]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/log
thirdparty/sqlalchemy/sqlalchemy.git
16 years agoImplemented recreate() for StaticPool
Jason Kirtland [Fri, 10 Jul 2009 21:48:45 +0000 (21:48 +0000)] 
Implemented recreate() for StaticPool

16 years ago- Fixed potential memory leak whereby previously pickled objects
Mike Bayer [Fri, 10 Jul 2009 20:01:56 +0000 (20:01 +0000)] 
- Fixed potential memory leak whereby previously pickled objects
placed back in a session would not be fully garbage collected
unless the Session were explicitly closed out.

16 years ago- Fixed bug whereby session.is_modified() would raise an exception
Mike Bayer [Thu, 9 Jul 2009 01:45:44 +0000 (01:45 +0000)] 
- Fixed bug whereby session.is_modified() would raise an exception
if any synonyms were in use.

16 years ago- Fixed a bug involving contains_eager(), which would apply itself
Mike Bayer [Tue, 7 Jul 2009 17:17:22 +0000 (17:17 +0000)] 
- Fixed a bug involving contains_eager(), which would apply itself
to a secondary (i.e. lazy) load in a particular rare case,
producing cartesian products.   improved the targeting
of query.options() on secondary loads overall [ticket:1461].

16 years agoSession.mapper is now *deprecated*.
Mike Bayer [Fri, 3 Jul 2009 15:31:29 +0000 (15:31 +0000)] 
Session.mapper is now *deprecated*.
Call session.add() if you'd like a free-standing object to be
part of your session.  Otherwise, a DIY version of
Session.mapper is now documented at
http://www.sqlalchemy.org/trac/wiki/UsageRecipes/SessionAwareMapper
The method will remain deprecated throughout 0.6.

M    test/ext/test_declarative.py
M    test/orm/test_scoping.py
M    lib/sqlalchemy/orm/scoping.py
M    CHANGES

16 years agoFix deprecated usage of on numeric type
Lele Gaifax [Tue, 23 Jun 2009 09:09:13 +0000 (09:09 +0000)] 
Fix deprecated usage of  on numeric type

16 years agoadded docs for post_update
Mike Bayer [Sun, 21 Jun 2009 16:21:52 +0000 (16:21 +0000)] 
added docs for post_update

16 years ago- repaired non-working attributes.set_committed_value function.
Mike Bayer [Thu, 18 Jun 2009 19:37:16 +0000 (19:37 +0000)] 
- repaired non-working attributes.set_committed_value function.

16 years ago- Trimmed the pickle format for InstanceState which should further
Mike Bayer [Tue, 16 Jun 2009 19:23:43 +0000 (19:23 +0000)] 
- Trimmed the pickle format for InstanceState which should further
reduce the memory footprint of pickled instances.  The format
should be backwards compatible with that of 0.5.4 and previous.

16 years agoassoc proxy object appends to list automatically [ticket:1351]
Mike Bayer [Mon, 15 Jun 2009 22:39:45 +0000 (22:39 +0000)] 
assoc proxy object appends to list automatically [ticket:1351]

16 years agoadded test to verify #1423
Mike Bayer [Mon, 15 Jun 2009 22:33:12 +0000 (22:33 +0000)] 
added test to verify #1423

16 years ago- Fixed bug whereby list-based attributes, like pickletype
Mike Bayer [Mon, 15 Jun 2009 22:23:08 +0000 (22:23 +0000)] 
- Fixed bug whereby list-based attributes, like pickletype
and PGArray, failed to be merged() properly.

16 years agosplit CHANGES into CHANGES and CHANGES_PRE_05, since I would like CHANGES to be viewa...
Mike Bayer [Mon, 15 Jun 2009 22:11:08 +0000 (22:11 +0000)] 
split CHANGES into CHANGES and CHANGES_PRE_05, since I would like CHANGES to be viewable in trac

16 years ago- The "foreign_keys" argument of relation() will now propagate
Mike Bayer [Sat, 13 Jun 2009 03:31:30 +0000 (03:31 +0000)] 
- The "foreign_keys" argument of relation() will now propagate
automatically to the backref in the same way that
primaryjoin and secondaryjoin do.   For the extremely
rare use case where the backref of a relation() has
intentionally different "foreign_keys" configured, both sides
now need to be configured explicity (if they do in fact require
this setting, see the next note...).

