]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/log
thirdparty/sqlalchemy/sqlalchemy.git
14 years agocallcounts for 2.4, 2.5, 2.6
Mike Bayer [Sun, 5 Dec 2010 21:39:15 +0000 (16:39 -0500)] 
callcounts for 2.4, 2.5, 2.6

14 years agohave a 'buffer' symbol present in util in py3k as well
Mike Bayer [Sun, 5 Dec 2010 21:12:35 +0000 (16:12 -0500)] 
have a 'buffer' symbol present in util in py3k as well

14 years ago- fix callcount for 2.6
Mike Bayer [Sun, 5 Dec 2010 21:09:53 +0000 (16:09 -0500)] 
- fix callcount for 2.6

14 years agomerge tip
Mike Bayer [Sun, 5 Dec 2010 19:58:35 +0000 (14:58 -0500)] 
merge tip

14 years agolog [ticket:1978]
Mike Bayer [Sun, 5 Dec 2010 19:58:02 +0000 (14:58 -0500)] 
log [ticket:1978]

14 years ago- merge default tip
Mike Bayer [Sun, 5 Dec 2010 19:56:26 +0000 (14:56 -0500)] 
- merge default tip

14 years ago- refer to 0.7 in docs
Mike Bayer [Sun, 5 Dec 2010 18:29:34 +0000 (13:29 -0500)] 
- refer to 0.7 in docs

14 years ago- move topological, queue into util
Mike Bayer [Sun, 5 Dec 2010 17:51:24 +0000 (12:51 -0500)] 
- move topological, queue into util
- move function_named into test.lib.util
- use @decorator for all decorators in test/

14 years ago- more inlining. nominal execution on sqlite down to 36 calls, from 51 in 0.6.
Mike Bayer [Sun, 5 Dec 2010 17:09:06 +0000 (12:09 -0500)] 
- more inlining.   nominal execution on sqlite down to 36 calls, from 51 in 0.6.

14 years ago- reduce some pool overhead
Mike Bayer [Sun, 5 Dec 2010 07:10:49 +0000 (02:10 -0500)] 
- reduce some pool overhead

14 years ago- inline the various construction methods of DefaultExecutionContext. This
Mike Bayer [Sun, 5 Dec 2010 06:41:37 +0000 (01:41 -0500)] 
- inline the various construction methods of DefaultExecutionContext.   This
makes it sort of easier to read, sort of harder, depending on where you look.
Cuts down two or three method calls on execute.

14 years ago- ResultProxy and friends always reference the DBAPI connection at the same time
Mike Bayer [Sun, 5 Dec 2010 05:46:11 +0000 (00:46 -0500)] 
- ResultProxy and friends always reference the DBAPI connection at the same time
as the cursor.  There is no reason for CursorFairy - the only use case would be,
end-user is using the pool or pool.manage with DBAPI connections, uses a cursor,
deferences the owning connection and continues using cursor.  This is an almost
nonexistent use case and isn't correct usage at a DBAPI level.  Take out CursorFairy.
- move the "check for a dot in the colname" logic out to the sqlite dialect.

14 years ago- ordering for tests, [ticket:1888]
Mike Bayer [Sun, 5 Dec 2010 03:59:32 +0000 (22:59 -0500)] 
- ordering for tests, [ticket:1888]

14 years ago- Backport of "optimized get" fix from 0.7,
Mike Bayer [Sun, 5 Dec 2010 03:13:10 +0000 (22:13 -0500)] 
- Backport of "optimized get" fix from 0.7,
improves the generation of joined-inheritance
"load expired row" behavior.  [ticket:1992]

14 years ago- added strictness to the optimized load, [ticket:1992]
Mike Bayer [Fri, 3 Dec 2010 23:45:42 +0000 (18:45 -0500)] 
- added strictness to the optimized load, [ticket:1992]

14 years agonew calling style: event.listen(target, identifier, fn)
Mike Bayer [Wed, 1 Dec 2010 19:10:11 +0000 (14:10 -0500)] 
new calling style: event.listen(target, identifier, fn)

