]> git.ipfire.org Git - thirdparty/sqlalchemy/sqlalchemy.git/log
thirdparty/sqlalchemy/sqlalchemy.git
13 years ago- reorganize changes so that the 0.7 changelog is present fully underneath the 0.8
Mike Bayer [Thu, 21 Jun 2012 18:50:25 +0000 (14:50 -0400)] 
- reorganize changes so that the 0.7 changelog is present fully underneath the 0.8

13 years agoa likely callcount
Mike Bayer [Thu, 21 Jun 2012 05:50:59 +0000 (01:50 -0400)] 
a likely callcount

13 years agoneed some default dialect in here
Mike Bayer [Thu, 21 Jun 2012 00:19:53 +0000 (20:19 -0400)] 
need some default dialect in here

13 years agotest didn't get added
Mike Bayer [Wed, 20 Jun 2012 23:46:00 +0000 (19:46 -0400)] 
test didn't get added

13 years agocallcount
Mike Bayer [Wed, 20 Jun 2012 23:39:35 +0000 (19:39 -0400)] 
callcount

13 years ago- [feature] The of_type() construct on attributes
Mike Bayer [Wed, 20 Jun 2012 23:28:29 +0000 (19:28 -0400)] 
- [feature] The of_type() construct on attributes
now accepts aliased() class constructs as well
as with_polymorphic constructs, and works with
query.join(), any(), has(), and also
eager loaders subqueryload(), joinedload(),
contains_eager()
[ticket:2438] [ticket:1106]
- a rewrite of the query path system to use an
object based approach for more succinct usage.  the system
has been designed carefully to not add an excessive method overhead.
- [feature] select() features a correlate_except()
method, auto correlates all selectables except those
passed.   Is needed here for the updated any()/has()
functionality.
- remove some old cruft from LoaderStrategy, init(),debug_callable()
- use a namedtuple for _extended_entity_info.  This method should
become standard within the orm internals
- some tweaks to the memory profile tests, number of runs can
be customized to work around pysqlite's very annoying behavior
- try to simplify PropertyOption._get_paths(), rename to _process_paths(),
returns a single list now.  overall works more completely as was needed
for of_type() functionality

13 years ago- [feature] Dramatic improvement in memory
Mike Bayer [Wed, 20 Jun 2012 22:55:13 +0000 (18:55 -0400)] 
- [feature] Dramatic improvement in memory
usage of the event system; instance-level
collections are no longer created for a
particular type of event until
instance-level listeners are established
for that event.  [ticket:2516] Also in 0.7.9.

13 years agoMerged in msabramo/sqlalchemy (pull request #9)
Mike Bayer [Mon, 18 Jun 2012 14:06:49 +0000 (10:06 -0400)] 
Merged in msabramo/sqlalchemy (pull request #9)

13 years agoremove ambiguous language
Mike Bayer [Mon, 18 Jun 2012 04:08:46 +0000 (00:08 -0400)] 
remove ambiguous language

13 years agoCorrect error in README.unittests; --first-package-wins option does not
Marc Abramowitz [Sun, 17 Jun 2012 14:59:25 +0000 (07:59 -0700)] 
Correct error in README.unittests; --first-package-wins option does not
take a value

~/dev/git-repos/sqlalchemy$ ./sqla_nose.py --first-package-wins=True
Usage: sqla_nose.py [options]

sqla_nose.py: error: --first-package-wins option does not take a value

13 years agosome fixes here
Mike Bayer [Sun, 17 Jun 2012 01:29:08 +0000 (21:29 -0400)] 
some fixes here

13 years agodoc fixes
Mike Bayer [Sun, 17 Jun 2012 00:56:32 +0000 (20:56 -0400)] 
doc fixes

13 years agodbs like oracle can't handle these tests since we aren't applying a sequence
Mike Bayer [Sun, 17 Jun 2012 00:10:30 +0000 (20:10 -0400)] 
dbs like oracle can't handle these tests since we aren't applying a sequence
to the reflected table

13 years agoobnoxious callcounts
Mike Bayer [Sat, 16 Jun 2012 23:55:35 +0000 (19:55 -0400)] 
obnoxious callcounts

13 years agoremove this test now covered in test_execute
Mike Bayer [Sat, 16 Jun 2012 23:14:36 +0000 (19:14 -0400)] 
remove this test now covered in test_execute

13 years ago- [bug] Fixed bug whereby append_column()
Mike Bayer [Sat, 16 Jun 2012 23:02:48 +0000 (19:02 -0400)] 
- [bug] Fixed bug whereby append_column()
wouldn't function correctly on a cloned
select() construct, courtesy
Gunnlaugur Por Briem.  [ticket:2482]
Also in 0.7.8.

