Previously, the proxy class was prepared lazily:
lazy_identity = lazy(identity, int)
lazy_identity(10) # prepared here
lazy_identity(10)
This has a slight advantage that if the lazy doesn't end up getting
used, the preparation work is skipped, however that's not very likely.
Besides this laziness, it is also inconsistent in that the methods which
are wrapped directly (__str__ etc.) are prepared already when __proxy__
is defined, and there is a weird half-initialized state.
This change it so that everything is prepared already on the first line
of the example above.
An n-ary logical XOR Q(…) ^ Q(…) ^ … ^ Q(…) should evaluate to true
when an odd number of its operands evaluate to true, not when exactly
one operand evaluates to true.
If the result type is bytes, then calling bytes() on it does nothing.
If the result type is not bytes, we should not cast to bytes, just
because the return value may be bytes.
This was problematic for screen reader users because they use headings
to navigate. Having two <h1> is confusing, and the one in the header
wasn’t particularly helpful since it’s the same on all pages.
Regression in b3db6c8dcb.
Thanks Ian Cubitt for the report.
This also corrected test_inheritance_deferred2() test which was
previously properly defined and marked as an expected failure but was
then wrongly adjusted to mask the lack of support for per-alias
deferral that was fixed by #21204.
While deferring many-to-many and GFK has no effect, the previous
implementation of QuerySet.defer() ignore them instead of crashing.
Regression in b3db6c8dcb.
Thanks Paco Martínez for the report.
Special thanks to Hannes Ljungberg for finding multiple implementation
gaps.
Thanks also to Simon Charette, Adam Johnson, and Mariusz Felisiak for
reviews.
Proper escaping of % in string literals used when defining constaints
was attempted (a8b3f96f6) by overriding quote_value of Postgres and
Oracle schema editor. The same approach was used when adding support for
constraints to the MySQL/MariaDB backend (1fc2c70).
Later on it was discovered that this approach was not appropriate and
that a preferable one was to pass params=None when executing the
constraint creation DDL to avoid any form of interpolation in the first
place (42e8cf47).
When the second patch was applied the corrective of the first were not
removed which caused % literals to be unnecessary doubled. This flew
under the radar because the existings test were crafted in a way that
consecutive %% didn't catch regressions.
This commit introduces an extra test for __exact lookups which
highlights more adequately % doubling problems but also adjust a
previous __endswith test to cover % doubling problems (%\% -> %%\%%).
Thanks Thomas Kolar for the report.
Refs #32369, #30408, #30593.
This ensures that constraint violations are tested in isolation from
each other as an IntegrityError only ensures a least one constraint is
violated.
For example, the assertion added in 42e8cf4 break both the
name_constraint_rhs and the rebate_constraint constraints and thus
doesn't constitute a proper regression test. Refs #32369.