- ...the only known (and really, really rare) use case where a
different foreign_keys setting was used on the forwards/backwards
side, a composite foreign key that partially points to its own
columns, has been enhanced such that the fk->itself aspect of the
relation won't be used to determine relation direction.

16 years ago- unit tests have been migrated from unittest to nose.
Mike Bayer [Wed, 10 Jun 2009 21:18:24 +0000 (21:18 +0000)] 
- unit tests have been migrated from unittest to nose.
See README.unittests for information on how to run
the tests.  [ticket:970]

16 years ago- Fixed Query being able to join() from individual columns of
Mike Bayer [Fri, 5 Jun 2009 21:23:11 +0000 (21:23 +0000)] 
- Fixed Query being able to join() from individual columns of
a joined-table subclass entity, i.e.
query(SubClass.foo, SubcClass.bar).join(<anything>).
In most cases, an error "Could not find a FROM clause to join
from" would be raised. In a few others, the result would be
returned in terms of the base class rather than the subclass -
so applications which relied on this erroneous result need to be
adjusted. [ticket:1431]

16 years ago- removed test.testing.ORMTest, test.fixtures, and all
Mike Bayer [Tue, 2 Jun 2009 21:42:14 +0000 (21:42 +0000)] 
- removed test.testing.ORMTest, test.fixtures, and all
dependencies on those.

16 years agomove from pXXX to 0.5.5 [ticket:1427]
Mike Bayer [Tue, 2 Jun 2009 15:50:04 +0000 (15:50 +0000)] 
move from pXXX to 0.5.5 [ticket:1427]

16 years ago- Fixed another 0.5.4 bug whereby mutable attributes (i.e. PickleType)
Mike Bayer [Mon, 1 Jun 2009 22:42:14 +0000 (22:42 +0000)] 
- Fixed another 0.5.4 bug whereby mutable attributes (i.e. PickleType)
wouldn't be deserialized correctly when the whole object
was serialized.  [ticket:1426]

16 years agoslight cleanup i want in 0.5/0.6
Mike Bayer [Sun, 31 May 2009 21:27:56 +0000 (21:27 +0000)] 
slight cleanup i want in 0.5/0.6

16 years agoremoved needless "thread" imports from util
Mike Bayer [Sat, 30 May 2009 01:09:16 +0000 (01:09 +0000)] 
removed needless "thread" imports from util

16 years ago- sql
Mike Bayer [Fri, 29 May 2009 18:56:50 +0000 (18:56 +0000)] 
- sql
    - Removed an obscure feature of execute() (including connection,
      engine, Session) whereby a bindparam() construct can be sent as
      a key to the params dictionary.  This usage is undocumented
      and is at the core of an issue whereby the bindparam() object
      created implicitly by a text() construct may have the same
      hash value as a string placed in the params dictionary and
      may result in an inappropriate match when computing the final
      bind parameters.   Internal checks for this condition would
      add significant latency to the critical task of parameter
      rendering, so the behavior is removed.  This is a backwards
      incompatible change for any application that may have been
      using this feature, however the feature has never been
      documented.

16 years ago- Fixed bug introduced in 0.5.4 whereby Composite types
Mike Bayer [Tue, 26 May 2009 22:45:56 +0000 (22:45 +0000)] 
- Fixed bug introduced in 0.5.4 whereby Composite types
fail when default-holding columns are flushed.

