ea70779adff294140eddb0229c382e12f32f3.
The tests wouldn't start due to a "Dependency on unknown app" error.
Backport of b8cc5f3cd4c70518dceb894edf9717abff79f0b4 from master
Creating a new cluster is neither required not recommendable for
most users. The previous section explains how to create a user
with sufficient permissions to automatically create a database
during tests.
Previous algo only worked if the first item was a part of the loop,
and you would get an infinite loop otherwise (see test).
To fix this, it was much easier to do a pre-pass.
A bonus is that you now get an error message that actually helps you debug
the dependency problem.
Backport of ff3d746e8d8e8fbe6de287bd0f4c3a9fa23c18dc from master, with
additional tests from c5def493d0993d65bf7d96f0a204006cbeaa6178
This change allows dynamically created inlines "Add related" button to work
correcly as long as their associated foreign key is pointing to the primary
key of the related model.
Thanks to amorce for the report, Julien Phalip for the initial patch,
and Collin Anderson for the review.
Backport of f9c4e14aeca7df79991bca8ac2d743953cbd095c from master
This rename is very tricky for the case of subclasses which define
get_query_set and haven't been updated yet, which applies to all projects in
the form of RelatedManager from Django 1.5.
Backport of 0c623da66406d1f20b5e26d497d57da5ad0de066 from master
The Preventing header injection example included
classes that are not imported.
Thanks to Collin Anderson and Berker Peksağ for the
reviews.
Backport of 0d74209ef66111f29f122d5f5fdf8e2964890cc5 from master.
This reverts commit 8014001d9287d516c58be80ad71fb63593648b3d.
Adding kwargs to deconstructed objects does not achieve useful
forward-compatibility in general, since the additional kwargs are silently
dropped rather than having their expected effect. In fact, it can cause the
failure to be more difficult to debug. Thanks Shai Berger for discussion.
Also added a unit test wit the simpler syntax which we have documented,
where the dictionary values are strings.
Backport of 5b26a014a81ba0d404d46e11d2b45c01d92b97e5 from master
Added relabeled_clone() method to sql.Query to fix the problem. It
manifested itself in rare cases where at least double nested subquery's
filter condition might target non-existing alias.
Thanks to Trac alias ris for reporting the problem.
Backport of 5c481db29572a387651681b43d5d4523f96b3793 from master
Removed inaccurate info about partial evaluation after refs #18702.
Added information on modifying sliced QuerySets; refs #22503.
Backport of 327df551e89a505c5756becee97c40198f38aff2 from master
Documentation for rendering form fields manually is now updated to use fields id_for_label instead of hardcoded values with additional mention of label_tag for alternative generation of complete label tag.
Backport of 48515ee1f7bec437544a0a958332483298f4028b from master
The example for django.contrib.admin.ModelAdmin.get_form modifies
self.exclude. However, since ModelAdmin instances are global and have no
thread- or request-locality, this is not safe for concurrent
requests.[1] This updated documentation demonstrates a safe method to
override admin forms on a per-request basis.
[1] https://groups.google.com/forum/#!topic/django-users/AmoUDtEefyA
Backport of 0d1a9d203a970a82a2f81edf0ba7d4b55442fd78 from master
