]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/log
thirdparty/sqlalchemy/sqlalchemy.git
13 years ago- [bug] Added ORA-03135 to the never ending
Mike Bayer [Wed, 25 Jan 2012 17:49:58 +0000 (12:49 -0500)] 
- [bug] Added ORA-03135 to the never ending
list of oracle "connection lost" errors
[ticket:2388]

13 years agoplay the hg dance with pjenvey
Mike Bayer [Tue, 24 Jan 2012 07:24:33 +0000 (02:24 -0500)] 
play the hg dance with pjenvey

13 years agoadd a thing about migration tools
Mike Bayer [Tue, 24 Jan 2012 07:24:01 +0000 (02:24 -0500)] 
add a thing about migration tools

13 years agoo null check PyObject_Repr results
Philip Jenvey [Tue, 24 Jan 2012 05:40:09 +0000 (21:40 -0800)] 
o null check PyObject_Repr results
o limit size of strings passed to PyErr_Format

13 years agodon't need to use __builtin__ for these things, doesn't work in py3k
Mike Bayer [Tue, 24 Jan 2012 00:16:36 +0000 (19:16 -0500)] 
don't need to use __builtin__ for these things, doesn't work in py3k

13 years agofix image link
Mike Bayer [Mon, 23 Jan 2012 21:35:16 +0000 (16:35 -0500)] 
fix image link

13 years agointegrate new readthedocs/sqla.org doc build
Mike Bayer [Mon, 23 Jan 2012 21:24:24 +0000 (16:24 -0500)] 
integrate new readthedocs/sqla.org doc build

13 years agoexception pickling fails on cx_oracle
Mike Bayer [Mon, 23 Jan 2012 03:37:23 +0000 (22:37 -0500)] 
exception pickling fails on cx_oracle

13 years ago2.4 doesn't have any()
Mike Bayer [Mon, 23 Jan 2012 01:51:41 +0000 (20:51 -0500)] 
2.4 doesn't have any()

13 years ago- [feature] Added new capability to relationship
Mike Bayer [Mon, 23 Jan 2012 01:11:03 +0000 (20:11 -0500)] 
- [feature] Added new capability to relationship
loader options to allow "default" loader strategies.
Pass '*' to any of joinedload(), lazyload(),
subqueryload(), or noload() and that becomes the
loader strategy used for all relationships,
except for those explicitly stated in the
Query.  Thanks to up-and-coming contributor
Kent Bower for an exhaustive and well
written test suite !  [ticket:2351]

13 years ago - [bug] Dropped the "30 char" limit on pymssql,
Mike Bayer [Mon, 23 Jan 2012 00:25:41 +0000 (19:25 -0500)] 
  - [bug] Dropped the "30 char" limit on pymssql,
    based on reports that it's doing things
    better these days.  pymssql hasn't been
    well tested and as the DBAPI is in flux
    it's still not clear what the status
    is on this driver and how SQLAlchemy's
    implementation should adapt.  [ticket:2347]

13 years ago- [bug] Fixed large_collection.py to close the
Mike Bayer [Mon, 23 Jan 2012 00:21:51 +0000 (19:21 -0500)] 
- [bug] Fixed large_collection.py to close the
session before dropping tables. [ticket:2346]

13 years ago - [bug] Adjusted the regexp used in the
Mike Bayer [Mon, 23 Jan 2012 00:17:21 +0000 (19:17 -0500)] 
  - [bug] Adjusted the regexp used in the
    mssql.TIME type to ensure only six digits
    are received for the "microseconds" portion
    of the value, which is expected by
    Python's datetime.time().  Note that
    support for sending microseconds doesn't
    seem to be possible yet with pyodbc
    at least.  [ticket:2340]

13 years ago- [feature] Simplified the versioning example
Mike Bayer [Mon, 23 Jan 2012 00:03:06 +0000 (19:03 -0500)] 
- [feature] Simplified the versioning example
a bit to use a declarative mixin as well
as an event listener, instead of a metaclass +
SessionExtension.  [ticket:2313]

13 years agomodernize/update cascade documentation, [ticket:2302]
Mike Bayer [Sun, 22 Jan 2012 23:32:41 +0000 (18:32 -0500)] 
modernize/update cascade documentation, [ticket:2302]