14 years ago- modernize multiple warning assertions for PG reflect index test
Mike Bayer [Wed, 1 Dec 2010 01:39:06 +0000 (20:39 -0500)] 
- modernize multiple warning assertions for PG reflect index test

14 years ago- replace GenericBackrefExtension with straight events
Mike Bayer [Wed, 1 Dec 2010 01:25:22 +0000 (20:25 -0500)] 
- replace GenericBackrefExtension with straight events
- add "backref" argument to register_attribute_impl

14 years ago- Fixed bug whereby KeyError would occur with non-ENUM
Mike Bayer [Tue, 30 Nov 2010 18:55:17 +0000 (13:55 -0500)] 
- Fixed bug whereby KeyError would occur with non-ENUM
supported PG versions after a pool dispose+recreate
would occur, [ticket:1989]

14 years ago- format guid hexstring explicit with 32 chars, [ticket:1988]
Mike Bayer [Tue, 30 Nov 2010 18:29:23 +0000 (13:29 -0500)] 
- format guid hexstring explicit with 32 chars, [ticket:1988]

14 years ago- use class name by itself in flush warnings to prevent overflow of warnings registry
Mike Bayer [Mon, 29 Nov 2010 23:38:44 +0000 (18:38 -0500)] 
- use class name by itself in flush warnings to prevent overflow of warnings registry
- test suite swaps out warnings.warn with warnings.warn_explicit, to solve warnings registry issue
- explicitly disallow functions from inside test.bootstrap, test.lib being interpreted as tests

14 years ago- repair MS-SQL handling of LIMIT/OFFSET as binds, add bindparam() to the compile...
Mike Bayer [Mon, 29 Nov 2010 22:28:36 +0000 (17:28 -0500)] 
- repair MS-SQL handling of LIMIT/OFFSET as binds, add bindparam() to the compile stream
so they function correctly

14 years ago- add docs for hybrid
Mike Bayer [Sun, 28 Nov 2010 22:13:07 +0000 (17:13 -0500)] 
- add docs for hybrid

14 years ago- fix bug preventing multiple SessionExtensions from being correctly converted to...
Mike Bayer [Sun, 28 Nov 2010 21:43:30 +0000 (16:43 -0500)] 
- fix bug preventing multiple SessionExtensions from being correctly converted to event listeners

14 years ago- restore the updates to test_assorted_eager that were part of
Mike Bayer [Sun, 28 Nov 2010 21:37:49 +0000 (16:37 -0500)] 
- restore the updates to test_assorted_eager that were part of
[ticket:1974] 's update

14 years agomerge backout
Mike Bayer [Sun, 28 Nov 2010 21:29:39 +0000 (16:29 -0500)] 
merge backout

14 years agoafter some usage, its clear that [ticket:1974] should not be implemented. backrefs
Mike Bayer [Sun, 28 Nov 2010 21:27:44 +0000 (16:27 -0500)] 
after some usage, its clear that [ticket:1974] should not be implemented.   backrefs
add to collections so its expected that collection membership would mirror in session
membership.
Backed out changeset e836366c843cd64a0df569582534868e3fb00f3b

14 years ago- assuming I read pep-386 correctly, the first 0.7 release will be 0.7b1
Mike Bayer [Sun, 28 Nov 2010 20:52:40 +0000 (15:52 -0500)] 
- assuming I read pep-386 correctly, the first 0.7 release will be 0.7b1

14 years ago- mappers can be created against select() constructs, or others,
Mike Bayer [Sun, 28 Nov 2010 20:50:00 +0000 (15:50 -0500)] 
- mappers can be created against select() constructs, or others,
that don't ultimately have any Table objects referenced.
[ticket:1876]

14 years agocombine test/engine/test_metadata.py and test/sql/test_columns.py into new
Mike Bayer [Sun, 28 Nov 2010 20:34:41 +0000 (15:34 -0500)] 
combine test/engine/test_metadata.py and test/sql/test_columns.py into new
test/sql/test_metadata.py, [ticket:1970]

14 years ago- replace util.py with util/ package, [ticket:1986]
Mike Bayer [Sun, 28 Nov 2010 20:32:09 +0000 (15:32 -0500)] 
- replace util.py with util/ package, [ticket:1986]

14 years ago- sqlalchemy.test and nose plugin moves back to being entirely
Mike Bayer [Sun, 28 Nov 2010 19:19:44 +0000 (14:19 -0500)] 
- sqlalchemy.test and nose plugin moves back to being entirely
outside of "sqlalchemy" and under "test/".

