]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/log
thirdparty/sqlalchemy/sqlalchemy.git
15 years agoo don't need to str() Jython arrays twice
Philip Jenvey [Sat, 12 Sep 2009 00:18:06 +0000 (00:18 +0000)] 
o don't need to str() Jython arrays twice
o map the Binary types w/ new str handling in mssql's BinaryTest.test_binary

15 years agodon't assume dict ordering
Philip Jenvey [Fri, 11 Sep 2009 23:48:56 +0000 (23:48 +0000)] 
don't assume dict ordering

15 years agotypo
Philip Jenvey [Fri, 11 Sep 2009 23:18:27 +0000 (23:18 +0000)] 
typo

15 years agooracle/mssql+zxjdbc blurb
Philip Jenvey [Fri, 11 Sep 2009 23:16:20 +0000 (23:16 +0000)] 
oracle/mssql+zxjdbc blurb

15 years agoclose cursors: mostly fetchone -> first
Philip Jenvey [Fri, 11 Sep 2009 22:53:50 +0000 (22:53 +0000)] 
close cursors: mostly fetchone -> first

15 years agoupdate per new Binary str handling
Philip Jenvey [Fri, 11 Sep 2009 22:51:25 +0000 (22:51 +0000)] 
update per new Binary str handling

15 years agoall about DDL events
Mike Bayer [Fri, 11 Sep 2009 22:37:13 +0000 (22:37 +0000)] 
all about DDL events

15 years agomssql+zxjdbc support
Philip Jenvey [Fri, 11 Sep 2009 08:10:32 +0000 (08:10 +0000)] 
mssql+zxjdbc support
original patch from Victor Ng
fixes #1505

15 years agodocs
Mike Bayer [Thu, 10 Sep 2009 21:23:04 +0000 (21:23 +0000)] 
docs

15 years agomove foreign keys stuff into the constraints section
Mike Bayer [Wed, 9 Sep 2009 21:00:14 +0000 (21:00 +0000)] 
move foreign keys stuff into the constraints section

15 years agoedits plus redid "sequences".
Mike Bayer [Wed, 9 Sep 2009 20:18:48 +0000 (20:18 +0000)] 
edits plus redid "sequences".

15 years agofix update examples for [ticket:1533]
Mike Bayer [Wed, 9 Sep 2009 16:26:26 +0000 (16:26 +0000)] 
fix update examples for [ticket:1533]

15 years agofix the new Binary str handling under Jython
Philip Jenvey [Wed, 9 Sep 2009 06:48:06 +0000 (06:48 +0000)] 
fix the new Binary str handling under Jython

15 years ago- the Binary type now returns data as a Python string
Mike Bayer [Sat, 5 Sep 2009 03:29:20 +0000 (03:29 +0000)] 
- the Binary type now returns data as a Python string
(or a "bytes" type in Python 3), instead of the built-
in "buffer" type.  This allows symmetric round trips
of binary data. [ticket:1524]

15 years ago- distill contextual parameters for the purpose of default functions into context...
Mike Bayer [Thu, 3 Sep 2009 22:26:16 +0000 (22:26 +0000)] 
- distill contextual parameters for the purpose of default functions into context.current_parameters
- metadata docs continued

15 years agomysql doesn't like the 0's. not sure why but its not the point of the test in any...
Mike Bayer [Wed, 2 Sep 2009 16:22:47 +0000 (16:22 +0000)] 
mysql doesn't like the 0's.  not sure why but its not the point of the test in any case

15 years agoadd a length for mysql
Philip Jenvey [Wed, 2 Sep 2009 00:30:44 +0000 (00:30 +0000)] 
add a length for mysql

15 years ago- Fixed bug which prevented two entities from mutually
Mike Bayer [Tue, 1 Sep 2009 22:55:59 +0000 (22:55 +0000)] 
- Fixed bug which prevented two entities from mutually
replacing each other's primary key values within a single
flush() for some orderings of operations.  [ticket:1519]

