9029622920
Improve _test_get_unique_constraints()
...
Call eq_() in a loop instead of comparing two lists
directly. This makes it much easier to find out which
element is not equal to the reference value.
2013-06-23 10:35:46 +03:00
eba9d1b58f
Provided a new attribute for :class:.TypeDecorator
...
called :attr:`.TypeDecorator.coerce_to_is_types`,
to make it easier to control how comparisons using
``==`` or ``!=`` to ``None`` and boolean types goes
about producing an ``IS`` expression, or a plain
equality expression with a bound parameter.
[ticket:2744]
2013-06-22 12:19:41 -04:00
38b5495511
versionadds
2013-06-22 11:48:33 -04:00
2f2115cf0c
0.9 changelog
2013-06-22 11:40:49 -04:00
62dc627219
- 0.8 changelog
...
- some whitespace
2013-06-22 11:40:32 -04:00
db24d2e813
this locale is not needed. maybe it is somewhere but the locale I have ("C", whatever that means,
...
cannot find meaning of this anywhere in Postgresql documentation) seems to work
2013-06-22 11:25:41 -04:00
29fa6913be
Merge pull request #5 from cjw296/pg-ranges
...
Support for Postgres range types.
2013-06-22 07:47:02 -07:00
a2a570e9f4
make tests pass
2013-06-20 16:42:19 +05:30
27d9972829
check if compostite_class is of class type
...
before checking if it is a subclass of `MutableComposite`
2013-06-19 16:03:23 +05:30
8c555f24b1
Merged in goodscloud/sqlalchemy (pull request #3 )
...
Forgot final check that a delete in a backref also doesn't create a new version
2013-06-18 14:03:09 -04:00
4517a854bc
fix up the isolation level docs which were a mess
2013-06-17 23:12:04 -04:00
dde15ec545
remove double methods
2013-06-17 19:48:49 -04:00
247c0471c7
turn this requirement back down as the simpler serialize works better now
2013-06-17 19:46:07 -04:00
7f82168cb6
- rework PropComparator.adapted() to be PropComparator.adapt_to_entity(),
...
passes in AliasedInsp and allows more flexibility.
- rework the AliasedClass/AliasedInsp relationship so that AliasedInsp has
all state and functionality. AliasedClass is just a facade.
[ticket:2756]
2013-06-17 19:42:43 -04:00
42d58f8b6e
migration for [ticket:2751]
2013-06-17 17:48:10 -04:00
f499b7b883
Forgot final check that a delete in a backref also doesn't create a new version
...
follow up to https://bitbucket.org/zzzeek/sqlalchemy/pull-request/2
2013-06-17 23:30:07 +02:00
7ff503855f
changelog for 0.8
2013-06-17 16:50:26 -04:00
dec742c8c1
changelog
2013-06-17 16:47:16 -04:00
83090f35d7
Merge branch 'master' of bitbucket.org:zzzeek/sqlalchemy
2013-06-17 16:06:50 -04:00
6b5192395a
Merged in goodscloud/sqlalchemy (pull request #2 )
...
don't create a history entry when an object in a backref has changed
2013-06-17 16:06:29 -04:00
0e83f757a8
can remove this, issue is fixed
2013-06-17 16:05:11 -04:00
07e6161c6b
- clean up this test (really we don't even need this, it's not testing much)
...
- for the moment, put a catch in it to see if we can trap that issue
on jenkins
2013-06-16 23:09:31 -04:00
d826d99544
- changelog
...
- docs
2013-06-15 22:58:04 -04:00
9612e2fbb5
Merge pull request #7 from malor/psycopg2_execution_options
...
Add AUTOCOMMIT isolation level support for psycopg2
2013-06-15 19:52:00 -07:00
b92007da7d
Fixed bug in polymorphic SQL generation where multiple joined-inheritance
...
entities against the same base class joined to each other as well
would not track columns on the base table independently of each other if
the string of joins were more than two entities long. Also in 0.8.2.
[ticket:2759]
2013-06-15 15:20:48 -04:00
d6c60cb2f3
don't create a history entry when an object in a backref has changed
...