16 years agomore fixes to bound parameter exception reporting rel_0_5_4p2
Mike Bayer [Tue, 26 May 2009 17:03:03 +0000 (17:03 +0000)] 
more fixes to bound parameter exception reporting

16 years ago- added unit test for exception formatting
Mike Bayer [Tue, 26 May 2009 01:00:46 +0000 (01:00 +0000)] 
- added unit test for exception formatting
- Deprecated the hardcoded TIMESTAMP function, which when
used as func.TIMESTAMP(value) would render "TIMESTAMP value".
This breaks on some platforms as Postgres doesn't allow
bind parameters to be used in this context.  The hard-coded
uppercase is also inappropriate and there's lots of other
PG casts that we'd need to support.  So instead, use
text constructs i.e. select(["timestamp '12/05/09'"]).

16 years ago- Repaired the printing of SQL exceptions which are not
Mike Bayer [Mon, 25 May 2009 15:26:16 +0000 (15:26 +0000)] 
- Repaired the printing of SQL exceptions which are not
based on parameters.

16 years ago- Fixed an attribute error introduced in 0.5.4 which would rel_0_5_4p1
Mike Bayer [Mon, 18 May 2009 16:21:42 +0000 (16:21 +0000)] 
- Fixed an attribute error introduced in 0.5.4 which would
occur when merge() was used with an incomplete object.

16 years ago- Back-ported the "compiler" extension from SQLA 0.6. This rel_0_5_4
Mike Bayer [Sun, 17 May 2009 22:58:21 +0000 (22:58 +0000)] 
- Back-ported the "compiler" extension from SQLA 0.6.  This
is a standardized interface which allows the creation of custom
ClauseElement subclasses and compilers.  In particular it's
handy as an alternative to text() when you'd like to
build a construct that has database-specific compilations.
See the extension docs for details.

16 years agounusual ...
Mike Bayer [Sun, 17 May 2009 22:44:35 +0000 (22:44 +0000)] 
unusual ...

16 years ago- The "polymorphic discriminator" column may be part of a
Mike Bayer [Sun, 17 May 2009 22:20:28 +0000 (22:20 +0000)] 
- The "polymorphic discriminator" column may be part of a
primary key, and it will be populated with the correct
discriminator value.  [ticket:1300]

16 years ago- Reflecting a FOREIGN KEY construct will take into account
Mike Bayer [Sun, 17 May 2009 22:00:33 +0000 (22:00 +0000)] 
- Reflecting a FOREIGN KEY construct will take into account
a dotted schema.tablename combination, if the foreign key
references a table in a remote schema. [ticket:1405]

16 years ago- Exception messages are truncated when the list of bound
Mike Bayer [Sun, 17 May 2009 21:54:17 +0000 (21:54 +0000)] 
- Exception messages are truncated when the list of bound
parameters is larger than 10, preventing enormous
multi-page exceptions from filling up screens and logfiles
for large executemany() statements. [ticket:1413]

16 years ago- Removed all* O(N) scanning behavior from the flush() process,
Mike Bayer [Sun, 17 May 2009 21:51:40 +0000 (21:51 +0000)] 
- Removed all* O(N) scanning behavior from the flush() process,
i.e. operations that were scanning the full session,
including an extremely expensive one that was erroneously
assuming primary key values were changing when this
was not the case.

* one edge case remains which may invoke a full scan,
  if an existing primary key attribute is modified
  to a new value.

16 years ago - Significant performance enhancements regarding Sessions/flush()
Mike Bayer [Sun, 17 May 2009 18:17:46 +0000 (18:17 +0000)] 
- Significant performance enhancements regarding Sessions/flush()
      in conjunction with large mapper graphs, large numbers of
      objects:

      - The Session's "weak referencing" behavior is now *full* -
        no strong references whatsoever are made to a mapped object
        or related items/collections in its __dict__.  Backrefs and
        other cycles in objects no longer affect the Session's ability
        to lose all references to unmodified objects.  Objects with
        pending changes still are maintained strongly until flush.
        [ticket:1398]

        The implementation also improves performance by moving
        the "resurrection" process of garbage collected items
        to only be relevant for mappings that map "mutable"
        attributes (i.e. PickleType, composite attrs).  This removes
        overhead from the gc process and simplifies internal
        behavior.