13 years ago- adjust the test for [ticket:2377] to be less controversial on
Mike Bayer [Sun, 22 Jan 2012 22:40:50 +0000 (17:40 -0500)] 
- adjust the test for [ticket:2377] to be less controversial on
problematic backends like Oracle.i
- move the check generated in r85017c4310d2 up for both label name/name
comparisions, fixes additional mismatches which can occur

13 years agofix a callcount issue here
Mike Bayer [Sun, 22 Jan 2012 22:30:52 +0000 (17:30 -0500)] 
fix a callcount issue here

13 years agocouple more vars not needed outside
Mike Bayer [Sun, 22 Jan 2012 21:54:58 +0000 (16:54 -0500)] 
couple more vars not needed outside

13 years ago- apply the pyflakes stick to strategies.py. In particular, try
Mike Bayer [Sun, 22 Jan 2012 21:49:08 +0000 (16:49 -0500)] 
- apply the pyflakes stick to strategies.py.  In particular, try
to get almost every method to be less than 50 lines.   For the relationship
loaders this meant finding some decent boundaries to split things up.
I find this style a bit unnatural as in some cases we are just doing
three things with a handful of variables, why shuffle them between
three methods that are entirely dependent on each other, but not
sure what other option there is for a long list of steps like
subquery loading.
Also renamed all the "loader callables" to something descriptive,
might help when they come up in stack traces.

13 years agopg8000 fix
Mike Bayer [Sun, 22 Jan 2012 20:20:36 +0000 (15:20 -0500)] 
pg8000 fix

13 years ago- [bug] Fixed bug whereby a table-bound Column
Mike Bayer [Sun, 22 Jan 2012 19:04:20 +0000 (14:04 -0500)] 
- [bug] Fixed bug whereby a table-bound Column
object named "<a>_<b>" which matched a column
labeled as "<tablename>_<colname>" could match
inappropriately when targeting in a result
set row.  [ticket:2377]
- requires that we change the tuple format in RowProxy.
Makes an improvement to the cases tested against
an unpickled RowProxy as well though doesn't solve the
problem there entirely.

13 years ago- rework the test suite to make use of SkipTest for tests skipped, unsupported, etc.
Mike Bayer [Sun, 22 Jan 2012 18:19:22 +0000 (13:19 -0500)] 
- rework the test suite to make use of SkipTest for tests skipped, unsupported, etc.
so that we can get an accurate picture what's really running/not, what's installed on jenkins, etc.
Tested in cpython 2.7 so far, we'll see what jenkins says about other platforms

13 years ago- [bug] Improved error messages when a non-string
Mike Bayer [Sun, 22 Jan 2012 17:11:13 +0000 (12:11 -0500)] 
- [bug] Improved error messages when a non-string
or invalid string is passed to any of the
date/time processors used by SQLite, including
C and Python versions.  [ticket:2382]
- changed the import model of processors.py so that we can
get at the pure python versions and C versions simultaneously
in tests.

13 years ago- [feature] Added "false()" and "true()" expression
Mike Bayer [Wed, 18 Jan 2012 17:42:54 +0000 (12:42 -0500)] 
- [feature] Added "false()" and "true()" expression
constructs to sqlalchemy.sql namespace, though
not part of __all__ as of yet.
- [bug] sql.false() and sql.true() compile to
0 and 1, respectively in sqlite [ticket:2368]

13 years ago- [bug] fixed regexp that filters out warnings
Mike Bayer [Wed, 18 Jan 2012 00:35:55 +0000 (19:35 -0500)] 
- [bug] fixed regexp that filters out warnings
for non-reflected "PARTITION" directives,
thanks to George Reilly [ticket:2376]

13 years agoanother serializable for [ticket:2371]
Mike Bayer [Thu, 12 Jan 2012 05:49:02 +0000 (00:49 -0500)] 
another serializable for [ticket:2371]

13 years ago- [bug] ensure pickleability of all ORM exceptions
Mike Bayer [Thu, 12 Jan 2012 01:58:10 +0000 (20:58 -0500)] 
- [bug] ensure pickleability of all ORM exceptions
for multiprocessing compatibility. [ticket:2371]

13 years agofurther fixes for column/table errors
Mike Bayer [Thu, 12 Jan 2012 01:47:53 +0000 (20:47 -0500)] 
further fixes for column/table errors