Rationale:

- coverage plugin works without issue, without need for an awkward
additional package install
- command line for "nosetests" isn't polluted with SQLAlchemy options

[ticket:1949]

14 years ago- a column with a "mutable" type mapped as "deferred" will not
Mike Bayer [Sun, 28 Nov 2010 18:45:51 +0000 (13:45 -0500)] 
- a column with a "mutable" type mapped as "deferred" will not
emit a load for the "old" value upon a set event,
unless the "active_history" flag is set to True.  [ticket:1976]

14 years ago- Pooling methods now underscored or removed:
Mike Bayer [Sun, 28 Nov 2010 18:40:56 +0000 (13:40 -0500)] 
- Pooling methods now underscored or removed:

    Pool.create_connection() -> Pool._create_connection()
    Pool.do_get() -> Pool._do_get()
    Pool.do_return_conn() -> Pool._do_return_conn()
    Pool.do_return_invalid() -> removed, was not used
    Pool.return_conn() -> Pool._return_conn()
    Pool.get() -> Pool._get(), public API is Pool.connect()
    SingletonThreadPool.cleanup() -> _cleanup()
    SingletonThreadPool.dispose_local() -> removed, use conn.invalidate()

[ticket:1982]

14 years ago- the "type_map" dictionary in sqlalchemy.types is now private,
Mike Bayer [Sun, 28 Nov 2010 18:07:00 +0000 (13:07 -0500)] 
- the "type_map" dictionary in sqlalchemy.types is now private,
i.e. is named "_type_map". [ticket:1870]

14 years ago- cascade_backrefs flag on relationship() now set to False by default.
Mike Bayer [Sun, 28 Nov 2010 18:01:23 +0000 (13:01 -0500)] 
- cascade_backrefs flag on relationship() now set to False by default.
[ticket:1974]

14 years ago- fix test related to [ticket:1980]
Mike Bayer [Sun, 28 Nov 2010 17:55:59 +0000 (12:55 -0500)] 
- fix test related to [ticket:1980]

14 years ago- the "mutable" flag on PickleType, postgresql.ARRAY is now off
Mike Bayer [Sun, 28 Nov 2010 17:16:39 +0000 (12:16 -0500)] 
- the "mutable" flag on PickleType, postgresql.ARRAY is now off
by default.  [ticket:1980]

14 years ago- the names used in the NamedTuple returned by Query.__iter__()
Mike Bayer [Sun, 28 Nov 2010 17:14:51 +0000 (12:14 -0500)] 
- the names used in the NamedTuple returned by Query.__iter__()
will be transferred from the individual Query objects that compose
a union() or similar, as opposed to returning the "table_colname"
alias generated when subqueries are generated.  [ticket:1942]

14 years ago- NullPool is now used by default for SQLite file-
Mike Bayer [Sun, 28 Nov 2010 17:00:01 +0000 (12:00 -0500)] 
- NullPool is now used by default for SQLite file-
 based databases.  :memory: databases will continue
 to select SingletonThreadPool by default.
 [ticket:1921]

14 years ago- post_process_text() is called for DDL() constructs, in particular allowing
Mike Bayer [Sun, 28 Nov 2010 16:52:24 +0000 (11:52 -0500)] 
- post_process_text() is called for DDL() constructs, in particular allowing
'%' with only one level of escaping.  Note this is backwards-incompatible
with previously triple-escaped sections.  [ticket:1897]

