Mike Bayer [Mon, 5 Dec 2022 21:11:59 +0000 (16:11 -0500)]
disable polymorphic adaption in most cases
Improved a fix first made in version 1.4 for :ticket:`8456` which scaled
back the usage of internal "polymorphic adapters", that are used to render
ORM queries when the :paramref:`_orm.Mapper.with_polymorphic` parameter is
used. These adapters, which are very complex and error prone, are now used
only in those cases where an explicit user-supplied subquery is used for
:paramref:`_orm.Mapper.with_polymorphic`, which includes only the use case
of concrete inheritance mappings that use the
:func:`_orm.polymorphic_union` helper, as well as the legacy use case of
using an aliased subquery for joined inheritance mappings, which is not
needed in modern use.
For the most common case of joined inheritance mappings that use the
built-in polymorphic loading scheme, which includes those which make use of
the :paramref:`_orm.Mapper.polymorphic_load` parameter set to ``inline``,
polymorphic adapters are now no longer used. This has both a positive
performance impact on the construction of queries as well as a
substantial simplification of the internal query rendering process.
The specific issue targeted was to allow a :func:`_orm.column_property`
to refer to joined-inheritance classes within a scalar subquery, which now
works as intuitively as is feasible.
ORM context, mapper, strategies now use ORMAdapter in all cases
instead of straight ColumnAdapter; added some more parameters
to ORMAdapter to make this possible. ORMAdapter now includes a
"trace" enumeration that identifies the use path for the
adapter and can aid in debugging.
implement __slots__ for the ExternalTraversal hierarchy up
to ORMAdapter. Within this change, we have to change the
ClauseAdapter.wrap() method, which is only used in one polymorphic
codepath, to use copy.copy() instead of
`__dict__` access (apparently `__reduce_ex__` is implemented for
objects with `__slots__`), and we also remove pickling ability,
which should not be needed for adapters (this might have been needed
for 1.3 and earlier in order for Query to be picklable, but none
of that state is present within Query / select() / etc. anymore).
Mike Bayer [Wed, 7 Dec 2022 20:54:59 +0000 (15:54 -0500)]
Oracle COLUMN_VALUE is a column name, not a keyword
Fixed issue in Oracle compiler where the syntax for
:meth:`.FunctionElement.column_valued` was incorrect, rendering the name
``COLUMN_VALUE`` without qualifying the source table correctly.
Mike Bayer [Wed, 7 Dec 2022 19:23:45 +0000 (14:23 -0500)]
move more Session weakref tests to test_memusage
py311 may be more sensitive here, or maybe the machines are
acting differently these days, in any case move
memory / GC sensitive tests to test_memusage.
Mike Bayer [Tue, 6 Dec 2022 14:08:55 +0000 (09:08 -0500)]
take mariadb-connector out of dbapi main job
mariadb-connector seems to have their release situation "fixed",
but for 1.1.5post3 they are requiring mariadb-connector-c 3.3.1, Fedora 36
which is quite bleeding edge only has 3.2.7. There's no way we
can easily have a version specifier that will look at underlying
client libraries.
since mariadb-connector's main branch seems to work way ahead of
where we will normally have client libraries set up, it's not really
worth testing in CI right now.
Mike Bayer [Sun, 4 Dec 2022 17:02:37 +0000 (12:02 -0500)]
Add tests for issue #8168; slight internal adjustments
The issue in #8168 was improved, but not completely fixed,
by #8456.
This includes some small changes to ORM context that
are a prerequisite for getting ORM adaptation to be
better. Have these in 2.0.0b4 so that we have at
least a better starting point.
Lele Gaifax [Mon, 5 Dec 2022 01:51:13 +0000 (20:51 -0500)]
Add compatibility properties to Range; implement pep-484
This adds a bunch of properties to new PG Range class for compatibility
with other implementations, providing a more similar API to access
emptiness and bounds status.
The naming conventions here derive from PostgreSQL itself,
see https://www.postgresql.org/docs/9.3/functions-range.html .
pep-484 also implemented by Mike, as this is a pretty type-intensive
module.