13 years agosome adjustments for py3k
Mike Bayer [Thu, 12 Jan 2012 01:45:28 +0000 (20:45 -0500)] 
some adjustments for py3k

13 years agofix failures for oursql, python 2.4
Mike Bayer [Thu, 12 Jan 2012 01:35:57 +0000 (20:35 -0500)] 
fix failures for oursql, python 2.4

13 years agoadd examples for multi metadata under __abstract__, custom vertical partitioning
Mike Bayer [Wed, 11 Jan 2012 21:35:34 +0000 (16:35 -0500)] 
add examples for multi metadata under __abstract__, custom vertical partitioning

13 years ago- Fixed regression from 0.6 whereby if
Mike Bayer [Tue, 10 Jan 2012 16:43:47 +0000 (11:43 -0500)] 
- Fixed regression from 0.6 whereby if
"load_on_pending" relationship() flag were used
where a non-"get()" lazy clause needed to be
emitted on a pending object, it would fail
to load.

13 years ago- [bug] Added __reduce__ to StatementError,
Mike Bayer [Tue, 10 Jan 2012 16:15:11 +0000 (11:15 -0500)] 
- [bug] Added __reduce__ to StatementError,
DBAPIError so that exceptions are pickleable,
as when using multiprocessing.  However, not
all DBAPIs support this yet, such as
psycopg2. [ticket:2371]

13 years agosqlsoup is released separately
Mike Bayer [Sun, 8 Jan 2012 19:05:07 +0000 (14:05 -0500)] 
sqlsoup is released separately

13 years agofix the gettext to be the same way sphinx-quickstart gives you
Mike Bayer [Sun, 8 Jan 2012 18:27:50 +0000 (13:27 -0500)] 
fix the gettext to be the same way sphinx-quickstart gives you

13 years agofix small update doc issue [ticket:2158]
Mike Bayer [Sun, 8 Jan 2012 16:54:17 +0000 (11:54 -0500)] 
fix small update doc issue [ticket:2158]

13 years agothat's not a "name=0", that's a counter. so name is None unconditonally.
Mike Bayer [Sun, 8 Jan 2012 00:49:25 +0000 (19:49 -0500)] 
that's not a "name=0", that's a counter.  so name is None unconditonally.
[ticket:2348]

13 years agoadd support to generate gettext
Mike Bayer [Sat, 7 Jan 2012 22:19:07 +0000 (17:19 -0500)] 
add support to generate gettext

13 years ago- [bug] the "name" of an FK constraint in SQLite
Mike Bayer [Sat, 7 Jan 2012 20:37:51 +0000 (15:37 -0500)] 
- [bug] the "name" of an FK constraint in SQLite
is reflected as "None", not "0" [ticket:2364].
SQLite does not appear to support constraint
naming in any case (the names are ignored).

13 years ago- [feature] query.filter() accepts multiple
Mike Bayer [Sat, 7 Jan 2012 20:15:27 +0000 (15:15 -0500)] 
- [feature] query.filter() accepts multiple
criteria which will join via AND, i.e.
query.filter(x==y, z>q, ...)

13 years ago- [bug] Fixed bug whereby event.listen(SomeClass)
Mike Bayer [Sat, 7 Jan 2012 17:57:22 +0000 (12:57 -0500)] 
- [bug] Fixed bug whereby event.listen(SomeClass)
forced an entirely unnecessary compile of the
mapper, making events very hard to set up
at module import time (nobody noticed this ??)
[ticket:2367]

13 years ago- fix the index, [ticket:2366]
Mike Bayer [Thu, 5 Jan 2012 21:33:14 +0000 (16:33 -0500)] 
- fix the index, [ticket:2366]
- the most explicit warning ever about mapper events

13 years agohappy new year
Mike Bayer [Wed, 4 Jan 2012 20:38:26 +0000 (15:38 -0500)] 
happy new year

13 years agodate change; hook test
Mike Bayer [Wed, 4 Jan 2012 20:32:32 +0000 (15:32 -0500)] 
date change; hook test

13 years agoanother date update/test a hook
Mike Bayer [Wed, 4 Jan 2012 20:29:12 +0000 (15:29 -0500)] 
another date update/test a hook

13 years agoupdate date here; I know it's in every file too but am testing some hooks too :)
Mike Bayer [Wed, 4 Jan 2012 20:25:09 +0000 (15:25 -0500)] 
update date here; I know it's in every file too but am testing some hooks too :)

13 years ago- rework "sql expressions as mapped attributes" to also stress hybrids as the first
Mike Bayer [Tue, 3 Jan 2012 22:07:44 +0000 (17:07 -0500)] 
- rework "sql expressions as mapped attributes" to also stress hybrids as the first
method to try.    now there's two sections here that introduce hybrids, which
is less than ideal.