13 years ago- [bug] The ResultProxy methods inserted_primary_key,
Mike Bayer [Sat, 16 Jun 2012 22:41:54 +0000 (18:41 -0400)] 
- [bug] The ResultProxy methods inserted_primary_key,
last_updated_params(), last_inserted_params(),
postfetch_cols(), prefetch_cols() all
assert that the given statement is a compiled
construct, and is an insert() or update()
statement as is appropriate, else
raise InvalidRequestError.  [ticket:2498]

- ResultProxy.last_inserted_ids is removed,
replaced by inserted_primary_key.

13 years ago- [bug] removed unnecessary table clause when
Mike Bayer [Sat, 16 Jun 2012 21:53:57 +0000 (17:53 -0400)] 
- [bug] removed unnecessary table clause when
reflecting enums, [ticket:2510].

13 years ago- [bug] Fixed bug whereby populate_existing
Mike Bayer [Sat, 16 Jun 2012 21:40:15 +0000 (17:40 -0400)] 
- [bug] Fixed bug whereby populate_existing
option would not propagate to subquery
eager loaders.  [ticket:2497].  Also
in 0.7.8.

13 years ago- [bug] Repaired common table expression
Mike Bayer [Wed, 13 Jun 2012 22:21:42 +0000 (18:21 -0400)] 
- [bug] Repaired common table expression
rendering to function correctly when the
SELECT statement contains UNION or other
compound expressions, courtesy btbuilder.
[ticket:2490]

13 years ago- [bug] Added BIGINT, BINARY, VARBINARY to
Mike Bayer [Wed, 13 Jun 2012 21:55:10 +0000 (17:55 -0400)] 
- [bug] Added BIGINT, BINARY, VARBINARY to
types.__all__, sqlalchemy namespaces,
plus tests to make sure new types
remain importable.
[ticket:2499]

13 years agodont call connection()/get_bind() all that here if we don't have to
Mike Bayer [Wed, 13 Jun 2012 14:13:17 +0000 (10:13 -0400)] 
dont call connection()/get_bind() all that here if we don't have to

13 years agocallcount fix
Mike Bayer [Wed, 13 Jun 2012 14:12:16 +0000 (10:12 -0400)] 
callcount fix

13 years agoweird, seems like hasattr() doesn't count as a function call in the same way as isins...
Mike Bayer [Tue, 12 Jun 2012 14:32:52 +0000 (10:32 -0400)] 
weird, seems like hasattr() doesn't count as a function call in the same way as isinstance()...

13 years ago- rewrite rowcount documentation to be as absolutely clear as possible
Mike Bayer [Tue, 12 Jun 2012 14:21:58 +0000 (10:21 -0400)] 
- rewrite rowcount documentation to be as absolutely clear as possible

13 years agosigh. go back to __iter__ because of silly rowproxy as argument requirement
Mike Bayer [Tue, 12 Jun 2012 00:46:43 +0000 (20:46 -0400)] 
sigh.  go back to __iter__ because of silly rowproxy as argument requirement

13 years agocomments regarding this
Mike Bayer [Tue, 12 Jun 2012 00:24:16 +0000 (20:24 -0400)] 
comments regarding this

13 years ago- [bug] Fixed bug affecting Py3K whereby
Mike Bayer [Tue, 12 Jun 2012 00:11:05 +0000 (20:11 -0400)] 
- [bug] Fixed bug affecting Py3K whereby
string positional parameters passed to
engine/connection execute() would fail to be
interpreted correctly, due to __iter__
being present on Py3K string.
[ticket:2503].  Also in 0.7.8.