Co-authored-by: Mike Bayer <mike_mp@zzzcomputing.com> Closes: #8927
Pull-request: https://github.com/sqlalchemy/sqlalchemy/pull/8927
Pull-request-sha: 8b9e7b7e3345673b43aeabd7ec88b88dc3cfa7eb
Changed how the positional compilation is performed. It's rendered by the compiler
the same as the pyformat compilation. The string is then processed to replace
the placeholders with the correct ones, and to obtain the correct order of the
parameters.
This vastly simplifies the computation of the order of the parameters, that in
case of nested CTE is very hard to compute correctly.
Reworked how numeric paramstyle behavers:
- added support for repeated parameter, without duplicating them like in normal
positional dialects
- implement insertmany support. This requires that the dialect supports out of
order placehoders, since all parameters that are not part of the VALUES clauses
are placed at the beginning of the parameter tuple
- support for different identifiers for a numeric parameter. It's for example
possible to use postgresql style placeholder $1, $2, etc
Added two new dialect based on sqlite to test "numeric" fully using
both :1 style and $1 style. Includes a workaround for SQLite's
not-really-correct numeric implementation.
Changed parmstyle of asyncpg dialect to use numeric, rendering with its native
$ identifiers
Gleb Kisenkov [Mon, 5 Dec 2022 13:45:25 +0000 (08:45 -0500)]
Type annotations for sqlalchemy.ext.automap
An attempt to annotate `lib/sqlalchemy/ext/automap.py` with type hints (issue [#6810](https://github.com/sqlalchemy/sqlalchemy/issues/6810#issuecomment-1127062951)).
More info on how I approach it could be found in [the earlier PR](https://github.com/sqlalchemy/sqlalchemy/pull/8775).
This pull request is:
- [ ] A documentation / typographical error fix
- Good to go, no issue or tests are needed
- [ ] A short code fix
- please include the issue number, and create an issue if none exists, which
must include a complete example of the issue. one line code fixes without an
issue and demonstration will not be accepted.
- Please include: `Fixes: #<issue number>` in the commit message
- please include tests. one line code fixes without tests will not be accepted.
- [x] A new feature implementation
- please include the issue number, and create an issue if none exists, which must
include a complete example of how the feature would look.
- Please include: `Fixes: #<issue number>` in the commit message
- please include tests.
wiseaidev [Mon, 5 Dec 2022 13:44:35 +0000 (08:44 -0500)]
Fixed an invalid syntax in an except statement
### Description
As the title suggests, I have fixed an invalid syntax in the docs for an `except` statement while reading the unusual.
### Checklist
<!-- go over following points. check them with an `x` if they do apply, (they turn into clickable checkboxes once the PR is submitted, so no need to do everything at once)
-->
This pull request is:
- [x] A documentation / typographical error fix
- Good to go, no issue or tests are needed
- [ ] A short code fix
- please include the issue number, and create an issue if none exists, which
must include a complete example of the issue. one line code fixes without an
issue and demonstration will not be accepted.
- Please include: `Fixes: #<issue number>` in the commit message
- please include tests. one line code fixes without tests will not be accepted.
- [ ] A new feature implementation
- please include the issue number, and create an issue if none exists, which must
include a complete example of how the feature would look.
- Please include: `Fixes: #<issue number>` in the commit message
- please include tests.
j00356287 [Mon, 5 Dec 2022 13:44:09 +0000 (08:44 -0500)]
doc change - Add new external dialect for openGauss
<!-- Provide a general summary of your proposed changes in the Title field above -->
### Description
<!-- Describe your changes in detail -->
Added new external dialect for [openGauss](https://www.opengauss.org/en/).
### Checklist
<!-- go over following points. check them with an `x` if they do apply, (they turn into clickable checkboxes once the PR is submitted, so no need to do everything at once)
-->
This pull request is:
- [x] A documentation / typographical error fix
- Good to go, no issue or tests are needed
- [ ] A short code fix
- please include the issue number, and create an issue if none exists, which
must include a complete example of the issue. one line code fixes without an
issue and demonstration will not be accepted.
- Please include: `Fixes: #<issue number>` in the commit message
- please include tests. one line code fixes without tests will not be accepted.
- [ ] A new feature implementation
- please include the issue number, and create an issue if none exists, which must
include a complete example of how the feature would look.
- Please include: `Fixes: #<issue number>` in the commit message
- please include tests.