13 years agoadd a "jinja2 fallback" renderer for render_string(), which is used for the js/css...
Mike Bayer [Fri, 30 Dec 2011 20:29:44 +0000 (15:29 -0500)] 
add a "jinja2 fallback" renderer for render_string(), which is used for the js/css templates
we don't really care about.  not sure if this is going to break again with a newer sphinx
or what, but fixes [ticket:2359] for now.

13 years agodocument that strings are accepted with declarative/secondary
Mike Bayer [Thu, 29 Dec 2011 16:27:50 +0000 (11:27 -0500)] 
document that strings are accepted with declarative/secondary

13 years ago- [feature] Added "class_registry" argument to
Mike Bayer [Wed, 28 Dec 2011 16:10:49 +0000 (11:10 -0500)] 
- [feature] Added "class_registry" argument to
declarative_base().  Allows two or more declarative
bases to share the same registry of class names.

13 years agoadd test for #2193
Mike Bayer [Tue, 27 Dec 2011 22:03:53 +0000 (17:03 -0500)] 
add test for #2193

13 years agofix a whole bunch of note:: / warning:: that were inline,
Mike Bayer [Sun, 25 Dec 2011 22:34:24 +0000 (17:34 -0500)] 
fix a whole bunch of note:: / warning:: that were inline,
no longer compatible with docutils 0.8

13 years agoadd a conditional here
Mike Bayer [Thu, 15 Dec 2011 16:42:50 +0000 (11:42 -0500)] 
add a conditional here

13 years ago- [bug] Fixed inappropriate usage of util.py3k
Mike Bayer [Thu, 15 Dec 2011 15:31:55 +0000 (10:31 -0500)] 
- [bug] Fixed inappropriate usage of util.py3k
flag and renamed it to util.py3k_warning, since
this flag is intended to detect the -3 flag
series of import restrictions only.
[ticket:2348]

13 years agofollow the pattern set earlier in the tests
Diana Clarke [Sun, 11 Dec 2011 23:13:35 +0000 (18:13 -0500)] 
follow the pattern set earlier in the tests

13 years agowhitespace pass
Diana Clarke [Sun, 11 Dec 2011 21:51:27 +0000 (16:51 -0500)] 
whitespace pass

13 years ago0.7.5 vers
Mike Bayer [Sun, 11 Dec 2011 15:42:18 +0000 (10:42 -0500)] 
0.7.5 vers

13 years ago - [bug] Fixed bug whereby hybrid_property didn't
Mike Bayer [Sun, 11 Dec 2011 15:41:33 +0000 (10:41 -0500)] 
  - [bug] Fixed bug whereby hybrid_property didn't
    work as a kw arg in any(), has().

13 years agoAdded tag rel_0_7_4 for changeset 5421b51a479a4c04da5175b58c8d8557ba2bc51c
Mike Bayer [Fri, 9 Dec 2011 22:19:59 +0000 (17:19 -0500)] 
Added tag rel_0_7_4 for changeset 5421b51a479a4c04da5175b58c8d8557ba2bc51c

13 years agomake the comparable_property() example somewhat meaningful rel_0_7_4
Mike Bayer [Fri, 9 Dec 2011 22:15:02 +0000 (17:15 -0500)] 
make the comparable_property() example somewhat meaningful

13 years ago- 0.7.4 prep
Mike Bayer [Fri, 9 Dec 2011 21:57:57 +0000 (16:57 -0500)] 
- 0.7.4 prep

13 years agoshould refer to Mapper class here
Mike Bayer [Fri, 9 Dec 2011 21:57:29 +0000 (16:57 -0500)] 
should refer to Mapper class here

13 years agoadd autoincrement directive for oracle
Mike Bayer [Fri, 9 Dec 2011 21:28:55 +0000 (16:28 -0500)] 
add autoincrement directive for oracle

13 years ago- Standalone expressions in polymorphic_on
Mike Bayer [Fri, 9 Dec 2011 16:12:41 +0000 (11:12 -0500)] 
- Standalone expressions in polymorphic_on
propagate to single-table inheritance
subclasses so that they are used in the
WHERE /JOIN clause to limit rows to that
subclass as is the usual behavior.
- make sure implicit map to polymorphic_on expr
handles creating a label().  Use an explicit name
here as _sa_polymorphic_on makes more sense when
poking around in _props.

