Commit Graph

2294 Commits

Author SHA1 Message Date
Aymeric Augustin f574220f09 Fixed #22508 -- Avoided overwriting select_related.
Previously, known related objects overwrote related objects loaded
though select_related. This could cancel the effect of select_related
when it was used over more than one level.

Thanks boxm for the bug report and timo for bisecting the regression.
2014-05-10 16:39:20 +02:00
Aymeric Augustin fb90b7cda2 Adjusted refactoring of vendor checks.
Thanks Shai for the thorough review.
2014-05-10 14:40:42 +02:00
Andrew Godwin 5400b29ebf Fixed #22325: Ignore __first__ dependencies to your own app 2014-05-08 21:38:56 -07:00
Andrew Godwin 5182efce8d Fixed #22563: Ignore AUTH_USER_MODEL errors in from_state 2014-05-08 21:34:30 -07:00
Aymeric Augustin 99d9fa329a Added feature flags for introspection capabilities. 2014-05-08 22:11:20 +02:00
Aymeric Augustin cff59bedc2 Split ignores_nulls_in_unique_constraints feature.
Oracle and SQL Server don't have exactly the same limitations. It's
worth treating them differently.
2014-05-08 22:11:20 +02:00
Aymeric Augustin c70a61eb49 Replaced vendor checks by three feature flags. 2014-05-08 22:11:15 +02:00
Aymeric Augustin b1432bfc22 Appeased flake8. 2014-05-08 21:49:54 +02:00
Andrew Godwin b25aee3b7b Fixed #22476: Couldn't alter attributes on M2Ms with through= set 2014-05-08 10:34:24 -07:00
Andrew Godwin 5a917cfef3 Fixed #22496: Data migrations get transactions again! 2014-05-07 14:28:34 -07:00
Andrew Godwin e9a456d11b Fixed #22581: Pass default values for schema through get_db_prep_save() 2014-05-07 13:46:23 -07:00
Andrew Godwin fc974313b8 Fixed #22563: Better error message when trying to change AUTH_USER_MODEL
You're not allowed to do this after you've made migrations; see ticket
for more details.
2014-05-07 13:05:11 -07:00
Andrew Godwin f9d7e18dc5 Fixed #22576: Ensure makemigrations doesn't touch the database. 2014-05-06 22:41:59 -07:00
Claude Paroz da9cf53cb5 Fixed #22564 -- Prevented unneeded bytestrings in migrations
In some cases, this could lead to migrations written with Python 2
being incompatible with Python 3.
Thanks Tim Graham for the report and Loïc Bistuer for the advices.
2014-05-06 09:13:07 +02:00
Claude Paroz 12474dacef Added a bunch of missing unicode_literals
Refs #22564.
2014-05-06 09:13:07 +02:00
Anssi Kääriäinen 3b7c66a3ac Fixed #22466 -- ordering by reverse foreign key
Ordering by reverse foreign key was broken by custom lookups patch
(commit 20bab2cf9d).

Thanks to everybody who helped solving this issue. Special thanks to
Trac alias takis for reporting this.
2014-05-05 15:28:07 +03:00
Jakub Roztocil 7131e14d00 Fixed #22489 -- missing implemenation for search lookup
When custom lookups were added, converting the search lookup to use
the new Lookup infrastructure wasn't done.

Some changes were needed to the added test, main change done by
committer was ensuring the test works on MySQL versions prior to 5.6.
2014-05-05 14:36:15 +03:00
Anssi Kääriäinen 5e1f4656b9 Fixed #22429 -- Incorrect SQL when using ~Q and F 2014-05-05 13:02:11 +03:00
Claude Paroz faa8c71fd9 Moved misplaced transaction.atomic from 0dce44e16b
Thanks Florian Apolloner for noticing the issue and Aymeric
Augustin for the expertise.
Refs #22540.
2014-05-03 23:44:40 +02:00
Tim Graham 3818d96426 Fixed #22435 -- Prevented adding a ManyToManyField from prompting for a default.
Thanks andrewsg for the report.
2014-05-02 20:46:47 -04:00
Claude Paroz 0dce44e16b Fixed #22540 -- Made _create_object_from_params() call create()
So as the save step is centralized in create(), especially useful
when customizing behavior in subclasses.
Thanks craig.labenz@gmail.com for the report.
2014-05-02 09:06:15 +02:00
Chris Beaven a0c4b8465d Fix migration autodector to work correctly with custom deconstructed values 2014-05-02 15:08:19 +12:00
Yehonatan Daniv e2e4cdba11 Fixed #22539 -- Copied exclude argument in Model.full_clean() to prevent side effects. 2014-05-01 09:28:11 -04:00
Tim Graham 45c2d1f5d9 flake8 fixes. 2014-04-30 18:33:46 -04:00
Andrew Godwin 8f6dff372b Fixed #22485: Include all unmigrated apps in project state by default. 2014-04-30 12:25:12 -07:00
Simon Charette 24ec9538b7 Fixed #19195 -- Allow explicit ordering by a relation `_id` field.
Thanks to chrisedgemon for the report and shaib, akaariai and
timgraham for the review.
2014-04-30 14:23:19 -04:00
Claude Paroz 7c54f8cced Fixed #22474 -- Made migration recorder aware of multiple databases
Thanks Tim Graham for the review.
2014-04-30 16:53:20 +02:00
Simon Charette a2340ac6d6 Use the new implementation of `six.with_metaclass`.
No more `NewBase` horrors.