14 years ago- push polymorphic_on not recognized warning to an exception, [ticket:1875]
Mike Bayer [Sun, 28 Nov 2010 16:28:27 +0000 (11:28 -0500)] 
- push polymorphic_on not recognized warning to an exception, [ticket:1875]

14 years ago- _literal_as_text raises if the incoming arg is not a Visitable or basestring.
Mike Bayer [Sun, 28 Nov 2010 16:24:57 +0000 (11:24 -0500)] 
- _literal_as_text raises if the incoming arg is not a Visitable or basestring.
[ticket:1847]

14 years ago- Query.get() will raise if the number of params
Mike Bayer [Sun, 28 Nov 2010 16:10:41 +0000 (11:10 -0500)] 
- Query.get() will raise if the number of params
in a composite key is too large, as well as too
small. [ticket:1977]
- the above change smoked out an old mistake in a unit test.

14 years agofixed a small potential memory leak in UnicodeResultProcessor (for some weird
Gaëtan de Menten [Sat, 27 Nov 2010 21:27:42 +0000 (22:27 +0100)] 
fixed a small potential memory leak in UnicodeResultProcessor (for some weird
reason, it didn't actually leak in my tests) by providing a dealloc method to
the type, and added a test to ensure it stays that way. Closes #1981.

14 years ago- merge 2-arg form of query.join(), [ticket:1923]
Mike Bayer [Fri, 26 Nov 2010 21:40:41 +0000 (16:40 -0500)] 
- merge 2-arg form of query.join(), [ticket:1923]

14 years ago- merge mapper simpler compile branch, [ticket:1966]
Mike Bayer [Fri, 26 Nov 2010 21:28:41 +0000 (16:28 -0500)] 
- merge mapper simpler compile branch, [ticket:1966]

14 years ago- Fixed operator precedence rules for multiple
Mike Bayer [Thu, 25 Nov 2010 17:20:13 +0000 (12:20 -0500)] 
- Fixed operator precedence rules for multiple
chains of a single non-associative operator.
I.e. "x - (y - z)" will compile as "x - (y - z)"
and not "x - y - z".  Also works with labels,
i.e. "x - (y - z).label('foo')"
[ticket:1984]
- Single element tuple expressions inside an IN clause
parenthesize correctly, also from [ticket:1984],
added tests for PG
- re-fix again importlater, [ticket:1983]

14 years ago- adjust importlater to import the full path, break out of "import" problems with...
Mike Bayer [Thu, 25 Nov 2010 15:36:31 +0000 (10:36 -0500)] 
- adjust importlater to import the full path, break out of "import" problems with ImportError,
[ticket:1983]

14 years agothis is new
Mike Bayer [Wed, 24 Nov 2010 17:27:22 +0000 (12:27 -0500)] 
this is new

14 years agomerge tip
Mike Bayer [Wed, 24 Nov 2010 17:22:17 +0000 (12:22 -0500)] 
merge tip

14 years ago- SqlSoup overhaul
Mike Bayer [Wed, 24 Nov 2010 17:21:59 +0000 (12:21 -0500)] 
- SqlSoup overhaul
- Added "map_to()" method to SqlSoup, which is a "master"
method which accepts explicit arguments for each aspect of
the selectable and mapping, including a base class per
mapping. [ticket:1975]

- Mapped selectables used with the map(), with_labels(),
join() methods no longer put the given argument into the
internal "cache" dictionary.  Particularly since the
join() and select() objects are created in the method
itself this was pretty much a pure memory leaking behavior.