15 years ago- Fixed bug which disallowed one side of a many-to-many
Mike Bayer [Tue, 1 Sep 2009 22:42:35 +0000 (22:42 +0000)] 
- Fixed bug which disallowed one side of a many-to-many
bidirectional reference to declare itself as "viewonly"
[ticket:1507]

15 years ago- A column can be added to a joined-table subclass after
Mike Bayer [Tue, 1 Sep 2009 22:26:23 +0000 (22:26 +0000)] 
- A column can be added to a joined-table subclass after
the class has been constructed (i.e. via class-level
attribute assignment).  The column is added to the underlying
Table as always, but now the mapper will rebuild its
"join" to include the new column, instead of raising
an error about "no such column, use column_property()
instead".  [ticket:1523]
- added an additional test in test_mappers for "added nonexistent column",
even though this test is already in test_query its more appropriate within
"mapper configuration" tests.

15 years ago- Fixed the error message for "could not find a FROM clause"
Mike Bayer [Tue, 1 Sep 2009 22:14:22 +0000 (22:14 +0000)] 
- Fixed the error message for "could not find a FROM clause"
in query.join() which would fail to issue correctly
if the query was against a pure SQL construct.
[ticket:1522]

15 years agodoh
Mike Bayer [Mon, 31 Aug 2009 21:59:45 +0000 (21:59 +0000)] 
doh

15 years ago- Fixed incorrect exception raise in
Mike Bayer [Mon, 31 Aug 2009 21:53:49 +0000 (21:53 +0000)] 
- Fixed incorrect exception raise in
Weak/StrongIdentityMap.add()
[ticket:1506]

15 years ago- Fixed recursion issue which occured if a mapped object's
Mike Bayer [Mon, 31 Aug 2009 21:24:52 +0000 (21:24 +0000)] 
- Fixed recursion issue which occured if a mapped object's
`__len__()` or `__nonzero__()` method resulted in state
changes.  [ticket:1501]

15 years ago- the allow_null_pks flag on mapper() is deprecated, and
Mike Bayer [Mon, 31 Aug 2009 20:54:19 +0000 (20:54 +0000)] 
- the allow_null_pks flag on mapper() is deprecated, and
the feature is turned "on" by default.  This means that
a row which has a non-null value for any of its primary key
columns will be considered an identity.  The need for this
scenario typically only occurs when mapping to an outer join.
[ticket:1339]
- streamlined the NULL check to use set operations

15 years ago- Inserting NULL into a primary key + foreign key column
Mike Bayer [Mon, 31 Aug 2009 20:38:14 +0000 (20:38 +0000)] 
- Inserting NULL into a primary key + foreign key column
will allow the "not null constraint" error to raise,
not an attempt to execute a nonexistent "col_id_seq"
sequence.  [ticket:1516]

- autoincrement SELECT statements, i.e. those which
select from a procedure that modifies rows, now work
with server-side cursor mode (the named cursor isn't
used for such statements.)

15 years agoRemove NCLOB from types.__all__, since it's not defined there. Not sure if this
Gaëtan de Menten [Mon, 31 Aug 2009 15:14:28 +0000 (15:14 +0000)] 
Remove NCLOB from types.__all__, since it's not defined there. Not sure if this
is the proper fix but the former situation made it impossible to do "from
sqlalchemy.types import *", which Elixir does.

15 years agorename jdbc suffix to zxjdbc, add ReturningParam.__ne__
Philip Jenvey [Sun, 30 Aug 2009 20:18:34 +0000 (20:18 +0000)] 
rename jdbc suffix to zxjdbc, add ReturningParam.__ne__

15 years agoassert -> eq_
Philip Jenvey [Sat, 29 Aug 2009 19:12:14 +0000 (19:12 +0000)] 
assert -> eq_

15 years agoassert -> eq_
Philip Jenvey [Sat, 29 Aug 2009 07:12:15 +0000 (07:12 +0000)] 
assert -> eq_