13 years agocallcount
Mike Bayer [Mon, 11 Jun 2012 23:44:58 +0000 (19:44 -0400)] 
callcount

13 years agomake sure integer is here. only the C ext doesn't try to call "key_fallback" for...
Mike Bayer [Mon, 11 Jun 2012 16:15:15 +0000 (12:15 -0400)] 
make sure integer is here.  only the C ext doesn't try to call "key_fallback" for integer here,
pure python one does.

13 years ago- [bug] Fixed identity_key() function which
Mike Bayer [Mon, 11 Jun 2012 14:08:33 +0000 (10:08 -0400)] 
- [bug] Fixed identity_key() function which
was not accepting a scalar argument
for the identity.  [ticket:2508]. Also
in 0.7.8.

13 years agoAdd some `Sphinx` related version informations paragraph-level markups,
Mike Bayer [Fri, 8 Jun 2012 19:56:58 +0000 (15:56 -0400)] 
Add some `Sphinx` related version informations paragraph-level markups,
such as ``.. versionaddedd::``, ``.. versionchanged::`` and ``.. deprecated::``.

13 years agoAdd some `Sphinx` paragraph level versions informations markups,
Mike Bayer [Fri, 8 Jun 2012 19:55:08 +0000 (15:55 -0400)] 
Add some `Sphinx` paragraph level versions informations markups,
such as ``.. versionadded::``, ``.. versionchanged::`` and ``.. deprecated::``.

13 years agoAdd some basic CSS styles for version informations paragraph markups.
Mike Bayer [Fri, 8 Jun 2012 19:55:02 +0000 (15:55 -0400)] 
Add some basic CSS styles for version informations paragraph markups.

13 years agocallcount adjust
Mike Bayer [Fri, 8 Jun 2012 18:43:52 +0000 (14:43 -0400)] 
callcount adjust

13 years agoremove deprecateds
Mike Bayer [Fri, 8 Jun 2012 18:40:45 +0000 (14:40 -0400)] 
remove deprecateds

13 years ago- [feature] The 'objects' argument to
Mike Bayer [Fri, 8 Jun 2012 18:35:24 +0000 (14:35 -0400)] 
- [feature] The 'objects' argument to
flush() is no longer deprecated, as some
valid use cases have been identified.

13 years ago- add before_attach event, [ticket:2464]
Mike Bayer [Fri, 8 Jun 2012 01:25:37 +0000 (21:25 -0400)] 
- add before_attach event, [ticket:2464]

13 years agoalso in 0.7.8
Mike Bayer [Fri, 1 Jun 2012 20:31:10 +0000 (16:31 -0400)] 
also in 0.7.8

13 years ago - [bug] Fixed memory leak in C version of
Mike Bayer [Fri, 1 Jun 2012 20:22:44 +0000 (16:22 -0400)] 
  - [bug] Fixed memory leak in C version of
    result proxy whereby DBAPIs which don't deliver
    pure Python tuples for result rows would
    fail to decrement refcounts correctly.
    The most prominently affected DBAPI
    is pyodbc.  [ticket:2489]

13 years agocredit!
Mike Bayer [Sat, 26 May 2012 17:21:05 +0000 (13:21 -0400)] 
credit!

13 years ago- upgrade DBAPI index to have links to all dialect toplevel pages
Mike Bayer [Sat, 26 May 2012 17:11:26 +0000 (13:11 -0400)] 
- upgrade DBAPI index to have links to all dialect toplevel pages
- add line for google DBAPI

13 years ago- add new google app engine DBAPI, [ticket:2484]
Mike Bayer [Sat, 26 May 2012 17:10:40 +0000 (13:10 -0400)] 
- add new google app engine DBAPI, [ticket:2484]

13 years ago- [bug] Dialect no longer emits expensive server
Mike Bayer [Thu, 24 May 2012 22:49:23 +0000 (18:49 -0400)] 
- [bug] Dialect no longer emits expensive server
collations query, as well as server casing,
on first connect.  These functions are still
available as semi-private. [ticket:2404]

13 years agoget __clause_element__ for remote()/foreign() annotation, [ticket:2493]
Mike Bayer [Thu, 24 May 2012 15:12:39 +0000 (11:12 -0400)] 
get __clause_element__ for remote()/foreign() annotation, [ticket:2493]

13 years ago- identify another condition, indirectly via #2491, where
Mike Bayer [Thu, 24 May 2012 14:58:43 +0000 (10:58 -0400)] 
- identify another condition, indirectly via #2491, where
remote/local detection was regressing vs. 0.7.   use a more
liberal check for "tables overlap".