        If a "mutable" attribute change is the sole change on an object
        which is then dereferenced, the mapper will not have access to
        other attribute state when the UPDATE is issued.  This may present
        itself differently to some MapperExtensions.

        The change also affects the internal attribute API, but not
        the AttributeExtension interface nor any of the publically
        documented attribute functions.

      - The unit of work no longer genererates a graph of "dependency"
        processors for the full graph of mappers during flush(), instead
        creating such processors only for those mappers which represent
        objects with pending changes.  This saves a tremendous number
        of method calls in the context of a large interconnected
        graph of mappers.

      - Cached a wasteful "table sort" operation that previously
        occured multiple times per flush, also removing significant
        method call count from flush().

      - Other redundant behaviors have been simplified in
        mapper._save_obj().

16 years agofix mysql tests
Mike Bayer [Thu, 14 May 2009 01:16:18 +0000 (01:16 +0000)] 
fix mysql tests

16 years ago- Fixed obscure mapper compilation issue when inheriting
Mike Bayer [Wed, 13 May 2009 19:34:21 +0000 (19:34 +0000)] 
- Fixed obscure mapper compilation issue when inheriting
mappers are used which would result in un-initialized
attributes.

16 years ago- It is now an error to specify both columns of a binary primaryjoin
Mike Bayer [Fri, 8 May 2009 01:41:51 +0000 (01:41 +0000)] 
- It is now an error to specify both columns of a binary primaryjoin
condition in the foreign_keys or remote_side collection.  Whereas
previously it was just nonsensical, but would succeed in a
non-deterministic way.

16 years ago- Fixed bug which prevented "mutable primary key" dependency
Mike Bayer [Fri, 8 May 2009 01:07:36 +0000 (01:07 +0000)] 
- Fixed bug which prevented "mutable primary key" dependency
logic from functioning properly on a one-to-one
relation().  [ticket:1406]
- moved MySQL to use innodb for naturalpks tests

16 years agoCorrected the SQLite SLBoolean type so that it properly treats 1 only as True. Fixes...
Michael Trier [Tue, 5 May 2009 00:36:37 +0000 (00:36 +0000)] 
Corrected the SQLite SLBoolean type so that it properly treats 1 only as True. Fixes #1402

16 years ago- MapperOptions and other state associated with query.options()
Mike Bayer [Sat, 2 May 2009 17:41:04 +0000 (17:41 +0000)] 
- MapperOptions and other state associated with query.options()
is no longer bundled within callables associated with each
lazy/deferred-loading attribute during a load.
The options are now associated with the instance's
state object just once when it's populated.  This removes
the need in most cases for per-instance/attribute loader
objects, improving load speed and memory overhead for
individual instances. [ticket:1391]

16 years agothis falls back to "expire" in any case since concat_op is not supported by the evaluator
Mike Bayer [Sat, 2 May 2009 15:35:37 +0000 (15:35 +0000)] 
this falls back to "expire" in any case since concat_op is not supported by the evaluator

16 years agoCorrected missing stop in the ORM Tutorial. Fixes #1395.
Michael Trier [Wed, 29 Apr 2009 00:14:09 +0000 (00:14 +0000)] 
Corrected missing stop in the ORM Tutorial. Fixes #1395.

16 years agoModified savepoint logic in mssql to ensure that it does not step on non-savepoint...
Michael Trier [Tue, 28 Apr 2009 03:35:35 +0000 (03:35 +0000)] 
Modified savepoint logic in mssql to ensure that it does not step on non-savepoint oriented routines. Savepoint support is still very experimental.

16 years ago- Allowed pickling of PropertyOption objects constructed with
Mike Bayer [Sun, 26 Apr 2009 21:57:18 +0000 (21:57 +0000)] 
- Allowed pickling of PropertyOption objects constructed with
instrumented descriptors; previously, pickle errors would occur
when pickling an object which was loaded with a descriptor-based
option, such as query.options(eagerload(MyClass.foo)).