15 years agofix, comes back funky on PG reflection
Mike Bayer [Fri, 28 Aug 2009 20:36:28 +0000 (20:36 +0000)] 
fix, comes back funky on PG reflection

15 years ago- Fixed an obscure issue whereby a joined-table subclass
Mike Bayer [Fri, 28 Aug 2009 20:29:08 +0000 (20:29 +0000)] 
- Fixed an obscure issue whereby a joined-table subclass
with a self-referential eager load on the base class
would populate the related object's "subclass" table with
data from the "subclass" table of the parent.
[ticket:1485]

15 years ago- Fixed column.copy() to copy defaults and onupdates.
Mike Bayer [Fri, 28 Aug 2009 19:00:44 +0000 (19:00 +0000)] 
- Fixed column.copy() to copy defaults and onupdates.
[ticket:1373]

15 years agofix port specification
Philip Jenvey [Fri, 28 Aug 2009 03:27:29 +0000 (03:27 +0000)] 
fix port specification

15 years agouse *args with log.debug()/log.info(), [ticket:1520]
Mike Bayer [Wed, 26 Aug 2009 19:46:04 +0000 (19:46 +0000)] 
use *args with log.debug()/log.info(), [ticket:1520]

15 years agocorrect the create_engine url
Philip Jenvey [Wed, 26 Aug 2009 01:27:04 +0000 (01:27 +0000)] 
correct the create_engine url
fixes #1515
thanks Randall

15 years agodoc fixes
Mike Bayer [Tue, 25 Aug 2009 01:31:27 +0000 (01:31 +0000)] 
doc fixes

15 years agoinstance_dict may be modified before the GC triggers _cleanup on Jython, so eat
Philip Jenvey [Fri, 21 Aug 2009 02:19:07 +0000 (02:19 +0000)] 
instance_dict may be modified before the GC triggers _cleanup on Jython, so eat
state mismatch AssertionErrors

15 years agoget mysql+pyodbc connections to work again
Mike Bayer [Thu, 20 Aug 2009 22:55:59 +0000 (22:55 +0000)] 
get mysql+pyodbc connections to work again

15 years agothis workaround isn't necessary
Philip Jenvey [Tue, 18 Aug 2009 05:51:32 +0000 (05:51 +0000)] 
this workaround isn't necessary

15 years agofix oracle+zxjdbc asdecimal conversions
Philip Jenvey [Tue, 18 Aug 2009 05:30:50 +0000 (05:30 +0000)] 
fix oracle+zxjdbc asdecimal conversions

15 years agooracle+zxjdbc returning support
Philip Jenvey [Tue, 18 Aug 2009 05:28:05 +0000 (05:28 +0000)] 
oracle+zxjdbc returning support

15 years agoalways visit returning clauses in the right order for positional paramstyle
Philip Jenvey [Tue, 18 Aug 2009 02:43:37 +0000 (02:43 +0000)] 
always visit returning clauses in the right order for positional paramstyle
sanity

15 years agodont see the need to re-call post_configure inside a reentrant loop. since we're
Mike Bayer [Fri, 14 Aug 2009 17:02:56 +0000 (17:02 +0000)] 
dont see the need to re-call post_configure inside a reentrant loop.  since we're
in 0.6 let's just simplify.

15 years agoadd a really contrived test that tests the _already_compiling flag. but this seems...
Mike Bayer [Fri, 14 Aug 2009 16:54:49 +0000 (16:54 +0000)] 
add a really contrived test that tests the _already_compiling flag.  but this seems silly
so far.