13 years ago- [feature] polymorphic_on now accepts many
Mike Bayer [Fri, 9 Dec 2011 05:56:12 +0000 (00:56 -0500)] 
- [feature] polymorphic_on now accepts many
new kinds of values:

- standalone expressions that aren't
otherwise mapped
- column_property() objects
- string names of any column_property()
or attribute name of a mapped Column

The docs include an example using
the case() construct, which is likely to be
a common constructed used here.
[ticket:2345] and part of [ticket:2238]

13 years agocallcount
Mike Bayer [Wed, 7 Dec 2011 20:14:46 +0000 (15:14 -0500)] 
callcount

13 years ago- [bug] Fixed bug in get_history() when referring
Mike Bayer [Wed, 7 Dec 2011 18:24:11 +0000 (13:24 -0500)] 
- [bug] Fixed bug in get_history() when referring
to a composite attribute that has no value;
added coverage for get_history() regarding
composites which is otherwise just a userland
function.

13 years agotests for FK/autoincrement selection
Mike Bayer [Wed, 7 Dec 2011 03:00:35 +0000 (22:00 -0500)] 
tests for FK/autoincrement selection

13 years ago- doc updates per [ticket:2251]
Mike Bayer [Wed, 7 Dec 2011 01:59:04 +0000 (20:59 -0500)] 
- doc updates per [ticket:2251]
- [feature] Added new value for Column autoincrement
called "ignore_fk", can be used to force autoincrement
on a column that's still part of a ForeignKeyConstraint.
New example in the relationship docs illustrates
its use.

13 years ago- enable SAVEPOINT support fully, remove warning, [ticket:822].
Mike Bayer [Tue, 6 Dec 2011 22:41:19 +0000 (17:41 -0500)] 
- enable SAVEPOINT support fully, remove warning, [ticket:822].
It's not known what the potential "data loss" issues are, tests seem to pass.

13 years ago- [bug] Decode incoming values when retrieving
Mike Bayer [Tue, 6 Dec 2011 21:31:52 +0000 (16:31 -0500)] 
- [bug] Decode incoming values when retrieving
list of index names and the names of columns
within those indexes.  [ticket:2269]
- rewrite unicode reflection test to be of more general use on
broken backends

13 years ago- [bug] don't cast "table name" as NVARCHAR
Mike Bayer [Tue, 6 Dec 2011 20:35:06 +0000 (15:35 -0500)] 
- [bug] don't cast "table name" as NVARCHAR
on SQL Server 2000.  Still mostly in the dark
what incantations are needed to make PyODBC
work fully with FreeTDS 0.91 here, however.
[ticket:2343]

13 years agoadd docs regarding flags only working with setuptools/distribute, [ticket:2341]
Mike Bayer [Tue, 6 Dec 2011 19:28:54 +0000 (14:28 -0500)] 
add docs regarding flags only working with setuptools/distribute, [ticket:2341]

13 years ago- [feature] IdentitySet supports the - operator
Mike Bayer [Tue, 6 Dec 2011 18:22:59 +0000 (13:22 -0500)] 
- [feature] IdentitySet supports the - operator
as the same as difference(), handy when dealing
with Session.dirty etc. [ticket:2301]

13 years agoadd missing "data" column from association example, [ticket:2259]
Mike Bayer [Tue, 6 Dec 2011 18:06:30 +0000 (13:06 -0500)] 
add missing "data" column from association example, [ticket:2259]

13 years ago- [bug] the @compiles decorator raises an
Mike Bayer [Tue, 6 Dec 2011 17:49:39 +0000 (12:49 -0500)] 
- [bug] the @compiles decorator raises an
informative error message when no "default"
compilation handler is present, rather
than KeyError.