16 years agoModified query_cls on DynamicAttribteImpl to accept a full mixin version of the Appen...
Michael Trier [Sat, 25 Apr 2009 15:35:52 +0000 (15:35 +0000)] 
Modified query_cls on DynamicAttribteImpl to accept a full mixin version of the AppenderQuery.

16 years agowe work with sphinx 0.6.1 now
Mike Bayer [Sat, 25 Apr 2009 15:13:30 +0000 (15:13 +0000)] 
we work with sphinx 0.6.1 now

16 years agoQuery.update() and Query.delete() should turn off eagerloads. Fixes #1378.
Ants Aasma [Mon, 20 Apr 2009 15:00:41 +0000 (15:00 +0000)] 
Query.update() and Query.delete() should turn off eagerloads. Fixes #1378.

16 years agoFixed adding of deferred or othe column properties to a declarative class.
Michael Trier [Sat, 18 Apr 2009 15:35:07 +0000 (15:35 +0000)] 
Fixed adding of deferred or othe column properties to a declarative class.

16 years agoRemoved allow_column_override documentation. Fixes #1381.
Michael Trier [Sat, 18 Apr 2009 01:21:38 +0000 (01:21 +0000)] 
Removed allow_column_override documentation. Fixes #1381.

16 years agoAdded copy and __copy__ methods to the OrderedDict. Fixes #1377.
Michael Trier [Mon, 13 Apr 2009 04:25:41 +0000 (04:25 +0000)] 
Added copy and __copy__ methods to the OrderedDict. Fixes #1377.

16 years agoCleaned up the deprecation problems with the examples.
Michael Trier [Mon, 13 Apr 2009 03:23:19 +0000 (03:23 +0000)] 
Cleaned up the deprecation problems with the examples.

16 years agoCorrected the sqlite float type so that it properly gets reflected as a SLFloat type...
Michael Trier [Mon, 13 Apr 2009 03:05:03 +0000 (03:05 +0000)] 
Corrected the sqlite float type so that it properly gets reflected as a SLFloat type. Fixes #1273.

16 years agoAdded in MSSQL reserved words list. Fixes #1310
Michael Trier [Sun, 12 Apr 2009 02:12:41 +0000 (02:12 +0000)] 
Added in MSSQL reserved words list. Fixes #1310

16 years agoAdded multi part schema name support. Closes #594 and #1341.
Michael Trier [Sat, 11 Apr 2009 21:36:45 +0000 (21:36 +0000)] 
Added multi part schema name support. Closes #594 and #1341.

16 years ago- Fixed a unit of work issue whereby the foreign
Mike Bayer [Sat, 11 Apr 2009 20:20:38 +0000 (20:20 +0000)] 
- Fixed a unit of work issue whereby the foreign
key attribute on an item contained within a collection
owned by an object being deleted would not be set to
None if the relation() was self-referential. [ticket:1376]

16 years agoCorrected duplication of serializer docs. Fixes #1375.
Michael Trier [Thu, 9 Apr 2009 22:57:16 +0000 (22:57 +0000)] 
Corrected duplication of serializer docs. Fixes #1375.

16 years ago- Fixed documentation for session weak_identity_map -
Mike Bayer [Thu, 9 Apr 2009 21:47:49 +0000 (21:47 +0000)] 
- Fixed documentation for session weak_identity_map -
the default value is True, indicating a weak
referencing map in use.

16 years agotest multi-level eager load without the limiting subquery
Mike Bayer [Wed, 8 Apr 2009 04:14:16 +0000 (04:14 +0000)] 
test multi-level eager load without the limiting subquery

16 years ago- Fixed the evaluator not being able to evaluate IS NULL clauses.
Ants Aasma [Mon, 6 Apr 2009 07:03:13 +0000 (07:03 +0000)] 
- Fixed the evaluator not being able to evaluate IS NULL clauses.
- Added evaluator tests to orm/alltests.py

16 years agodisabling triggers for Mysql since it requires SUPER privs
Mike Bayer [Sun, 5 Apr 2009 02:29:45 +0000 (02:29 +0000)] 
disabling triggers for Mysql since it requires SUPER privs

16 years agoAdded indexed to the list of reserved keywords (added in 3.6.4). Fixes #1358.
Michael Trier [Sat, 4 Apr 2009 02:54:09 +0000 (02:54 +0000)] 
Added indexed to the list of reserved keywords (added in 3.6.4). Fixes #1358.