15 years agolower call counts for 2.4, hurray
Philip Jenvey [Fri, 14 Aug 2009 04:01:54 +0000 (04:01 +0000)] 
lower call counts for 2.4, hurray

15 years agodon't use the deprecated driver name
Philip Jenvey [Fri, 14 Aug 2009 03:04:28 +0000 (03:04 +0000)] 
don't use the deprecated driver name

15 years agoReplace HTML entities with reST markup
Lele Gaifax [Tue, 11 Aug 2009 08:19:37 +0000 (08:19 +0000)] 
Replace HTML entities with reST markup

15 years agomove postgresql's % escape handling out of base
Philip Jenvey [Tue, 11 Aug 2009 05:12:50 +0000 (05:12 +0000)] 
move postgresql's % escape handling out of base

15 years agoo default Connector/J's characterEncoding=UTF-8 for generally better JDBC
Philip Jenvey [Tue, 11 Aug 2009 04:16:48 +0000 (04:16 +0000)] 
o default Connector/J's characterEncoding=UTF-8 for generally better JDBC
unicode handling
o pass url query params down as jdbc connect opts

15 years agosome doc work
Mike Bayer [Mon, 10 Aug 2009 23:06:07 +0000 (23:06 +0000)] 
some doc work

15 years ago0.6
Mike Bayer [Mon, 10 Aug 2009 21:44:14 +0000 (21:44 +0000)] 
0.6

15 years ago- simplify MySQLIdentifierPreparer into standard pattern,
Mike Bayer [Mon, 10 Aug 2009 04:48:00 +0000 (04:48 +0000)] 
- simplify MySQLIdentifierPreparer into standard pattern,
thus allowing easy subclassing
- move % sign logic for MySQLIdentifierPreparer into MySQLdb dialect
- paramterize the escape/unescape quote char in IdentifierPreparer
- cut out MySQLTableDefinitionParser cruft

15 years ago- the Oracle dialect now features NUMBER which intends
Mike Bayer [Sun, 9 Aug 2009 23:46:06 +0000 (23:46 +0000)] 
- the Oracle dialect now features NUMBER which intends
to act justlike Oracle's NUMBER type.  It is the primary
numeric type returned by table reflection and attempts
to return Decimal()/float/int based on the precision/scale
parameters.  [ticket:885]

15 years ago- PG: somewhat better support for % signs in table/column names;
Mike Bayer [Sun, 9 Aug 2009 22:11:40 +0000 (22:11 +0000)] 
- PG: somewhat better support for % signs in table/column names;
psycopg2 can't handle a bind parameter name of
%(foobar)s however and SQLA doesn't want to add overhead
just to treat that one non-existent use case.
[ticket:1279]
- MySQL: somewhat better support for % signs in table/column names;
MySQLdb can't handle % signs in SQL when executemany() is used,
and SQLA doesn't want to add overhead just to treat that one
non-existent use case. [ticket:1279]

15 years agoclose out py3k + pg8000 bugs that are fixable for now without pg8000 decimal fix
Mike Bayer [Sun, 9 Aug 2009 21:41:56 +0000 (21:41 +0000)] 
close out py3k + pg8000 bugs that are fixable for now without pg8000 decimal fix

15 years agopython3k fixes
Mike Bayer [Sun, 9 Aug 2009 20:50:46 +0000 (20:50 +0000)] 
python3k fixes

15 years agopull from identity_map atomically to avoid a race with weakref cleanup
Philip Jenvey [Sun, 9 Aug 2009 20:35:24 +0000 (20:35 +0000)] 
pull from identity_map atomically to avoid a race with weakref cleanup

15 years agoo oracle+zxjdbc type handling additions
Philip Jenvey [Sun, 9 Aug 2009 00:56:52 +0000 (00:56 +0000)] 
o oracle+zxjdbc type handling additions
o avoid returning tests on oracle+zxjdbc for now

15 years agomerge 06CHANGES into CHANGES and add more stuff
Mike Bayer [Sat, 8 Aug 2009 23:10:00 +0000 (23:10 +0000)] 
merge 06CHANGES into CHANGES and add more stuff