Mike Bayer [Mon, 5 Dec 2022 04:25:14 +0000 (23:25 -0500)]
adjustments for unreliable gc
sporadic (and at the moment persistent) test failures
related to aiosqlite seem to have in common that Python
gc stops working fully when we run a lot of tests with
aiosqlite. The failures are not limited to aiosqlite
as they are more involving places where we assume or
expect gc.collect() to get rid of things, and it doesn't.
Identify (based on reproducible case on the d3 CI runner)
the spots where this happens and add fixes.
test/orm/test_transaction.py test_gced_delete_on_rollback
has always been a very sensitive test with a lot of issues,
so here we move it to the test_memusage suite and limit
it only to when the memusage suite is running.
Mike Bayer [Sat, 3 Dec 2022 14:11:14 +0000 (09:11 -0500)]
disallow same-named columns, unchecked replacement in Table
Fixed issue where table reflection using :paramref:`.Table.extend_existing`
would fail to deduplicate a same-named column if the existing
:class:`.Table` used a separate key. The
:paramref:`.Table.autoload_replace` parameter would allow the column to be
skipped but under no circumstances should a :class:`.Table` ever have the
same-named column twice.
Additionally, changed deprecation warnings to exceptions
as were implemented in I1d58c8ebe081079cb669e7ead60886ffc1b1a7f5 .
Mike Bayer [Fri, 2 Dec 2022 22:00:10 +0000 (17:00 -0500)]
add spaces, leading underscore to oracle checks
Expand the test suite from #8708 which unfortunately did
not exercise the bound parameter codepaths completely.
Continued fixes for Oracle fix :ticket:`8708` released in 1.4.43 where
bound parameter names that start with underscores, which are disallowed by
Oracle, were still not being properly escaped in all circumstances.
Federico Caselli [Sat, 19 Nov 2022 19:39:10 +0000 (20:39 +0100)]
Fix positional compiling bugs
Fixed a series of issues regarding positionally rendered bound parameters,
such as those used for SQLite, asyncpg, MySQL and others. Some compiled
forms would not maintain the order of parameters correctly, such as the
PostgreSQL ``regexp_replace()`` function as well as within the "nesting"
feature of the :class:`.CTE` construct first introduced in :ticket:`4123`.
Mike Bayer [Thu, 1 Dec 2022 21:03:58 +0000 (16:03 -0500)]
disable setinputsizes only for true DBAPI executemany
Fixed regression caused by the combination of :ticket:`8177`, re-enable
setinputsizes for SQL server unless fast_executemany + DBAPI executemany is
used for a statement, along with :ticket:`6047`, implement
"insertmanyvalues", which bypasses DBAPI executemany in place of a custom
DBAPI execute for INSERT statements. setinputsizes would incorrectly not be
used for a multiple parameter-set INSERT statement that used
"insertmanyvalues" if fast_executemany were turned on, as the check would
incorrectly assume this is a DBAPI executemany call. The "regression"
would then be that the "insertmanyvalues" statement format is apparently
slightly more sensitive to multiple rows that don't use the same types
for each row, so in such a case setinputsizes is especially needed.
The fix repairs the fast_executemany check so that it only disables
setinputsizes if true DBAPI executemany is to be used.
Mike Bayer [Thu, 17 Nov 2022 17:03:46 +0000 (12:03 -0500)]
add new pattern for single inh column override
Added a new parameter :paramref:`_orm.mapped_column.use_existing_column` to
accommodate the use case of a single-table inheritance mapping that uses
the pattern of more than one subclass indicating the same column to take
place on the superclass. This pattern was previously possible by using
:func:`_orm.declared_attr` in conjunction with locating the existing column
in the ``.__table__`` of the superclass, however is now updated to work
with :func:`_orm.mapped_column` as well as with pep-484 typing, in a
simple and succinct way.
Simon Schiele [Wed, 30 Nov 2022 13:40:50 +0000 (08:40 -0500)]
Add "compare" on dataclass fields
Added :paramref:`_orm.mapped_column.compare` parameter to relevant ORM
attribute constructs including :func:`_orm.mapped_column`,
:func:`_orm.relationship` etc. to provide for the Python dataclasses
``compare`` parameter on ``field()``, when using the
:ref:`orm_declarative_native_dataclasses` feature. Pull request courtesy
Simon Schiele.