13 years agofix typo, [ticket:2232]
Mike Bayer [Tue, 6 Dec 2011 17:43:16 +0000 (12:43 -0500)] 
fix typo, [ticket:2232]

13 years agoadd BEGIN workaround to pysqlite docs, [ticket:2219]
Mike Bayer [Tue, 6 Dec 2011 17:41:01 +0000 (12:41 -0500)] 
add BEGIN workaround to pysqlite docs, [ticket:2219]

13 years agoupdate pyodbc/SQL server connection strings, [ticket:2217]
Mike Bayer [Tue, 6 Dec 2011 17:25:36 +0000 (12:25 -0500)] 
update pyodbc/SQL server connection strings, [ticket:2217]

13 years agoupdate delete-orphan cascade docs per [ticket:2332]
Mike Bayer [Tue, 6 Dec 2011 17:15:06 +0000 (12:15 -0500)] 
update delete-orphan cascade docs per [ticket:2332]

13 years agoupdate query.update(), query.delete(), sessionevents.after_bulk_update(), sessioneven...
Mike Bayer [Tue, 6 Dec 2011 16:53:49 +0000 (11:53 -0500)] 
update query.update(), query.delete(), sessionevents.after_bulk_update(), sessionevents.after_bulk_delete() docstirngs, [ticket:2324]

13 years agorefine this a bit
Mike Bayer [Mon, 5 Dec 2011 22:25:53 +0000 (17:25 -0500)] 
refine this a bit

13 years ago- [feature] Added an example to the hybrid docs
Mike Bayer [Mon, 5 Dec 2011 22:06:15 +0000 (17:06 -0500)] 
- [feature] Added an example to the hybrid docs
of a "transformer" - a hybrid that returns a
query-transforming callable in combination
with a custom comparator.   Uses a new method
on Query called with_transformation().  The use
case here is fairly experimental, but only
adds one line of code to Query.

13 years agomarathon doc updating session including a rewrite of unicode paragraphs
Mike Bayer [Mon, 5 Dec 2011 04:08:35 +0000 (23:08 -0500)] 
marathon doc updating session including a rewrite of unicode paragraphs

13 years ago- update docs to declarative [ticket:2323]
Mike Bayer [Sun, 4 Dec 2011 21:58:14 +0000 (16:58 -0500)] 
- update docs to declarative [ticket:2323]

13 years ago- [bug] Updated warning message when setting
Mike Bayer [Sun, 4 Dec 2011 21:23:55 +0000 (16:23 -0500)] 
- [bug] Updated warning message when setting
delete-orphan without delete to no longer
refer to 0.6, as we never got around to
upgrading this to an exception.  Ideally
this might be better as an exception but
it's not critical either way.  [ticket:2325]
- add new test suite to test_cascades for cascade
arguments

13 years ago- reflection process always resets the primary key constraint
Mike Bayer [Sun, 4 Dec 2011 20:57:08 +0000 (15:57 -0500)] 
- reflection process always resets the primary key constraint
first, fixes extend_existing

13 years ago-re-document the is_modified method, and place significant caveats
Mike Bayer [Sun, 4 Dec 2011 20:47:13 +0000 (15:47 -0500)] 
-re-document the is_modified method, and place significant caveats
about the "passive" flag which is now known to be wrong.  Add
new is_modified tests illustrating the passive flag.
[ticket:2320]

13 years ago- [bug] __table_args__ can now be passed as
Mike Bayer [Sun, 4 Dec 2011 19:28:57 +0000 (14:28 -0500)] 
- [bug] __table_args__ can now be passed as
an empty tuple as well as an empty dict.
[ticket:2339].  Thanks to Fayaz Yusuf Khan
for the patch.