15 years ago- added **kw to ClauseElement.compare(), so that we can smarten up the "use_get"...
Mike Bayer [Sat, 8 Aug 2009 22:21:02 +0000 (22:21 +0000)] 
- added **kw to ClauseElement.compare(), so that we can smarten up the "use_get" operation
- many-to-one relation to a joined-table subclass now uses get()
  for a simple load (known as the "use_get" condition),
  i.e. Related->Sub(Base), without the need
  to redefine the primaryjoin condition in terms of the base
  table. [ticket:1186]
- specifying a foreign key with a declarative column,
  i.e. ForeignKey(MyRelatedClass.id) doesn't break the "use_get"
  condition from taking place [ticket:1492]

15 years agoclean up the way we detect MSSQL's form of RETURNING
Mike Bayer [Sat, 8 Aug 2009 17:38:45 +0000 (17:38 +0000)] 
clean up the way we detect MSSQL's form of RETURNING

15 years agounwrapped _get_colparams a bit, dropped out an isinstance() call
Mike Bayer [Sat, 8 Aug 2009 17:24:02 +0000 (17:24 +0000)] 
unwrapped _get_colparams a bit, dropped out an isinstance() call

15 years agofix up oracle tests, returning is on by default
Mike Bayer [Sat, 8 Aug 2009 16:49:28 +0000 (16:49 +0000)] 
fix up oracle tests, returning is on by default

15 years ago- turned on auto-returning for oracle, some errors
Mike Bayer [Sat, 8 Aug 2009 15:26:43 +0000 (15:26 +0000)] 
- turned on auto-returning for oracle, some errors
- added make_transient() [ticket:1052]
- ongoing refactor of compiler _get_colparams()  (more to come)

15 years agoensure order of larger comparisons
Philip Jenvey [Sat, 8 Aug 2009 02:01:46 +0000 (02:01 +0000)] 
ensure order of larger comparisons

15 years agodocumentation updates
Mike Bayer [Fri, 7 Aug 2009 22:17:24 +0000 (22:17 +0000)] 
documentation updates

15 years ago- renamed PASSIVE_NORESULT to PASSIVE_NO_RESULT
Mike Bayer [Fri, 7 Aug 2009 21:14:32 +0000 (21:14 +0000)] 
- renamed PASSIVE_NORESULT to PASSIVE_NO_RESULT
- renamed PASSIVE_NO_CALLABLES to PASSIVE_NO_FETCH
- passive now propagates all the way through lazy callables,
all the way into query._get(), so that many-to-one lazy load
can load the instance via the local session but not trigger
any SQL if not available, fixes [ticket:1298] without
messing up consistency of tests added in r6201
- many-to-one also handles returning PASSIVE_NO_RESULT
for the "old" value thus eliminating the need for the
previous value even if the new value is None
- query._get() uses identity_map.get(), which has been
changed to no longer raise KeyError, thus providing
mythical time savings that didn't seem to make any
difference in how fast the unit tests ran.

15 years agoFix #1451: take into account the actual coding system when determining the field...
Lele Gaifax [Fri, 7 Aug 2009 12:34:10 +0000 (12:34 +0000)] 
Fix #1451: take into account the actual coding system when determining the field length

15 years agoFix #1429: take into account possible spurious spaces around the DEFAULT keyword
Lele Gaifax [Fri, 7 Aug 2009 12:16:15 +0000 (12:16 +0000)] 
Fix #1429: take into account possible spurious spaces around the DEFAULT keyword

15 years agoM-x whitespace-cleanup
Lele Gaifax [Fri, 7 Aug 2009 12:10:33 +0000 (12:10 +0000)] 
M-x whitespace-cleanup

15 years agofix some profiles for 2.4
Mike Bayer [Thu, 6 Aug 2009 22:37:38 +0000 (22:37 +0000)] 
fix some profiles for 2.4

15 years agofix non2.4 gremlin
Mike Bayer [Thu, 6 Aug 2009 22:16:53 +0000 (22:16 +0000)] 
fix non2.4 gremlin

15 years agodont need this anymore
Mike Bayer [Thu, 6 Aug 2009 21:19:20 +0000 (21:19 +0000)] 
dont need this anymore

15 years agomerge 0.6 series to trunk.
Mike Bayer [Thu, 6 Aug 2009 21:11:27 +0000 (21:11 +0000)] 
merge 0.6 series to trunk.