Added an additional case for associationproxy into
test_dc_transforms.py -> test_attribute_options
Michael Gorven [Tue, 29 Nov 2022 23:36:19 +0000 (18:36 -0500)]
[sqlite] Reflect DEFERRABLE and INITIALLY options for foreign keys
Added support for the SQLite backend to reflect the "DEFERRABLE" and
"INITIALLY" keywords which may be present on a foreign key construct. Pull
request courtesy Michael Gorven.
Mike Bayer [Fri, 25 Nov 2022 19:29:30 +0000 (14:29 -0500)]
annotated / DC forms for association proxy
Added support for the :func:`.association_proxy` extension function to
take part within Python ``dataclasses`` configuration, when using
the native dataclasses feature described at
:ref:`orm_declarative_native_dataclasses`. Included are attribute-level
arguments including :paramref:`.association_proxy.init` and
:paramref:`.association_proxy.default_factory`.
Documentation for association proxy has also been updated to use
"Annotated Declarative Table" forms within examples, including type
annotations used for :class:`.AssocationProxy` itself.
Also modernized documentation examples in sqlalchemy.ext.mutable,
which was not up to date even for 1.4 style code.
Corrected typing for relationship(secondary) where "secondary"
accepts a callable (i.e. lambda) as well
Federico Caselli [Sun, 27 Nov 2022 17:11:34 +0000 (18:11 +0100)]
Improve support for enum in mapped classes
Add a new system by which TypeEngine objects have some
say in how the declarative type registry interprets them.
The Enum datatype is the primary target for this but it is
hoped the system may be useful for other types as well.
Lele Gaifax [Sun, 27 Nov 2022 16:28:51 +0000 (11:28 -0500)]
Add value-level hooks for SQL type detection; apply to Range
Added additional type-detection for the new PostgreSQL
:class:`_postgresql.Range` type, where previous cases that allowed the
psycopg2-native range objects to be received directly by the DBAPI without
SQLAlchemy intercepting them stopped working, as we now have our own value
object. The :class:`_postgresql.Range` object has been enhanced such that
SQLAlchemy Core detects it in otherwise ambiguous situations (such as
comparison to dates) and applies appropriate bind handlers. Pull request
courtesy Lele Gaifax.
Federico Caselli [Sun, 27 Nov 2022 17:11:34 +0000 (18:11 +0100)]
update for mypy 1.0 dev
As I need dmypy to work without facing [1], I am
running the latest build of mypy which seems so far
to finally not have that issue.
update constructs that latest mypy is being more picky
about, including better typing for the _NONE_NAME
symbol used in constraints (porting those elements
from the Enum patch at
I15ac3daee770408b5795746f47c1bbd931b7d26d)
Tobias Pfeiffer [Mon, 28 Nov 2022 12:52:31 +0000 (07:52 -0500)]
add partial index predicate to SQLiteDialect.get_indexes() result
Added support for reflection of expression-oriented WHERE criteria included
in indexes on the SQLite dialect, in a manner similar to that of the
PostgreSQL dialect. Pull request courtesy Tobias Pfeiffer.
Mike Bayer [Mon, 28 Nov 2022 15:58:49 +0000 (10:58 -0500)]
identify unresolvable Mapped types
Fixed issue where use of an unknown datatype within a :class:`.Mapped`
annotation for a column-based attribute would silently fail to map the
attribute, rather than reporting an exception; an informative exception
message is now raised.
tighten up iteration of names on mapped classes to more fully
exclude a large number of underscored names, so that we can avoid trying
to look at annotations for them or anything else. centralize the
"list of names we care about" more fully within _cls_attr_resolver
and base it on underscore conventions we should usually ignore,
with the exception of the few underscore names we want to see.
The boolness of the range was defined to be equal to its emptiness. As this has been identified as a typo rather than the intended, this inverts the logic, to match common Python behaviour as well as how other popular PG drivers do.
Added :class:`_expression.ScalarValues` that can be used as a column
element allowing using :class:`_expression.Values` inside IN clauses
or in conjunction with ``ANY`` or ``ALL`` collection aggregates.
This new class is generated using the method
:meth:`_expression.Values.scalar_values`.
The :class:`_expression.Values` instance is now coerced to a
:class:`_expression.ScalarValues` when used in a ``IN`` or ``NOT IN``
operation.