13 years ago- [bug] Fixed bug in declarative
Mike Bayer [Thu, 24 May 2012 14:17:54 +0000 (10:17 -0400)] 
- [bug] Fixed bug in declarative
whereby the precedence of columns
in a joined-table, composite
column (typically for id) would fail to
be correct if the columns contained
names distinct from their attribute
names.  This would cause things like
primaryjoin conditions made against the
entity attributes to be incorrect.  Related
to [ticket:1892] as this was supposed
to be part of that, this is [ticket:2491].
Also in 0.7.8.

13 years ago add a test for r2491
Mike Bayer [Wed, 23 May 2012 16:23:56 +0000 (12:23 -0400)] 
 add a test for r2491

13 years agolooks like I finally stepped on someone elses dunders...will have to figure out
Mike Bayer [Sun, 20 May 2012 13:40:55 +0000 (09:40 -0400)] 
looks like I finally stepped on someone elses dunders...will have to figure out
how to make a public API for this anyway

13 years agofix logic here broken by the change for enable_relationship_loading
Mike Bayer [Sat, 19 May 2012 23:40:13 +0000 (19:40 -0400)] 
fix logic here broken by the change for enable_relationship_loading

13 years agoignore columns that don't resolve here either.
Mike Bayer [Sat, 19 May 2012 23:33:58 +0000 (19:33 -0400)] 
ignore columns that don't resolve here either.

13 years ago- reorganize the usage of __mapper_args__ so that it's only
Mike Bayer [Fri, 18 May 2012 14:55:28 +0000 (10:55 -0400)] 
- reorganize the usage of __mapper_args__ so that it's only
called after the __prepare__() step, if any, so that everything to
do with the mapping occurs after the table is reflected.

13 years ago- [bug] Added ROWID to oracle.*, [ticket:2483]
Mike Bayer [Thu, 17 May 2012 20:11:15 +0000 (16:11 -0400)] 
- [bug] Added ROWID to oracle.*, [ticket:2483]
Also in 0.7.8.

13 years agoadd a note about sqlites lack of concurrency by design, [ticket:2447]
Mike Bayer [Thu, 17 May 2012 20:03:51 +0000 (16:03 -0400)] 
add a note about sqlites lack of concurrency by design, [ticket:2447]

13 years ago- [feature] Added utility feature
Mike Bayer [Thu, 17 May 2012 19:48:39 +0000 (15:48 -0400)] 
- [feature] Added utility feature
Session.enable_relationship_loading(),
supersedes relationship.load_on_pending.
Both features should be avoided, however.
[ticket:2372]

13 years ago- [bug] Fixed bug whereby subqueryload() from
Mike Bayer [Thu, 17 May 2012 18:18:05 +0000 (14:18 -0400)] 
- [bug] Fixed bug whereby subqueryload() from
a polymorphic mapping to a target would incur
a new invocation of the query for each
distinct class encountered in the polymorphic
result.  [ticket:2480].  Also in 0.7.8.

13 years agofix mysql
Mike Bayer [Thu, 17 May 2012 16:12:05 +0000 (12:12 -0400)] 
fix mysql

13 years ago- [feature] The "deferred declarative
Mike Bayer [Thu, 17 May 2012 15:45:05 +0000 (11:45 -0400)] 
- [feature] The "deferred declarative
reflection" system has been moved into the
declarative extension itself, using the
new DeferredReflection class.  This
class is now tested with both single
and joined table inheritance use cases.
[ticket:2485]
- [bug] The autoload_replace flag on Table,
when False, will cause any reflected foreign key
constraints which refer to already-declared
columns to be skipped, assuming that the
in-Python declared column will take over
the task of specifying in-Python ForeignKey
or ForeignKeyConstraint declarations.

13 years ago- add DeferredReflection to declarative itself
Mike Bayer [Thu, 17 May 2012 14:57:31 +0000 (10:57 -0400)] 
- add DeferredReflection to declarative itself
- split out test_declarative into four separate modules

13 years agoclean this up some more. __mapper__ isn't even set up.
Mike Bayer [Tue, 15 May 2012 16:01:00 +0000 (18:01 +0200)] 
clean this up some more.  __mapper__ isn't even set up.