15 years agodisabled examples test pending necesary repairs
Mike Bayer [Thu, 6 Aug 2009 20:16:31 +0000 (20:16 +0000)] 
disabled examples test pending necesary repairs

15 years ago- UPDATE and DELETE do not support ORDER BY, LIMIT, OFFSET,
Mike Bayer [Sun, 2 Aug 2009 18:13:07 +0000 (18:13 +0000)] 
- UPDATE and DELETE do not support ORDER BY, LIMIT, OFFSET,
etc. in standard SQL.  Query.update() and Query.delete()
now raise an exception if any of limit(), offset(),
order_by(), group_by(), or distinct() have been
called. [ticket:1487]

16 years ago- Simplified the sweep of instrumentation in strategies._register_attribute
Mike Bayer [Sun, 2 Aug 2009 17:51:33 +0000 (17:51 +0000)] 
- Simplified the sweep of instrumentation in strategies._register_attribute

- Improved support for MapperProperty objects overriding
that of an inherited mapper for non-concrete
inheritance setups - attribute extensions won't randomly
collide with each other.  [ticket:1488]

- Added AttributeExtension to sqlalchemy.orm.__all__

16 years agobackported 0.6 r6084 fix for oracle alias names, [ticket:1309]
Mike Bayer [Fri, 31 Jul 2009 23:10:46 +0000 (23:10 +0000)] 
backported 0.6 r6084 fix for oracle alias names, [ticket:1309]

16 years agofix broken orm debug logging
Philip Jenvey [Wed, 29 Jul 2009 00:41:26 +0000 (00:41 +0000)] 
fix broken orm debug logging

16 years agomerged [ticket:1486] fix from 0.6
Mike Bayer [Tue, 28 Jul 2009 17:47:54 +0000 (17:47 +0000)] 
merged [ticket:1486] fix from 0.6

16 years agoReverted my screw up of setup.cfg
Michael Trier [Tue, 28 Jul 2009 01:23:17 +0000 (01:23 +0000)] 
Reverted my screw up of setup.cfg

16 years agoCorrected examples tests. I was running from ./test instead of root.
Michael Trier [Tue, 28 Jul 2009 01:20:52 +0000 (01:20 +0000)] 
Corrected examples tests. I was running from ./test instead of root.

16 years agoAdded in Examples into the test suite so they get exercised regularly. Cleaned up...
Michael Trier [Mon, 27 Jul 2009 02:12:15 +0000 (02:12 +0000)] 
Added in Examples into the test suite so they get exercised regularly. Cleaned up some deprecation warnings in the examples.

16 years agoCorrected annoying deprecation warning on 2.6+ related to mssql and the __new__ calls.
Michael Trier [Sun, 26 Jul 2009 03:07:42 +0000 (03:07 +0000)] 
Corrected annoying deprecation warning on 2.6+ related to mssql and the __new__ calls.

16 years agoCorrected problem with binary test on mssql. Still having issues with prepared statem...
Michael Trier [Sun, 26 Jul 2009 03:07:32 +0000 (03:07 +0000)] 
Corrected problem with binary test on mssql. Still having issues with prepared statements.

16 years ago- Squeezed a few more unnecessary "lazy loads" out of
Mike Bayer [Sun, 26 Jul 2009 01:46:41 +0000 (01:46 +0000)] 
- Squeezed a few more unnecessary "lazy loads" out of
relation().  When a collection is mutated, many-to-one
backrefs on the other side will not fire off to load
the "old" value, unless "single_parent=True" is set.
A direct assignment of a many-to-one still loads
the "old" value in order to update backref collections
on that value, which may be present in the session
already, thus maintaining the 0.5 behavioral contract.
[ticket:1483]

16 years ago- Fixed bug in Table and Column whereby passing empty
Mike Bayer [Sat, 25 Jul 2009 21:40:27 +0000 (21:40 +0000)] 
- Fixed bug in Table and Column whereby passing empty
dict for "info" argument would raise an exception.
[ticket:1482]