The code that determines whether an object in a relation has been added/removed does not take into account that that relation may be a backref. If the relation is a backref, then nothing on the current table is changing, and therefore no history entry should be created.
2013-06-15 18:15:29 +02:00
a1ceae2ed3
Add AUTOCOMMIT isolation level support for psycopg2
...
One can use this to emit statements, which can not be
executed within a transaction (e. g. CREATE DATABASE):
from sqlalchemy import create_engine
eng = create_engine('postgresql://test:test@localhost/test')
conn = eng.connect().execution_options(isolation_level='AUTOCOMMIT')
conn.execute('CREATE DATABASE test2;')
Fixes issue #2072 .
2013-06-15 16:54:50 +03:00
9d3e220636
- skip this test for py3k. serializer kind of a bust
2013-06-14 18:34:01 -04:00
02ec9afb15
changelog for this
2013-06-10 20:10:06 -04:00
ef03ecf606
changelog
2013-06-10 19:42:37 -04:00
bd1f232063
Merge pull request #6 from eevee/propagate-reset-on-return
...
Preserve reset_on_return when recreating a Pool.
2013-06-10 16:38:48 -07:00
23ce1d8b12
Preserve reset_on_return when recreating a Pool.
2013-06-10 15:04:39 -07:00
58c8c4ce77
Fixed bug where sending a composite attribute into :meth:.Query.order_by
...
would produce a parenthesized expression not accepted by some databases.
[ticket:2754]
2013-06-10 13:50:06 -04:00
b614a24c5d
Fixed the interaction between composite attributes and
...
the :func:`.aliased` function. Previously, composite attributes
wouldn't work correctly in comparison operations when aliasing
was applied. Also in 0.8.2. [ticket:2755]
2013-06-10 13:45:19 -04:00
b2da12e070
Documentation for the new range type support.
2013-06-10 13:24:02 +01:00
b2ea2eef5d
Implement EXCLUDE constraints for postgres.
2013-06-10 12:09:56 +01:00
f4020282b7
add support for range operators listed in http://www.postgresql.org/docs/9.2/interactive/functions-range.html
2013-06-10 12:09:55 +01:00
70edfa229c
Basic type support for the new range types in postgres 9.2
2013-06-10 12:09:55 +01:00
1652491cc6
this comment is ancient
2013-06-09 20:36:37 -04:00
c0ff3eb9d7
these notes about **kw are incorrect, we are talking about the return value here
2013-06-09 17:56:31 -04:00
258d2a8315
- get_unique_constraints() pull request
...
- version (0.9 for now)
- changelog
- move the test into the test suite so that all dialects can take advantage of it
2013-06-09 17:54:11 -04:00
decf47a770
Merge pull request #4 from malor/reflect_constraints
...
Add basic support of unique constraints reflection
2013-06-09 14:39:30 -07:00
c69fe4acf8
Add basic support of unique constraints reflection
...
Inspection API already supports reflection of table
indexes information and those also include unique
constraints (at least for PostgreSQL and MySQL).
But it could be actually useful to distinguish between
indexes and plain unique constraints (though both are
implemented in the same way internally in RDBMS).
This change adds a new method to Inspection API - get_unique_constraints()
and implements it for SQLite, PostgreSQL and MySQL dialects.
2013-06-09 23:49:55 +03:00
65f5887e9a
Added pool logging for "rollback-on-return" and the less used
...
"commit-on-return". This is enabled with the rest of pool
"debug" logging.
[ticket:2752]
2013-06-09 16:16:53 -04:00
f65ddee93a
...and...its a behavioral improvement
2013-06-08 19:50:40 -04:00
7ca321155a
fix this example
2013-06-08 19:48:10 -04:00
5a1d483b33
mention right joins, make it clear we're talking about left outer join as the bigger issue
2013-06-08 19:23:43 -04:00
10a672e605
oracle 9 does support this, oracle 8 is a special case.
2013-06-08 19:11:06 -04:00
a144f55e65
cleanup
2013-06-08 19:07:52 -04:00
85368d25ed
get nested joins to render on oracle 8
2013-06-08 18:54:14 -04:00
Roman Podolyaka