14 years ago- fixes for PG, mysql, twophase
Mike Bayer [Sat, 20 Nov 2010 22:57:30 +0000 (17:57 -0500)] 
- fixes for PG, mysql, twophase
- added "pool_events" arg to create_engine(), "events" to pool, allowing
establishment of listeners which fire before those of the dialect

14 years ago- logging has been overhauled such that engines no longer need to encode the
Mike Bayer [Sat, 20 Nov 2010 21:28:39 +0000 (16:28 -0500)] 
- logging has been overhauled such that engines no longer need to encode the
"hex id" string in their logging name in order to maintain separate loggers
per engine. thanks to Vinay Sajip for assistance.  merge of [ticket:1926]

14 years ago- the column assigned to polymorphic_on now behaves like any other
Mike Bayer [Sat, 20 Nov 2010 21:25:12 +0000 (16:25 -0500)] 
- the column assigned to polymorphic_on now behaves like any other
mapped attribute, in that it can be assigned to, mapped to multiple
columns.  It is also populated immediately upon object construction
with its class-based value, so the attribute can be read before
any flush occurs.  [ticket:1895]

14 years ago- bindparam() gets a new option "callable", which is a lambda or def
Mike Bayer [Sat, 20 Nov 2010 20:43:12 +0000 (15:43 -0500)] 
- bindparam() gets a new option "callable", which is a lambda or def
evaluated at execution time to determine the value.  This replaces
the implicit recognition of callables sent as the primary value
of bindparam(), which was an undocumented behavior used by the ORM.
The argument is separated now so that values can be passed to
bindparams that are also callables without ambiguity, such
as user defined objects that include a __call__() method.
[ticket:1950]

14 years ago- merge FromObj cleanup, immutable .c attribute patch, [ticket:1917]
Mike Bayer [Sat, 20 Nov 2010 20:35:43 +0000 (15:35 -0500)] 
- merge FromObj cleanup, immutable .c attribute patch, [ticket:1917]

14 years ago- merge metadata/schema/declarative branch, [ticket:1893]
Mike Bayer [Sat, 20 Nov 2010 20:30:12 +0000 (15:30 -0500)] 
- merge metadata/schema/declarative branch, [ticket:1893]

14 years ago- add distinct tests for mapping to join with "exclude" prop versus "explicit",
Mike Bayer [Sat, 20 Nov 2010 20:28:18 +0000 (15:28 -0500)] 
- add distinct tests for mapping to join with "exclude" prop versus "explicit",
should be the last thing needed for [ticket:1896], related to [ticket:1892]

14 years ago- merge no-save-cascade collection flushes branch, [ticket:1973]
Mike Bayer [Sat, 20 Nov 2010 20:11:12 +0000 (15:11 -0500)] 
- merge no-save-cascade collection flushes branch, [ticket:1973]

14 years agoone more orphan test
Mike Bayer [Sat, 20 Nov 2010 19:56:27 +0000 (14:56 -0500)] 
one more orphan test

14 years ago- the in-flush check for "pending orphan being flushed" has been removed.
Mike Bayer [Sat, 20 Nov 2010 19:52:21 +0000 (14:52 -0500)] 
- the in-flush check for "pending orphan being flushed" has been removed.
It's now possible to issue INSERT for an object that would otherwise
be an orphan due to no parent object specified, along a relationship
that specifies "delete-orphan" cascade.  The detection of this condition
as an error should now be accomplished via NOT NULL foreign keys.
[ticket:1912]

14 years ago- merge reverse inheriting col order, [ticket:1892]
Mike Bayer [Sat, 20 Nov 2010 18:49:17 +0000 (13:49 -0500)] 
- merge reverse inheriting col order, [ticket:1892]

14 years ago- the ordering of columns in a multi-column property now is in
Mike Bayer [Sat, 20 Nov 2010 18:44:03 +0000 (13:44 -0500)] 
- the ordering of columns in a multi-column property now is in
  reverse order of which they were added to the property. A
  typical effect of this is that the ".id" attribute on a
  joined-inheritance subclass, where both parent/child tables
  name the PK column ".id", will reference the ".id" column of
  the child table, not the parent, thus allowing join conditions
  and such to be constructed more intuitively. This is a
  behavior change for some joined-table inheritance queries.
  [ticket:1892]
- it's now an error condition to map to a join where multiple
  same-named columns from each table combine themselves
  implicitly. An explicit mention in the "properties" dictionary
  should be specified, using a list of columns, or
  column_property(*cols) with declarative. [ticket:1892]