16 years agoAdded Oracle examples showing how to use named parameters with a TNS. Fixes #1361.
Michael Trier [Sat, 4 Apr 2009 02:42:42 +0000 (02:42 +0000)] 
Added Oracle examples showing how to use named parameters with a TNS. Fixes #1361.

16 years agoCorrected examples in ORM tutorial to specify actual exception being thrown. Fixes...
Michael Trier [Sat, 4 Apr 2009 02:39:23 +0000 (02:39 +0000)] 
Corrected examples in ORM tutorial to specify actual exception being thrown. Fixes 1365.

16 years agoCorrected doc notations that suppressed some non-SQL output. Fixes #1366.
Michael Trier [Sat, 4 Apr 2009 00:48:26 +0000 (00:48 +0000)] 
Corrected doc notations that suppressed some non-SQL output. Fixes #1366.

16 years agoExplicit String length
Jason Kirtland [Fri, 3 Apr 2009 19:43:10 +0000 (19:43 +0000)] 
Explicit String length

16 years agoadd defaults to alltests
Mike Bayer [Fri, 3 Apr 2009 19:37:56 +0000 (19:37 +0000)] 
add defaults to alltests

16 years ago- Fixed bug in relation(), introduced in 0.5.3,
Mike Bayer [Thu, 2 Apr 2009 14:53:28 +0000 (14:53 +0000)] 
- Fixed bug in relation(), introduced in 0.5.3,
whereby a self referential relation
from a base class to a joined-table subclass would
not configure correctly.

16 years agoLots of fixes to the code examples to specify imports explicitly.
Michael Trier [Tue, 31 Mar 2009 22:31:08 +0000 (22:31 +0000)] 
Lots of fixes to the code examples to specify imports explicitly.

Explicit imports make it easier for users to understand the examples.
Additionally a lot of the examples were fixed to work with the changes in the
0.5.x code base. One small correction to the Case expression.  Thanks a bunch
to Adam Lowry! Fixes #717.

16 years ago- Fixed another location where autoflush was interfering
Mike Bayer [Tue, 31 Mar 2009 14:57:19 +0000 (14:57 +0000)] 
- Fixed another location where autoflush was interfering
with session.merge().  autoflush is disabled completely
for the duration of merge() now. [ticket:1360]

16 years agoextract() is now dialect-sensitive and supports SQLite and others.
Jason Kirtland [Mon, 30 Mar 2009 20:41:48 +0000 (20:41 +0000)] 
extract() is now dialect-sensitive and supports SQLite and others.

16 years agoremove needless print stuff
Mike Bayer [Mon, 30 Mar 2009 15:38:00 +0000 (15:38 +0000)] 
remove needless print stuff

16 years ago- Lazy loader will not use get() if the "lazy load"
Mike Bayer [Sun, 29 Mar 2009 21:21:10 +0000 (21:21 +0000)] 
- Lazy loader will not use get() if the "lazy load"
SQL clause matches the clause used by get(), but
contains some parameters hardcoded.  Previously
the lazy strategy would fail with the get().  Ideally
get() would be used with the hardcoded parameters
but this would require further development.
[ticket:1357]

16 years ago- coverage dumps out separate reports for individual packages
Mike Bayer [Sun, 29 Mar 2009 20:23:05 +0000 (20:23 +0000)] 
- coverage dumps out separate reports for individual packages
- other coverage tips

16 years ago- added a section on using aliased() with a subquery
Mike Bayer [Sun, 29 Mar 2009 18:39:54 +0000 (18:39 +0000)] 
- added a section on using aliased() with a subquery
- doctests needed huge number of +NORMALIZE_WHITESPACE not needed before for some reason

16 years agoCorrected docstring for class_mapper. It does not accept an object. Fixes #1316.
Michael Trier [Sun, 29 Mar 2009 04:29:50 +0000 (04:29 +0000)] 
Corrected docstring for class_mapper. It does not accept an object. Fixes #1316.

