Thanks to Elias Myllymäki for the report, and Shai Berger and Jake
Howard for the reviews.
Co-authored-by: Natalia <124304+nessita@users.noreply.github.com>
Backport of 9f3419b519799d69f2aba70b9d25abe2e70d03e0 from main.
Thanks to Baptiste Mispelon for the report and quick fix, and to Simon
Charette and Jacob Walls for the reviews.
Co-authored-by: Natalia <124304+nessita@users.noreply.github.com>
Backport of 66f9eb0ff1e7147406318c5ba609729678e4e6f6 from main.
Previously, any first field of a composite primary key with type
`INTEGER` was incorrectly introspected as an `AutoField` due to SQLite
treating `INTEGER PRIMARY KEY` as an alias for the `ROWID`.
This change ensures that integer fields in composite PKs are not
mistaken for auto-incrementing fields.
Thanks Jacob Walls and Sarah Boyce for the reviews.
Backport of 07100db6f46255ec6ef70b860495f977473684d6 from main.
Now that Django properly supports creating models with composite primary
keys, the tests should use a `CompositePrimaryKey` field instead of a
feature flag to inline backend specific SQL for creating a composite PK.
Specifcially, the inspectdb's test_composite_primary_key was adjusted to
use schema editor instead of per-backend raw SQL.
Backport of 4c75858135589f3a00e32eb4d476074536371a32 from main.
The issue was only manifesting itself when also filtering againt a related
model as that forces the usage of a subquery because SQLUpdateCompiler doesn't
support the UPDATE FROM syntax yet.
Regression in 65ad4ade74dc9208b9d686a451cd6045df0c9c3a.
Refs #28900.
Thanks Gav O'Connor for the detailed report.
Backport of 8ef4e0bd423ac3764004c73c3d1098e7a51a2945 from main.
Regression in aba0e541caaa086f183197eaaca0ac20a730bbe4 and in
d5bebc1c26d4c0ec9eaa057aefc5b38649c0ba3b.
Thanks Florent Messa for the report, and Jake Howard and Claude
Paroz for the review.
Backport of 0596263c3136bc26cffa670e5322bd0aa56c4d34 from main.
Regression in 55d89e25f4115c5674cdd9b9bcba2bb2bb6d820b.
This work improves the django.utils.text.wrap() function to ensure that
empty lines and lines with whitespace only are kept instead of being
dropped.
Thanks Matti Pohjanvirta for the report and fix.
Co-authored-by: Natalia <124304+nessita@users.noreply.github.com>
Backport of 1e9db35836d42a3c72f3d1015c2f302eb6fee046 from main.
This reverts commits b1324a680add78de24c763911d0eefa19b9263bc and
02a5cbfe76382da2a0414df17017185be5bd47f9. The former caused a regression
in admin sites that relied on the `object-tools` block being inside the
`content` block.
Thank you to Fabian Braun for the report.
Backport of 1bc805e23b73a580b82a1d416ab0fb59a1073047 from main.
Regression in ec7d69035a408b357f1803ca05a7c991cc358cfa.
Thank you Gabriel Trouvé for the report and Claude Paroz for the review.
Backport of d469db978ea6a705549b9519313d9adc198e4232 from main.
Now that selected aliases are stored in sql.Query.selected: dict[str, Any]
the values_list() method must ensures that duplicate field name references are
assigned unique aliases.
Refs #28900.
Regression in 65ad4ade74dc9208b9d686a451cd6045df0c9c3a.
Thanks Claude for the report.
Backport of 21f8be76d43aa1ee5ae41c1e0a428cfea1f231c1 from main.
Regression in 65ad4ade74dc9208b9d686a451cd6045df0c9c3a which allowed for
annotations to be SELECT'ed before model field references through
values()/values_list() and broke assumptions the select_for_update(of)
table infererence logic had about model fields always being first.
Refs #28900.
Thanks OutOfFocus4 for the report and Sarah for the test.
Backport of 71a19a0e475165dbc14c1fe02f552013ee670e4c from main
Regression in 58cd4902a71a3695dd6c21dc957f59c333db364c.
Thanks Baptiste Mispelon for the report.
Backport of 8ad3e80e88201f4c557f6fa79fcfc0f8a0961830 from main.
Regression in 65ad4ade74dc9208b9d686a451cd6045df0c9c3a.
Refs #28900.
Thanks Jeff Iadarola for the report and tests.
Co-Authored-By: OutOfFocus4 <jeff.iadarola@gmail.com>
Backport of 12b771a1ec4bbfe82405176f5601e6441855a303 from main
Swapped to an allow list instead of a deny list for field types to
determine if the UNNEST optimization can be enabled to avoid further
surprises with other types that would require further specialization to
adapt.
Regression in a16eedcf9c69d8a11d94cac1811018c5b996d491.
Thanks Joshua Goodwin for the report and Sarah Boyce for the test.
Backport of 764af7a3d6c0b543dcf659a2c327f214da768fe4 from main
Just like the In() lookup discards of None members TupleIn() should
discard tuples containing any None as NULL != NULL in SQL and the
framework expects such queries to be elided under some circumstances.
Refs #31667, #36116.
Thanks Basptise Mispelon for bisecting the regression to 626d77e.
Backport of f7f38f3a0b44d8c6d14344dae66b6ce52cd77b55 from main
Regression in 65ad4ade74dc9208b9d686a451cd6045df0c9c3a.
Refs #28900
Thanks Patrick Altman for the report.