Thanks to bendavis78 for his work on merging this into six.
2014-04-29 09:43:08 -04:00
Alex Gaynor 2bcb8bfc8d Fix many many typos in comments throughout the codebase 2014-04-26 10:18:45 -07:00
Aymeric Augustin 8b5b199e20 Fixed #3214 -- Stopped parsing SQL with regex.
Avoided introducing a new regex-based SQL splitter in the migrations
framework, before we're bound by backwards compatibility.

Adapted this change to the legacy "initial SQL data" feature, even
though it's already deprecated, in order to facilitate the transition
to migrations.

sqlparse becomes mandatory for RunSQL on some databases (all but
PostgreSQL). There's no API to provide a single statement and tell
Django not to attempt splitting. Since we have a more robust splitting
implementation, that seems like a good tradeoff. It's easier to add a
new keyword argument later if necessary than to remove one.

Many people contributed to both tickets, thank you all, and especially
Claude for the review.

Refs #22401.
2014-04-26 17:46:23 +02:00
Aymeric Augustin 3033a7193a Fixed #21166 -- Reset errors_occurred flag after commit and rollback. 2014-04-25 22:45:11 +02:00
Aymeric Augustin e368912902 Set some transaction-related feature flags on SQLite.
Refs #22496.
2014-04-25 11:43:20 +02:00
Shai Berger 843613add4 Fixed #22498 -- constraint name was not quoted in FK creation SQL 2014-04-25 01:30:43 +03:00
Shai Berger 53d97e4fe3 Made sure cursor.close() does not complain if cursor is already closed on Oracle
Refs #22483
2014-04-25 00:36:40 +03:00
Aymeric Augustin 25209715d4 Prevented a crash in the cursor wrappers on Oracle.
Fixed #22483 (again).

Forwardport of 5cd6429620 from stable/1.7.x
2014-04-24 09:14:42 -04:00
Aymeric Augustin b0f4eecfa7 Ignored repeated calls to connection.close().
Forwardport of 9bbb43dd1a from stable/1.7.x
2014-04-24 09:14:26 -04:00
Aymeric Augustin 0aa4c6c391 Used the same instance of atomic for entry and exit.
Since all state is maintained on the connection at this time and none in
the atomic, it doesn't matter, but it could introduce some subtle bugs
if the implementation changed in the future.
2014-04-23 21:44:36 +02:00
Aymeric Augustin e74d2183c2 Wrapped migrations in a transaction only on DBs with transactional DDL. 2014-04-23 21:44:36 +02:00
Erik Romijn 75c0d4ea3a Fixed queries that may return unexpected results on MySQL due to typecasting.
This is a security fix; disclosure to follow shortly.
2014-04-21 18:11:26 -04:00
Aymeric Augustin 428c0bbe1b Appeased flake8 2.1.0. 2014-04-21 12:27:34 +02:00
Tim Graham 471fb04a30 Fixed flake8 errors. 2014-04-20 13:08:04 -04:00
Andrew Gorcester 956bd64424 Fixed #22397 -- Issues removing M2M field with explicit through model
Changed the migration autodetector to remove models last so that FK
and M2M fields will not be left as dangling references. Added a check
in the migration state renderer to error out in the presence of
dangling references instead of leaving them as strings. Fixed a bug
in the sqlite backend to handle the deletion of M2M fields with
"through" models properly (i.e., do nothing successfully).

Thanks to melinath for report, loic for tests and andrewgodwin and
charettes for assistance with architecture.
2014-04-18 16:44:47 -04:00
Simon Charette 0d397e5a5b Revert "Fixed #22397 -- Issues removing M2M field with explicit through model."
This reverts commit 00e3b9a2a9.

It's causing a regression when tested with the proxy_model_inheritance tests.
2014-04-18 01:27:30 -04:00
Andrew Gorcester 00e3b9a2a9 Fixed #22397 -- Issues removing M2M field with explicit through model.
Changed the migration autodetector to remove models last so that FK
and M2M fields will not be left as dangling references. Added a check
in the migration state renderer to error out in the presence of
dangling references instead of leaving them as strings. Fixed a bug
in the sqlite backend to handle the deletion of M2M fields with
"through" models properly (i.e., do nothing successfully).

Thanks to melinath for report, loic for tests and andrewgodwin and
charettes for assistance with architecture.
2014-04-17 12:54:35 -04:00
Justin Hamade a13df671a5 Fixed #22434 -- Retain ordering on related sliced subqueries.
Thanks maciej.pawlisz for the report, and charettes for the review.
2014-04-16 17:57:11 -04:00
Simon Charette 79f05616fb Fixed flake8 warnings introduced in recent commits. 2014-04-16 16:49:37 -04:00
Motiejus Jakštys 0e45669fa9 Fixed #22460 -- Explicity remove constraints when dropping a related field. 2014-04-16 15:38:31 -04:00
Marc Egli 0bcc92c691 Fixed #22356 -- Added a check to make sure unique_together fields are local. 2014-04-14 16:49:40 -04:00
Simon Charette 72d3889db4 Fixed #22350 -- Consistently serialize bytes and text in migrations.
Thanks to @treyhunner and Loïc for their suggestions and review.
2014-04-13 18:22:26 -04:00
valtron d3b71b976d Fixed #21760 -- prefetch_related used an inefficient query for reverse FK.
Regression introduced by commit 9777442. Refs #21410.
2014-04-13 00:34:16 +07:00