16 years agoModified information_schema change to keep it backwards compatible.
Michael Trier [Sun, 29 Mar 2009 02:08:22 +0000 (02:08 +0000)] 
Modified information_schema change to keep it backwards compatible.

16 years agofix crappity
Mike Bayer [Fri, 27 Mar 2009 21:42:41 +0000 (21:42 +0000)] 
fix crappity

16 years ago- Fixed __repr__() and other _get_colspec() methods on
Mike Bayer [Fri, 27 Mar 2009 21:41:36 +0000 (21:41 +0000)] 
- Fixed __repr__() and other _get_colspec() methods on
ForeignKey constructed from __clause_element__() style
construct (i.e. declarative columns).  [ticket:1353]

16 years agoCorrected problem with information schema not working with binary collation on mssql...
Michael Trier [Fri, 27 Mar 2009 21:27:34 +0000 (21:27 +0000)] 
Corrected problem with information schema not working with binary collation on mssql. Fixes #1343.

16 years ago- Fixed the "set collection" function on "dynamic" relations
Mike Bayer [Fri, 27 Mar 2009 19:54:10 +0000 (19:54 +0000)] 
- Fixed the "set collection" function on "dynamic" relations
to initiate events correctly.  Previously a collection
could only be assigned to a pending parent instance,
otherwise modified events would not be fired correctly.
Set collection is now compatible with merge(),
fixes [ticket:1352].

16 years agomore tests rel_0_5_3
Mike Bayer [Tue, 24 Mar 2009 01:22:38 +0000 (01:22 +0000)] 
more tests

16 years ago - Fixed bug in dynamic_loader() where append/remove events
Mike Bayer [Tue, 24 Mar 2009 01:19:45 +0000 (01:19 +0000)] 
- Fixed bug in dynamic_loader() where append/remove events
      after construction time were not being propagated to the
      UOW to pick up on flush(). [ticket:1347]

16 years agoadd collections module to API ref for completeness. links to the mapper documentation
Mike Bayer [Sat, 21 Mar 2009 18:03:47 +0000 (18:03 +0000)] 
add collections module to API ref for completeness.  links to the mapper documentation
which is less wordy.

16 years ago- An alias() of a select() will convert to a "scalar subquery"
Mike Bayer [Sat, 21 Mar 2009 16:12:37 +0000 (16:12 +0000)] 
- An alias() of a select() will convert to a "scalar subquery"
when used in an unambiguously scalar context, i.e. it's used
in a comparison operation.  This applies to
the ORM when using query.subquery() as well.

16 years ago - Fixed SQLite reflection methods so that non-present
Mike Bayer [Tue, 17 Mar 2009 15:09:49 +0000 (15:09 +0000)] 
- Fixed SQLite reflection methods so that non-present
      cursor.description, which triggers an auto-cursor
      close, will be detected so that no results doesn't
      fail on recent versions of pysqlite which raise
      an error when fetchone() called with no rows present.

16 years ago- Query.join() can now construct multiple FROM clauses, if
Mike Bayer [Sun, 15 Mar 2009 03:02:42 +0000 (03:02 +0000)] 
- Query.join() can now construct multiple FROM clauses, if
needed.  Such as, query(A, B).join(A.x).join(B.y)
might say SELECT A.*, B.* FROM A JOIN X, B JOIN Y.
Eager loading can also tack its joins onto those
multiple FROM clauses.  [ticket:1337]

16 years ago - Fixed bug where column_prefix wasn't being checked before
Mike Bayer [Wed, 11 Mar 2009 21:45:57 +0000 (21:45 +0000)] 
- Fixed bug where column_prefix wasn't being checked before
      not mapping an attribute that already had class-level
      name present.