13 years agoimplement support for __prepare__ declarative method. this signals that
Mike Bayer [Tue, 15 May 2012 15:57:30 +0000 (17:57 +0200)] 
implement support for __prepare__ declarative method.  this signals that
the mapping should not be constructed immediately; instead, the info for
the mapping is placed into an interm object _MapperThingy which serves
as the mapper for the purposes of figuring out inheritance, etc.  Later,
the prepare_deferred_mapping() function can be called given a base which
will invoke __prepare__ for each mapped class.

13 years agoa fix to allow mysql to work
Mike Bayer [Sat, 5 May 2012 03:18:52 +0000 (23:18 -0400)] 
a fix to allow mysql to work

13 years agocallcount bump to work around sqlite's silly column name workaround
Mike Bayer [Sat, 5 May 2012 00:44:12 +0000 (20:44 -0400)] 
callcount bump to work around sqlite's silly column name workaround

13 years ago - [bug] Fixed issue in unit of work
Mike Bayer [Fri, 4 May 2012 23:27:57 +0000 (19:27 -0400)] 
  - [bug] Fixed issue in unit of work
    whereby setting a non-None self-referential
    many-to-one relationship to None
    would fail to persist the change if the
    former value was not already loaded.
    [ticket:2477].

13 years ago- [bug] removed unnecessary table clause
Mike Bayer [Fri, 4 May 2012 22:45:31 +0000 (18:45 -0400)] 
- [bug] removed unnecessary table clause
when reflecting domains, [ticket:2473]
also in 0.7.7

13 years ago - [bug] Quoting information is now passed along
Mike Bayer [Fri, 4 May 2012 22:40:55 +0000 (18:40 -0400)] 
  - [bug] Quoting information is now passed along
    from a Column with quote=True when generating
    a same-named bound parameter to the bindparam()
    object, as is the case in generated INSERT and UPDATE
    statements, so that unknown reserved names can
    be fully supported.  [ticket:2437]

13 years agodoc update regarding with_lockmode, [ticket:2440]
Mike Bayer [Fri, 4 May 2012 21:50:44 +0000 (17:50 -0400)] 
doc update regarding with_lockmode, [ticket:2440]

13 years agodocuemnt mock strategy, [ticket:2439]
Mike Bayer [Fri, 4 May 2012 21:46:06 +0000 (17:46 -0400)] 
docuemnt mock strategy, [ticket:2439]

13 years ago - [feature] Added SQLite execution option
Mike Bayer [Fri, 4 May 2012 21:44:22 +0000 (17:44 -0400)] 
  - [feature] Added SQLite execution option
    "sqlite_raw_colnames=True", will bypass
    attempts to remove "." from column names
    returned by SQLite cursor.description.
    [ticket:2475]

13 years ago - [bug] Add BINARY, VARBINARY to types.__all__,
Mike Bayer [Fri, 4 May 2012 21:37:56 +0000 (17:37 -0400)] 
  - [bug] Add BINARY, VARBINARY to types.__all__,
    [ticket:2474]

13 years agoAdd some documentation sugar for ``having`` criterion.
Priit Laes [Fri, 13 Jan 2012 14:03:41 +0000 (16:03 +0200)] 
Add some documentation sugar for ``having`` criterion.

13 years ago- refactor query.update() and query.delete() to use a pure
Mike Bayer [Sun, 29 Apr 2012 22:53:29 +0000 (18:53 -0400)] 
- refactor query.update() and query.delete() to use a pure
template method pattern using new class hierarchy BulkUD
in sqlalchemy.orm.persistence

13 years agoFix a typo in the documentation
Franck Cuny [Fri, 27 Apr 2012 22:21:02 +0000 (15:21 -0700)] 
Fix a typo in the documentation

s/Attrbutes/Attributes/

13 years agobackport of relationshipproperty comparison bug fix
Mike Bayer [Thu, 26 Apr 2012 15:59:50 +0000 (11:59 -0400)] 
backport of relationshipproperty comparison bug fix

13 years ago- [bug] Fixed bug in relationship comparisons
Mike Bayer [Wed, 25 Apr 2012 14:38:24 +0000 (10:38 -0400)] 
- [bug] Fixed bug in relationship comparisons
whereby calling unimplemented methods like
SomeClass.somerelationship.like() would
produce a recursion overflow, instead
of NotImplementedError.