14 years ago- merge limit/offset as binds branch, [ticket:805]
Mike Bayer [Sat, 20 Nov 2010 17:57:51 +0000 (12:57 -0500)] 
- merge limit/offset as binds branch, [ticket:805]

14 years ago- merge hybrid attributes branch, [ticket:1903]
Mike Bayer [Sat, 20 Nov 2010 17:55:46 +0000 (12:55 -0500)] 
- merge hybrid attributes branch, [ticket:1903]

14 years ago- merge events branch, [ticket:1902]
Mike Bayer [Sat, 20 Nov 2010 16:39:44 +0000 (11:39 -0500)] 
- merge events branch, [ticket:1902]

14 years ago- repair remaining tests
Mike Bayer [Sat, 20 Nov 2010 16:35:16 +0000 (11:35 -0500)] 
- repair remaining tests

14 years agomerge
Taavi Burns [Sat, 20 Nov 2010 13:35:24 +0000 (08:35 -0500)] 
merge

14 years agoRemove unused import.
Taavi Burns [Sat, 20 Nov 2010 13:32:00 +0000 (08:32 -0500)] 
Remove unused import.

14 years agoDecimalResultProcessor memory testing is likely to catch something only when cextensi...
Taavi Burns [Sat, 20 Nov 2010 13:30:07 +0000 (08:30 -0500)] 
DecimalResultProcessor memory testing is likely to catch something only when cextensions are enabled.  [ticket:1978]

14 years agoShorter tests for [ticket:1978]
Taavi Burns [Sat, 20 Nov 2010 02:27:30 +0000 (21:27 -0500)] 
Shorter tests for [ticket:1978]

14 years ago- modernize MutableTypesTest, add tests for expired/deferred to establish 0.6 behavior
Mike Bayer [Fri, 19 Nov 2010 23:49:06 +0000 (18:49 -0500)] 
- modernize MutableTypesTest, add tests for expired/deferred to establish 0.6 behavior
regarding [ticket:1976]

14 years ago- active_history flag also added to composite().
Mike Bayer [Fri, 19 Nov 2010 23:44:09 +0000 (18:44 -0500)] 
- active_history flag also added to composite().
The flag has no effect in 0.6, but is instead
a placeholder flag for forwards compatibility,
as it will be needed in 0.7 for composites.
[ticket:1976]

14 years agoFix memory leaks in the cprocessors DecimalResultProcessor, including tests. [ticket...
Taavi Burns [Fri, 19 Nov 2010 22:44:39 +0000 (17:44 -0500)] 
Fix memory leaks in the cprocessors DecimalResultProcessor, including tests.  [ticket:1978]

14 years agoskip on 2.4
Mike Bayer [Fri, 19 Nov 2010 01:12:24 +0000 (20:12 -0500)] 
skip on 2.4

14 years ago- remove really old, redundant test
Mike Bayer [Fri, 19 Nov 2010 00:21:37 +0000 (19:21 -0500)] 
- remove really old, redundant test

14 years agomerge tip
Mike Bayer [Fri, 19 Nov 2010 00:01:42 +0000 (19:01 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Fri, 19 Nov 2010 00:01:17 +0000 (19:01 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:57:33 +0000 (18:57 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:54:48 +0000 (18:54 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:53:46 +0000 (18:53 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:53:29 +0000 (18:53 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:52:51 +0000 (18:52 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:50:24 +0000 (18:50 -0500)] 
merge tip

14 years agomerge tip
Mike Bayer [Thu, 18 Nov 2010 23:49:42 +0000 (18:49 -0500)] 
merge tip

14 years ago- Added as_uuid=True flag to the UUID type, will receive
Mike Bayer [Thu, 18 Nov 2010 23:44:35 +0000 (18:44 -0500)] 
- Added as_uuid=True flag to the UUID type, will receive
and return values as Python UUID() objects rather than
strings.  Currently, the UUID type is only known to
work with psycopg2.  [ticket:1956]