16 years agofix formatting to match unit tests
Mike Bayer [Wed, 11 Mar 2009 05:48:02 +0000 (05:48 +0000)] 
fix formatting to match unit tests

16 years agoadd schema to entity method
Jonathan Ellis [Mon, 9 Mar 2009 21:24:43 +0000 (21:24 +0000)] 
add schema to entity method

16 years ago- a forward and complementing backwards reference which are both
Mike Bayer [Mon, 9 Mar 2009 01:20:29 +0000 (01:20 +0000)] 
- a forward and complementing backwards reference which are both
of the same direction, i.e. ONETOMANY or MANYTOONE,
is now detected, and an error message is raised.
Saves crazy CircularDependencyErrors later on.

16 years agotake 2
Mike Bayer [Sun, 8 Mar 2009 19:40:12 +0000 (19:40 +0000)] 
take 2

16 years agoattempt to exlude sqlite 3.5.9 and below, for buildbot failure. not clear which...
Mike Bayer [Sun, 8 Mar 2009 19:29:56 +0000 (19:29 +0000)] 
attempt to exlude sqlite 3.5.9 and below, for buildbot failure.  not clear which version of sqlite fixes this particular issue

16 years ago- Query.group_by() properly takes into account aliasing applied
Mike Bayer [Sun, 8 Mar 2009 19:06:12 +0000 (19:06 +0000)] 
- Query.group_by() properly takes into account aliasing applied
to the FROM clause, such as with select_from(), using
with_polymorphic(), or using from_self().

16 years agosome buildbot fixes
Mike Bayer [Sun, 8 Mar 2009 02:35:46 +0000 (02:35 +0000)] 
some buildbot fixes

16 years ago- updated builders for latest sphinx tip
Mike Bayer [Mon, 2 Mar 2009 00:01:09 +0000 (00:01 +0000)] 
- updated builders for latest sphinx tip
- applied patch from [ticket:1321]

16 years ago- The "objects" argument to session.flush() is deprecated.
Mike Bayer [Sun, 1 Mar 2009 23:53:58 +0000 (23:53 +0000)] 
- The "objects" argument to session.flush() is deprecated.
State which represents the linkage between a parent and
child object does not support "flushed" status on
one side of the link and not the other, so supporting
this operation leads to misleading results.
[ticket:1315]

16 years ago- Added PGUuid and PGBit types to
Mike Bayer [Sun, 1 Mar 2009 20:24:02 +0000 (20:24 +0000)] 
- Added PGUuid and PGBit types to
sqlalchemy.databases.postgres. [ticket:1327]

- Refection of unknown PG types won't crash when those
types are specified within a domain.  [ticket:1327]

- executemany() in conjunction with INSERT..RETURNING is documented as undefined by psycopg2.

16 years ago- Fixed adaptation of EXISTS clauses via any(), has(), etc.
Mike Bayer [Thu, 26 Feb 2009 15:16:06 +0000 (15:16 +0000)] 
- Fixed adaptation of EXISTS clauses via any(), has(), etc.
in conjunction with an aliased object on the left and
of_type() on the right.  [ticket:1325]

16 years agoFix markup glitch
Lele Gaifax [Wed, 25 Feb 2009 13:20:38 +0000 (13:20 +0000)] 
Fix markup glitch

16 years ago- pared down private and semi-private functions in the attributes package.
Mike Bayer [Mon, 23 Feb 2009 00:08:37 +0000 (00:08 +0000)] 
- pared down private and semi-private functions in the attributes package.
- simplified the process of establishment and unestablishment of
class management from a mapper perspective; class manager setup/teardown
is now symmetric (ClassManager would never be fully de-associated previously).
- class manager now unconditionally decorates __init__.  this has a slight
behavior change for an unmapped subclass of a mapped superclass, in that
InstanceState creation corresponds to that of the superclass.  This
still doesn't allow unmapped subclasses to be usable in mapper
situations, though.