Mike Bayer [Sat, 26 Nov 2022 16:03:45 +0000 (11:03 -0500)]
add new variation helper
I'm using a lot of @testing.combinations with either
a boolean True/False, or a series of string names, each indicating
some case to switch on. I want a descriptive name in the test
run (not True/False) and I don't want to compare strings.
So make a new helper around @combinations that provides an
object interface that has booleans inside of it, prints nicely
in the test output, raises an error if you name the case
incorrectly.
Federico Caselli [Sat, 26 Nov 2022 18:24:49 +0000 (19:24 +0100)]
Add new script to sync test files adding future annotation
Create a new test file test_tm_future_annotations_sync.py that's obtained
from test_tm_future_annotations.py by using the new script sync_test_files.
This files includes the ``from __future__ import annotations``
It also turns out we had some bugs in there and we can also support some
additional typing things
Mike Bayer [Fri, 25 Nov 2022 21:49:28 +0000 (16:49 -0500)]
improve column targeting issues with query_expression
Fixed issues in :func:`_orm.with_expression` where expressions that were
composed of columns within a subquery being SELECTed from, or when using
``.from_statement()``, would not render correct SQL **if** the expression
had a label name that matched the attribute which used
:func:`_orm.query_expression`, even when :func:`_orm.query_expression` had
no default expression. For the moment, if the :func:`_orm.query_expression`
**does** have a default expression, that label name is still used for that
default, and an additional label with the same name will be ignored.
Overall, this case is pretty thorny so further adjustments might be
warranted.
Mike Bayer [Wed, 23 Nov 2022 15:58:28 +0000 (10:58 -0500)]
add "merge" to viewonly cascades; propagate NO_RAISE when merging
Fixed bug where :meth:`_orm.Session.merge` would fail to preserve the
current loaded contents of relationship attributes that were indicated with
the :paramref:`_orm.relationship.viewonly` parameter, thus defeating
strategies that use :meth:`_orm.Session.merge` to pull fully loaded objects
from caches and other similar techniques. In a related change, fixed issue
where an object that contains a loaded relationship that was nonetheless
configured as ``lazy='raise'`` on the mapping would fail when passed to
:meth:`_orm.Session.merge`; checks for "raise" are now suspended within
the merge process assuming the :paramref:`_orm.Session.merge.load`
parameter remains at its default of ``True``.
Overall, this is a behavioral adjustment to a change introduced in the 1.4
series as of :ticket:`4994`, which took "merge" out of the set of cascades
applied by default to "viewonly" relationships. As "viewonly" relationships
aren't persisted under any circumstances, allowing their contents to
transfer during "merge" does not impact the persistence behavior of the
target object. This allows :meth:`_orm.Session.merge` to correctly suit one
of its use cases, that of adding objects to a :class:`.Session` that were
loaded elsewhere, often for the purposes of restoring from a cache.
Mike Bayer [Wed, 23 Nov 2022 22:44:30 +0000 (17:44 -0500)]
flake8 has very courteously referred us to another pyqca project and closed all comments, thank you! I'll get on that right away
pyqca/flake8-import-order does not seem to have a release or a commit
in the past two years, so while I have created an issue and PR [1] [2],
for now vendor our fork so we can get on with things.
Mike Bayer [Wed, 16 Nov 2022 17:45:54 +0000 (12:45 -0500)]
ensure implicit_returning is checked on a Table instance
Fixed regression where flushing a mapped class that's mapped against a
subquery, such as a direct mapping or some forms of concrete table
inheritance, would fail if the :paramref:`_orm.Mapper.eager_defaults`
parameter were used.
mapper.local_table can still be a non-table instance.
correct the check first added in 466ed5b53a3.
Mike Bayer [Mon, 21 Nov 2022 14:46:43 +0000 (09:46 -0500)]
fix optionalized forms for dict[]
Fixed a suite of issues involving :class:`.Mapped` use with dictionary
types, such as ``Mapped[dict[str, str] | None]``, would not be correctly
interpreted in Declarative ORM mappings. Support to correctly
"de-optionalize" this type including for lookup in type_annotation_map has
been fixed.
