This also removes unnecessary comments with the previous spelling.
AP Stylebook has a short entry to advise the preferred spelling for
"en-us". "Afterwards" is preferred in British English.
Migrations assumed that an import of the models.Model class must
already be included when it's serialized, but for models with only
custom fields this was not necessarily the case.
Thanks Jaap Joris Vens for the report.
This disables optimization of RenameField operation when an old field
name is referenced in subsequent operations.
Co-authored-by: InvalidInterrupt <InvalidInterrupt@users.noreply.github.com>
Migrations cannot be recorded in the same transaction as its associated
DDL operations when some of it is deferred until the schema editor
context exits.
Regression in c86a3d80a2.
The migrations loader prevents the use of PEP-420 namespace packages
for holding apps' migrations modules. Previously the loader tested for
this only by checking that app.migrations.__file__ is present. This
prevented migrations' being found in frozen Python environments that
don't set __file__ on any modules. Now the loader *additionally* checks
whether app.migrations.__path__ is a list because namespace packages
use a different type for __path__. Namespace packages continue to be
forbidden, and, in fact, users of normal Python environments should
experience no change whatsoever.
When the MigrationAutodetector creates more than one initial migration
in a app, name all initial migrations "initial" rather than the opaque
"auto_<DATE>_<TIME>" name.
Initial migrations that have a descriptive name continue to use the
descriptive name.
This allows the removal of its O(n) .get_field_by_name method and many
other awkward access patterns.
While fields were initially stored in a list to preserve the initial
model definiton field ordering the auto-detector doesn't take field
ordering into account and no operations exists to reorder fields of a
model.
This makes the preservation of the field ordering completely superflous
because field reorganization after the creation of the model state
wouldn't be taken into account.
Avoids inspecting the exception message, which is not considered a
stable API and can change across Python versions.
ModuleNotFoundError was introduced in Python 3.6. It is a subclass of
ImportError that is raised when the imported module does not exist. It
is not raised for other errors that can occur during an import. This
exception instance has the property "name" which holds the name of
module that failed to import.
This moves all the field referencing resolution methods to shared
functions instead of duplicating efforts amongst state_forwards and
references methods.
Non-delayed rendering is unnecessary and wasteful now that state models
relationship consistency on delayed reload is ensured.
This partly reverts commit fcc4e251db.
Jacob Walls
Seonghyeon Cho
Carlton Gibson
Mateo Radman
Mariusz Felisiak
Manav Agarwal
Chris Jerdonek
David Smith
abhiabhi94
manav014
David Sanders
David Wobrock
Adam Johnson
Hasan Ramezani
Simon Charette
Hannes Ljungberg
William Schwartz
Koen De Wit
Iuri de Silvio
Tim Graham
François Freitag
Nick Pope
Jon Dufresne
Nan Liu