14 years ago- Ensured every numeric, float, int code, scalar + array,
Mike Bayer [Thu, 18 Nov 2010 23:10:03 +0000 (18:10 -0500)] 
- Ensured every numeric, float, int code, scalar + array,
are recognized by psycopg2 and pg8000's "numeric"
base type. [ticket:1955]

14 years ago- fix test specification for oracle euro test
Mike Bayer [Thu, 18 Nov 2010 20:03:25 +0000 (15:03 -0500)] 
- fix test specification for oracle euro test

14 years ago- Added active_history flag to relationship()
Mike Bayer [Thu, 18 Nov 2010 19:55:19 +0000 (14:55 -0500)] 
- Added active_history flag to relationship()
and column_property(), forces attribute events to
always load the "old" value, so that it's available to
attributes.get_history(). [ticket:1961]
- modernize test_relationship, give test classes meaningful names

14 years ago- The mapper argument "primary_key" can be passed as a
Mike Bayer [Thu, 18 Nov 2010 17:33:25 +0000 (12:33 -0500)] 
- The mapper argument "primary_key" can be passed as a
single column as well as a list or tuple.  [ticket:1971]

14 years ago- The cx_oracle "decimal detection" logic, which takes place
Mike Bayer [Thu, 18 Nov 2010 17:19:31 +0000 (12:19 -0500)] 
- The cx_oracle "decimal detection" logic, which takes place
for for result set columns with ambiguous numeric characteristics,
now uses the decimal point character determined by the locale/
NLS_LANG setting, using an on-first-connect detection of
this character.  cx_oracle 5.0.3 or greater is also required
when using a non-period-decimal-point NLS_LANG setting.
[ticket:1953].

14 years ago- establish a consistent pattern of behavior along o2m, m2m, and m2o relationships
Mike Bayer [Thu, 18 Nov 2010 00:34:47 +0000 (19:34 -0500)] 
- establish a consistent pattern of behavior along o2m, m2m, and m2o relationships
when "save-update" cascade is disabled, or the target object is otherwise not
present in the session, and collection/scalar changes have taken place.  A warning
is emitted describing the type of operation, the target reference, and the relationship
description, stating that the operation will not take place.  The operation then doesn't
take place.   [ticket:1973]
- clean up test_cascade a little bit, remove cruft

14 years ago- initial patch for [ticket:1926]
Mike Bayer [Wed, 17 Nov 2010 22:54:52 +0000 (17:54 -0500)] 
- initial patch for [ticket:1926]

14 years ago- An error is raised if __table_args__ is not in tuple
Mike Bayer [Wed, 17 Nov 2010 15:55:10 +0000 (10:55 -0500)] 
- An error is raised if __table_args__ is not in tuple
or dict format, and is not None.  [ticket:1972]

14 years ago- primary_key arg to mapper() is a list
Mike Bayer [Tue, 16 Nov 2010 22:42:18 +0000 (17:42 -0500)] 
- primary_key arg to mapper() is a list
- other modernizations to mapper() doc

14 years ago- adapt initial patch from [ticket:1917] to current tip
Mike Bayer [Tue, 16 Nov 2010 20:53:14 +0000 (15:53 -0500)] 
- adapt initial patch from [ticket:1917] to current tip
- raise TypeError for immutability

14 years ago- set sqla_nose as the setup.py test runner - works !
Mike Bayer [Tue, 16 Nov 2010 18:33:48 +0000 (13:33 -0500)] 
- set sqla_nose as the setup.py test runner - works !

14 years ago- move sqlalchemy.test to test.lib
Mike Bayer [Tue, 16 Nov 2010 00:37:50 +0000 (19:37 -0500)] 
- move sqlalchemy.test to test.lib

14 years agofirst step of [ticket:1949], remove the setuptools aspect
Mike Bayer [Tue, 16 Nov 2010 00:25:34 +0000 (19:25 -0500)] 
first step of [ticket:1949], remove the setuptools aspect
of the plugin, move it to test/bootstrap