13 years agoanother pypy skip
Mike Bayer [Wed, 25 Apr 2012 03:40:00 +0000 (23:40 -0400)] 
another pypy skip

13 years agocallcount
Mike Bayer [Wed, 25 Apr 2012 02:17:53 +0000 (22:17 -0400)] 
callcount

13 years agodocument the remote/foreign/remote_foreign functions
Mike Bayer [Wed, 25 Apr 2012 02:11:15 +0000 (22:11 -0400)] 
document the remote/foreign/remote_foreign functions

13 years agoadjust some test exclusions
Mike Bayer [Wed, 25 Apr 2012 00:44:21 +0000 (20:44 -0400)] 
adjust some test exclusions

13 years agofix 2.5 failure
Mike Bayer [Tue, 24 Apr 2012 23:55:43 +0000 (19:55 -0400)] 
fix 2.5 failure

13 years agotop level version-specific links
Mike Bayer [Tue, 24 Apr 2012 23:32:30 +0000 (19:32 -0400)] 
top level version-specific links

13 years agopy3k fixes
Mike Bayer [Tue, 24 Apr 2012 23:15:00 +0000 (19:15 -0400)] 
py3k fixes

13 years ago- put the correct target here in the log message
Mike Bayer [Tue, 24 Apr 2012 22:16:52 +0000 (18:16 -0400)] 
- put the correct target here in the log message

13 years ago- [feature] Calling rollback() within a
Mike Bayer [Tue, 24 Apr 2012 22:06:27 +0000 (18:06 -0400)] 
- [feature] Calling rollback() within a
session.begin_nested() will now only expire
those objects that had net changes within the
scope of that transaction, that is objects which
were dirty or were modified on a flush.  This
allows the typical use case for begin_nested(),
that of altering a small subset of objects, to
leave in place the data from the larger enclosing
set of objects that weren't modified in
that sub-transaction.  [ticket:2452]
- inline the "register_newly_XYZ" functions to operate
upon collections to reduce method calls

13 years ago- [feature] postgresql.ARRAY features an optional
Mike Bayer [Tue, 24 Apr 2012 20:44:53 +0000 (16:44 -0400)] 
- [feature] postgresql.ARRAY features an optional
"dimension" argument, will assign a specific
number of dimensions to the array which will
render in DDL as ARRAY[][]..., also improves
performance of bind/result processing.
[ticket:2441]

13 years ago- [feature] The "unconsumed column names" warning emitted
Mike Bayer [Tue, 24 Apr 2012 20:27:58 +0000 (16:27 -0400)] 
- [feature] The "unconsumed column names" warning emitted
when keys are present in insert.values() or update.values()
that aren't in the target table is now an exception.
[ticket:2415]

13 years ago- [feature] Inspector.get_primary_keys() is
Mike Bayer [Tue, 24 Apr 2012 20:25:20 +0000 (16:25 -0400)] 
- [feature] Inspector.get_primary_keys() is
deprecated; use Inspector.get_pk_constraint().
Courtesy Diana Clarke.  [ticket:2422]
- restored default get_primary_keys()/get_pk_constraint() wrapper
to help maintain compatibility with third party dialects
created against 0.6 or 0.7

13 years ago- [bug] All of UniqueConstraint, ForeignKeyConstraint,
Mike Bayer [Tue, 24 Apr 2012 20:08:35 +0000 (16:08 -0400)] 
- [bug] All of UniqueConstraint, ForeignKeyConstraint,
CheckConstraint, and PrimaryKeyConstraint will
attach themselves to their parent table automatically
when they refer to a Table-bound Column object directly
(i.e. not just string column name), and refer to
one and only one Table.   Prior to 0.8 this behavior
occurred for UniqueConstraint and PrimaryKeyConstraint,
but not ForeignKeyConstraint or CheckConstraint.
[ticket:2410]

13 years agofix import here
Mike Bayer [Tue, 24 Apr 2012 20:03:51 +0000 (16:03 -0400)] 
fix import here