Backport of 543e17c4405dfdac4f18759fc78b190406d14239 from main
Thank you Adam Johnson for the report and fix. Thank you Sarah Boyce for
your spot on analysis.
Regression in c09bceef68e5abb79accedd12dade16aa6577a09, which is
partially reverted in this branch.
Co-authored-by: Sarah Boyce <42296566+sarahboyce@users.noreply.github.com>
Backport of 27b68bcadf1ab2e9f7fd223aed42db352ccdc62d from main.
Thank you Raffaella for the report. Thank you Tim Schilling and Natalia Bidart
for the reviews.
Backport of de1117ea8eabe0ee0aa048e5a4e249eab7c4245e from main.
Regression in 40b3975e7d3e1464a733c69171ad7d38f8814280.
Thanks smiling-watermelon for the report.
Co-authored-by: Sarah Boyce <42296566+sarahboyce@users.noreply.github.com>
Backport of c09bceef68e5abb79accedd12dade16aa6577a09 from main.
Avoids reports of bulk_update() sending Cast expressions
to JSONField.get_prep_value().
Co-authored-by: Simon Charette <charette.s@gmail.com>
Backport of 0bf412111be686b6b23e00863f5d449d63557dbf from main.
Regression in 66e47ac69a7e71cf32eee312d05668d8f1ba24bb.
Thanks mfontana-elem for the report and Sarah for the tests.
Backport of c3a23aa02faa1cf1d32e43d66858e793cd9ecac4 from main.
While the top-level objects fed to serialization are bound to the test
database being created nothing prevents code invoked during
serialization from performing queries against other connections entries
that haven't been swapped yet.
The reported example of that is a database router directing all reads to
a test mirror for a set of models involving auto-created many-to-many
fields. It might be tempting to address the many-to-many field case but
this a symptom of a larger problem where the test framework yields the
flow execution to user code that could interact with non-test databases
in unexpected ways.
Deferring test database fixture serialization until the point where all
connections entries have been swapped for their test equivalent ensures
that no code triggered during serialization can interact with non-test
databases.
Thanks Jake Howard for the report and Jacob Walls for the initial
investigation.
Backport of dc69a63f844b2ef3bc3371edde91644cf0bef0ee from main
When Expression.__init__() overrides make use of *args, **kwargs
captures their argument values are respectively bound as a tuple and
dict instances. These composite values might themselves contain values
that require special identity treatments such as Concat(output_field)
as it's a Field instance.
Refs #30628 which introduced bound Field differentiation but lacked
argument captures handling.
Thanks erchenstein for the report.
Backport of df2c4952df6d93c575fb8a3c853dc9d4c2449f36 from main
Non-tuple exact and in lookups have specialized logic for subqueries that can
be adapted to properly assign select mask if unspecified and ensure the number
of involved members are matching on both side of the operator.
Backport of 41239fe34d64e801212dccaa4585e4802d0fac68 from main.
In order for Expression.relabeled_clone to work appropriately its
get_source_expressions method must return all resolvable which wasn't the case
for Lookup when its right-hand-side is "direct" (not a compilable).
While refs #22288 added support for non-literals iterable right-hand-side
lookups it predated the subclassing of Lookup(Expression) refs #27021 which
could have been an opportunity to ensure right-hand-sides are always resolvable
(ValueList and ExpressionList).
Addressing all edge case with non-resolvable right-hand-sides would require
a significant refactor and deprecation of some parts of the Lookup interface so
this patch only focuses on FieldGetDbPrepValueIterableMixin (In and Range
lookups) by making sure that a right-hand-side containing resolvables are dealt
with appropriately during the resolving phase.
Thanks Aashay Amballi for the report.
Backport of 089deb82b9ac2d002af36fd36f288368cdac4b53 from main.
Added through_fields support to ManyToManyField.deconstruct.
Thanks to Simon Charette for pointers and the review.
Backport of b13b8684a04d0bc1081104c5973c62c27dc673b0 from main.
The original queryset._next_is_sticky() call never had the intended effect as
no further filtering was applied internally after the pk__in lookup making it
a noop.
In order to be coherent with how related filters are applied when retrieving
objects from a related manager the effects of what calling _next_is_sticky()
prior to applying annotations and filters to the queryset provided for
prefetching are emulated by allowing the reuse of all pre-existing JOINs.
Thanks David Glenck and Thiago Bellini Ribeiro for the detailed reports and
tests.
Backport of 2598b371a93e21d84b7a2a99b2329535c8c0c138 from main.
The get_(local|foreign)_related_value methods of GenericRelation must be
reversed because it defines (from|to)_fields and associated related_fields
in the reversed order as it's effectively a reverse GenericForeignKey
itself.
The related value methods must also account for the fact that referenced
primary key values might be stored as a string on the model defining the
GenericForeignKey but as integer on the model defining the GenericRelation.
This is achieved by calling the to_python method of the involved content type
in get_foreign_related_value just like GenericRelatedObjectManager does.
Lastly reverse many-to-one manager's prefetch_related_querysets should use
set_cached_value instead of direct attribute assignment as direct assignment
might are disallowed on ReverseManyToOneDescriptor descriptors. This is likely
something that was missed in f5233dc (refs #32511) when the is_cached guard
was added.
Thanks 1xinghuan for the report.
Backport of 198b30168d4e94af42e0dc7967bd3259b5c5790b from main.
Regression in e626716c28b6286f8cf0f8174077f3d2244f3eb3.
Thanks buffgecko12 for the report and Sarah Boyce for the review.
Backport of d15454a6e84a595ffc8dc1b926282f484f782a8f from main.