16 years ago- Fixed bug whereby a load/refresh of joined table
Mike Bayer [Sat, 25 Jul 2009 21:26:28 +0000 (21:26 +0000)] 
- Fixed bug whereby a load/refresh of joined table
inheritance attributes which were based on
column_property() or similar would fail to evaluate.
[ticket:1480]

16 years ago- Declarative will raise an informative exception if
Mike Bayer [Sat, 25 Jul 2009 20:43:11 +0000 (20:43 +0000)] 
- Declarative will raise an informative exception if
__table_args__ is passed as a tuple with no dict argument.
Improved documentation.  [ticket:1468]

16 years ago- fixed the test for FalseDiscriminator to use Boolean for picky postgresql
Mike Bayer [Sat, 25 Jul 2009 20:27:33 +0000 (20:27 +0000)] 
- fixed the test for FalseDiscriminator to use Boolean for picky postgresql
- added Query.enable_assertions(False) as a mediocre solution for [ticket:1424].
updated the recipe at http://www.sqlalchemy.org/trac/wiki/UsageRecipes/PreFilteredQuery to
reflect.
- moved most default Query state to be class level variables to start.  the dicts could
go as well but being overly careful to not place mutables there for the moment.
- a visit by the "dunder-private method names aren't cool" police
- continued undisciplined pep-8ness

16 years ago- Using False or 0 as a polymorphic discriminator now
Mike Bayer [Sat, 25 Jul 2009 19:42:15 +0000 (19:42 +0000)] 
- Using False or 0 as a polymorphic discriminator now
works on the base class as well as a subclass.
[ticket:1440]

16 years ago- Unary expressions such as DISTINCT propagate their
Mike Bayer [Sat, 25 Jul 2009 19:34:02 +0000 (19:34 +0000)] 
- Unary expressions such as DISTINCT propagate their
type handling to result sets, allowing conversions like
unicode and such to take place.  [ticket:1420]

16 years ago- Improved error message when query() is called with
Mike Bayer [Sat, 25 Jul 2009 18:59:56 +0000 (18:59 +0000)] 
- Improved error message when query() is called with
a non-SQL /entity expression. [ticket:1476]

16 years agobeefed up documentation for count(), [ticket:1465]
Mike Bayer [Sat, 25 Jul 2009 18:54:20 +0000 (18:54 +0000)] 
beefed up documentation for count(), [ticket:1465]

16 years ago- The collection proxies produced by associationproxy are now
Mike Bayer [Sat, 25 Jul 2009 17:08:38 +0000 (17:08 +0000)] 
- The collection proxies produced by associationproxy are now
pickleable.  A user-defined proxy_factory however
is still not pickleable unless it defines __getstate__
and __setstate__. [ticket:1446]

16 years agoFix small typos in docstring
Lele Gaifax [Fri, 24 Jul 2009 15:31:09 +0000 (15:31 +0000)] 
Fix small typos in docstring

16 years agoensure "rowswitch" for isdelete is supported
Mike Bayer [Wed, 22 Jul 2009 20:41:33 +0000 (20:41 +0000)] 
ensure "rowswitch" for isdelete is supported

16 years ago- relations() now have greater ability to be "overridden",
Mike Bayer [Tue, 21 Jul 2009 21:47:03 +0000 (21:47 +0000)] 
- relations() now have greater ability to be "overridden",
meaning a subclass that explicitly specifies a relation()
overriding that of the parent class will be honored
during a flush.  This is currently to support
many-to-many relations from concrete inheritance setups.
Outside of that use case, YMMV.  [ticket:1477]

16 years ago- Fixed bug whereby inheritance discriminator part of a
Mike Bayer [Tue, 21 Jul 2009 20:25:36 +0000 (20:25 +0000)] 
- Fixed bug whereby inheritance discriminator part of a
composite primary key would fail on updates.
Continuation of [ticket:1300].