13 years ago- [bug] Fixed bug whereby transaction.rollback()
Mike Bayer [Sun, 4 Dec 2011 19:25:00 +0000 (14:25 -0500)] 
- [bug] Fixed bug whereby transaction.rollback()
would throw an error on an invalidated
connection if the transaction were a
two-phase or savepoint transaction.
For plain transactions, rollback() is a no-op
if the connection is invalidated, so while
it wasn't 100% clear if it should be a no-op,
at least now the interface is consistent.
[ticket:2317]

13 years ago- [feature] The "extend_existing" flag on Table
Mike Bayer [Sun, 4 Dec 2011 19:16:42 +0000 (14:16 -0500)] 
- [feature] The "extend_existing" flag on Table
now allows for the reflection process to take
effect for a Table object that's already been
defined; when autoload=True and extend_existing=True
are both set, the full set of columns will be
reflected from the Table which will then
*overwrite* those columns already present,
rather than no activity occurring.  Columns that
are present directly in the autoload run
will be used as always, however.
[ticket:1410]

13 years ago- [bug] the "name" of a column-level CHECK constraint,
Mike Bayer [Sun, 4 Dec 2011 18:46:45 +0000 (13:46 -0500)] 
- [bug] the "name" of a column-level CHECK constraint,
if present, is now rendered in the CREATE TABLE
statement using "CONSTRAINT <name> CHECK <expression>".
[ticket:2305]

13 years ago- [bug] pyodbc-based dialects now parse the
Mike Bayer [Sun, 4 Dec 2011 18:29:29 +0000 (13:29 -0500)] 
- [bug] pyodbc-based dialects now parse the
pyodbc accurately as far as observed
pyodbc strings, including such gems
as "py3-3.0.1-beta4" [ticket:2318]
- [bug] use new pyodbc version detection for
_need_decimal_fix option, [ticket:2318]

13 years agosome call count reduction
Mike Bayer [Sun, 4 Dec 2011 01:59:56 +0000 (20:59 -0500)] 
some call count reduction

13 years ago- [bug] Calls to query.join() to a single-table
Mike Bayer [Sun, 4 Dec 2011 01:57:27 +0000 (20:57 -0500)] 
- [bug] Calls to query.join() to a single-table
inheritance subclass are now tracked, and
are used to eliminate the additional WHERE..
IN criterion normally tacked on with single
table inheritance, since the join should
accommodate it.  This allows OUTER JOIN
to a single table subclass to produce
the correct results, and overall will produce
fewer WHERE criterion when dealing with
single table inheritance joins.
[ticket:2328]

13 years ago- [bug] Fixed the error formatting raised when
Mike Bayer [Sun, 4 Dec 2011 01:09:27 +0000 (20:09 -0500)] 
- [bug] Fixed the error formatting raised when
a tuple is inadvertently passed to session.query()
[ticket:2297].

13 years ago- [bug] repaired the with_hint() feature which
Mike Bayer [Sun, 4 Dec 2011 01:04:55 +0000 (20:04 -0500)] 
- [bug] repaired the with_hint() feature which
wasn't implemented correctly on MSSQL -
usually used for the "WITH (NOLOCK)" hint
(which you shouldn't be using anyway !
use snapshot isolation instead :) )
[ticket:2336]

13 years ago- [bug] Fixed bug whereby column_property() created
Mike Bayer [Sun, 4 Dec 2011 00:50:01 +0000 (19:50 -0500)] 
- [bug] Fixed bug whereby column_property() created
against ORM-level column could be treated as
a distinct entity when producing certain
kinds of joined-inh joins.  [ticket:2316]

- [bug] related to [ticket:2316], made some
adjustments to the change from [ticket:2261]
regarding the "from" list on a select(). The
_froms collection is no longer memoized, as this
simplifies various use cases and removes the
need for a "warning" if a column is attached
to a table after it was already used in an
expression - the select() construct will now
always produce the correct expression.
There's probably no real-world
performance hit here; select() objects are
almost always made ad-hoc, and systems that
wish to optimize the re-use of a select()
would be using the "compiled_cache" feature.
A hit which would occur when calling select.bind
has been reduced, but the vast majority
of users shouldn't be using "bound metadata"
anyway :).

13 years agoadd supported implementations
Mike Bayer [Sat, 3 Dec 2011 17:39:41 +0000 (12:39 -0500)] 
add supported implementations

13 years ago- [bug] Fixed bug whereby "order_by='foreign_key'"
Mike Bayer [Thu, 1 Dec 2011 19:21:43 +0000 (14:21 -0500)] 
- [bug] Fixed bug whereby "order_by='foreign_key'"
option to Inspector.get_table_names
wasn't implementing the sort properly, replaced
with the existing sort algorithm
- clean up metadata usage in reflection tests