| Age | Commit message (Collapse) | Author |
|
Full disclosure and new release forthcoming.
|
|
malicious users.
|
|
Some backends might need special handling for auto-increment values.
This was introduced in 5fa5621f574. Thanks Michael Manfre for spotting
the issue.
Backport of 0cdfa76e68 from master.
|
|
Thanks dloewenherz for the report.
Backport of 4951932 from master.
|
|
Also fixes #9045.
Backport of 8248d1402 from master.
|
|
formfield_overrides
Thanks joebuyer at manycycles.com for the report and Simon Charette
for the review.
Backport of 04e6542b5 from master.
|
|
Thanks kujiu for the report and Aymeric Augustin for the review.
Backport of c91667338 from master.
|
|
is_valid
Thanks Simon Charette for the report and the initial patch.
Backport of 66dfcc10b from master.
|
|
This was not working properly when ContentFile was initialized with
an unicode string.
Thanks Alexey Boriskin for the report and the test.
Backport of 34dcf51e06 from master.
|
|
Applied to inlines what ec9d6b1122d did for main fieldsets.
Backport of 553838a28 from master.
|
|
templates consistent.
Thanks and at cloverfastfood dot com for the report.
b64d30405a3d5468dc8c6232747d45bbeee4f7bb from master.
|
|
The failure was caused by using None as a choice for a CharField. To
avoid Oracle's "" <-> NULL handling the field type was changed to
IntegerField.
Backpatch of 632cf32b43609f7d674c81f81f01eb9e4a059f1f
|
|
Backport of c196e01100b2
|
|
Backport of 795ac7deda from master.
|
|
can be displayed as selected even if the lookup's first element is not a string.
Backport of 88e17156393b
|
|
Thanks Aymeric Augustin for detecting this regression.
Backport of 5fa5621f5 from master.
|
|
Thanks gw 2012 at tnode com for the report.
Backport of baae4b8.
|
|
Backport of 349c4c37f8 from master.
|
|
Thanks markteisman at hotmail.com for the report.
Backport of 0eeae1505 from master.
|
|
|
|
Reverse o2o fields are now usable with defer.
Backpatch of [6ebf115206289bce8f3d86318871faac13d6e835]
|
|
Backport of c10aaa70a from master.
|
|
Thanks m3wolf for the report and akaariai for reproducing the problem.
Backport of 2ea80b9.
|
|
are shown in change forms when the raw value is None.
Backport of 29d59a879ea5b116
|
|
Thanks to rasca for the suggestion and to mateusgondim for the patch.
Backport of ae206d78f6d991e
|
|
Backport of f26b956e80 from master.
|
|
TemplateSyntaxError is expected at compile time, not at run time.
Refs #19280.
Backport of 7644800.
|
|
Backport of f89901d from master.
|
|
Refs #18417. Also fixed some test case classes which subclassed
the wrong parent.
Backport of a5d47415f from master.
|
|
with override_settings
Backport of 9f7cefd5 from master.
|
|
syntax.
Backport of 6b8a7ce.
|
|
These tests were silently skipped.
Backport of 95edeab.
|
|
Backpatch of dc569c880143db07e01b3293d698ad8fe4a0136f
|
|
The previous pattern didn't properly addressed cases where '>'
was present inside quoted tag content.
Backport of bf1871d87 from master.
|
|
This is required to allow the test to run without error under SQLite in the CI environment.
Backport of 18d7c1e.
|
|
created for swapped models.
Thanks to rizumu for the report.
Backport of c8985a8a7317042a641e870cb75b3005cc5d67b1.
|
|
Backpatch of 0a0a0d66b316598f7c296e8bf75749a14ce3ac49
|
|
SQLite used INSERT INTO tbl SELECT %s UNION SELECT %s, the problem
was that there should have been UNION ALL instead of UNION.
Refs #19351
Backpatch of a27582484cf814554907d2d1ad077852de36963f
|
|
F() expressions reuse joins like any lookup in a .filter() call -
reuse multijoins generated in the same .filter() call else generate
new joins. Also, lookups can now reuse joins generated by F().
This change is backwards incompatible, but it is required to prevent
dict randomization from generating different queries depending on
.filter() kwarg ordering. The new way is also more consistent in how
joins are reused.
Backpatch of 90b86291d022a09031d1df397d7aaebc30e435f7
|
|
Django cannot delete more than 999 objects at a time with SQLite.
Refs #16426, #16039.
Backport of 2875b5d from master.
|
|
Thanks dstufft for the draft patch.
Backport of 0e3690d230 from master
|
|
Also reverted 8fb7a9002669fb7ba7bec7df90b465b92e1ed3c2. Refs #17055.
Backport of a026e48 from master.
|
|
Backport of ea6b95db from master.
|
|
one single representation. The previously used character (ş) has two different representations (u'\u015f' and u's\u0327'), which caused spurious failures when the tests were run across multiple platforms (e.g. on a Linux VM hosted on a Mac).
|
|
Thanks to Lowe Thiderman for the patch and tests
|
|
(cherry picked from commit f824a951776db86c04aaefc1e7c1c12ffb84c798)
|
|
Thanks niko at neagee.net for the report and Emil Stenstrom for
the patch.
Backport of e0363c688 from master.
|
|
Thanks anthony at adsorbtion.org for the report, and johannesl for
bringing the patch up-to-date.
Backport of 2a67374b5 from master.
|
|
Thanks shadow for the report, and Melevir and thiderman for the
patch.
Backport of ec9d6b112 from master.
|
|
|