Sam Bull [Sun, 20 Nov 2022 00:38:04 +0000 (19:38 -0500)]
Fix inferred type of async_sessionmaker
Improved the typing for :class:`.sessionmaker` and
:class:`.asyncsessionmaker`, so that the default type of their return value
will be :class:`.Session` or :class:`.AsyncSession`, without the need to
type this explicitly. Previously, Mypy would not automaticaly infer these
return types from its generic base.
As part of this change, arguments for :class:`.Session`,
:class:`.AsyncSession`, :class:`.sessionmaker` and
:class:`.asyncsessionmaker` beyond the initial "bind" argument have been
made keyword-only, which includes parameters that have always been
documented as keyword arguments, such as :paramref:`.Session.autoflush`,
:paramref:`.Session.class_`, etc.
Mike Bayer [Sat, 19 Nov 2022 21:42:22 +0000 (16:42 -0500)]
add common base class for all SQL col expression objects
Added a new type :class:`.SQLColumnExpression` which may be indicated in
user code to represent any SQL column oriented expression, including both
those based on :class:`.ColumnElement` as well as on ORM
:class:`.QueryableAttribute`. This type is a real class, not an alias, so
can also be used as the foundation for other objects.
Mike Bayer [Mon, 21 Nov 2022 13:51:34 +0000 (08:51 -0500)]
fall back to eval() for names that have dots
Fixed regression in 2.0.0b3 caused by :ticket:`8759` where indicating the
:class:`.Mapped` name using a qualified name such as
``sqlalchemy.orm.Mapped`` would fail to be recognized by Declarative as
indicating the :class:`.Mapped` construct.
Mike Bayer [Thu, 17 Nov 2022 01:11:18 +0000 (20:11 -0500)]
accommodate NULL format_type()
Made an adjustment to how the PostgreSQL dialect considers column types
when it reflects columns from a table, to accommodate for alternative
backends which may return NULL from the PG ``format_type()`` function.
command run is "pyupgrade --py37-plus --keep-runtime-typing --keep-percent-format <files...>"
pyupgrade will change assert_ to assertTrue. That was reverted since assertTrue does not
exists in sqlalchemy fixtures
Mike Bayer [Mon, 14 Nov 2022 02:48:53 +0000 (21:48 -0500)]
perf improvements related to corresponding_column (2)
commit two of two. this reorganizes ColumnCollection
to build a new index up front that's used to optimize
the corresponding_column() method.
Additional performance enhancements within ORM-enabled SQL statements,
specifically targeting callcounts within the construction of ORM
statements, using combinations of :func:`_orm.aliased` with
:func:`_sql.union` and similar "compound" constructs, in addition to direct
performance improvements to the ``corresponding_column()`` internal method
that is used heavily by the ORM by constructs like :func:`_orm.aliased` and
similar.
Mike Bayer [Sun, 13 Nov 2022 16:49:43 +0000 (11:49 -0500)]
don't invoke fromclause.c when creating an annotated
The ``aliased()`` constructor calls upon ``__clause_element__()``,
which internally annotates a ``FromClause``, like a subquery.
This became expensive as ``AnnotatedFromClause`` has for
many years called upon ``element.c`` so that the full ``.c``
collection is transferred to the Annotated.
Taking this out proved to be challenging. A straight remove
seemed to not break any tests except for the one that
tested the exact condition. Nevertheless this seemed
"spooky" so I instead moved the get of ``.c`` to be in a
memoized proxy method. However, that then exposed
a recursion issue related to loader_criteria; so the
source of that behavior, which was an accidental behavioral
artifact, is now made into an explcicit option that
loader_criteria uses directly.
The accidental behavioral artifact in question is still
kind of strange since I was not able to fully trace out
how it works, but the end result is that fixing the
artifact to be "correct" causes loader_criteria, within
the particular test for #7491, creates a select/
subquery structure with a cycle in it, so compilation fails
with recursion overflow.
The "solution" is to cause the artifact to occur in this
case, which is that the ``AnnotatedFromClause`` will have a
different ``.c`` collection than its element, which is a
subquery. It's not totally clear how a cycle is generated
when this is not done.
This is commit one of two, which goes through
some hoops to make essentially a one-line change.
The next commit will rework ColumnCollection to optimize
the corresponding_column() method significantly.
projects / thirdparty / sqlalchemy / sqlalchemy.git / log