13 years ago- [feature] The behavior of column targeting
Mike Bayer [Tue, 24 Apr 2012 20:03:00 +0000 (16:03 -0400)] 
- [feature] The behavior of column targeting
in result sets is now case sensitive by
default.   SQLAlchemy for many years would
run a case-insensitive conversion on these values,
probably to alleviate early case sensitivity
issues with dialects like Oracle and
Firebird.   These issues have been more cleanly
solved in more modern versions so the performance
hit of calling lower() on identifiers is removed.
The case insensitive comparisons can be re-enabled
by setting "case_insensitive=False" on
create_engine().  [ticket:2423]

13 years ago - [feature] The after_attach event is now
Mike Bayer [Tue, 24 Apr 2012 19:52:09 +0000 (15:52 -0400)] 
  - [feature] The after_attach event is now
    emitted after the object is established
    in Session.new or Session.identity_map
    upon Session.add(), Session.merge(),
    etc., so that the object is represented
    in these collections when the event
    is called. [ticket:2464]

13 years ago - [feature] the SQLite date and time types
Mike Bayer [Tue, 24 Apr 2012 19:49:52 +0000 (15:49 -0400)] 
  - [feature] the SQLite date and time types
    have been overhauled to support a more open
    ended format for input and output, using
    name based format strings and regexps.  A
    new argument "microseconds" also provides
    the option to omit the "microseconds"
    portion of timestamps.  Thanks to
    Nathan Wright for the work and tests on
    this.  [ticket:2363]

13 years agodeprecate inspector.get_primary_keys() in favor of inspector.get_pk_constraint()
Diana Clarke [Tue, 3 Apr 2012 01:07:37 +0000 (21:07 -0400)] 
deprecate inspector.get_primary_keys() in favor of inspector.get_pk_constraint()
- see #2422

13 years ago- sql
Mike Bayer [Sun, 1 Apr 2012 23:42:54 +0000 (19:42 -0400)] 
- sql
  - [bug] Removed warning when Index is created
    with no columns; while this might not be what
    the user intended, it is a valid use case
    as an Index could be a placeholder for just an
    index of a certain name.

- mssql
  - [feature] Added interim create_engine flag
    supports_unicode_binds to PyODBC dialect,
    to force whether or not the dialect
    passes Python unicode literals to PyODBC
    or not.

13 years ago- [bug] Fixed bug which would prevent
Mike Bayer [Sun, 1 Apr 2012 15:01:56 +0000 (11:01 -0400)] 
- [bug] Fixed bug which would prevent
OrderingList from being pickleable
[ticket:2454].  Courtesy Jeff Dairiki

13 years ago- [bug] Fixed bug in expression annotation
Mike Bayer [Sat, 31 Mar 2012 17:35:05 +0000 (13:35 -0400)] 
- [bug] Fixed bug in expression annotation
mechanics which could lead to incorrect
rendering of SELECT statements with aliases
and joins, particularly when using
column_property().  [ticket:2453]

13 years ago - [bug] Fixed bug whereby polymorphic_on
Mike Bayer [Sat, 31 Mar 2012 16:55:42 +0000 (12:55 -0400)] 
  - [bug] Fixed bug whereby polymorphic_on
    column that's not otherwise mapped on the
    class would be incorrectly included
    in a merge() operation, raising an error.
    [ticket:2449]

13 years ago- [feature] Added new for_update/with_lockmode()
Mike Bayer [Thu, 22 Mar 2012 02:58:55 +0000 (22:58 -0400)] 
- [feature] Added new for_update/with_lockmode()
options for Postgresql: for_update="read"/
with_lockmode("read"),
for_update="read_nowait"/
with_lockmode("read_nowait").
These emit "FOR SHARE" and "FOR SHARE NOWAIT",
respectively.  Courtesy Diana Clarke
[ticket:2445]

13 years ago- 0.7.7
Mike Bayer [Wed, 21 Mar 2012 00:12:26 +0000 (20:12 -0400)] 
- 0.7.7
- [feature] Added prefix_with() method
to Query, calls upon select().prefix_with()
to allow placement of MySQL SELECT
directives in statements.  Courtesy
Diana Clarke [ticket:2443]

13 years agoI'm pretty sure this should be entities
Diana Clarke [Sun, 18 Mar 2012 04:05:30 +0000 (00:05 -0400)] 
I'm pretty sure this should be entities