]> git.ipfire.org Git - thirdparty/sqlalchemy/alembic.git/log
thirdparty/sqlalchemy/alembic.git
11 years ago- This releases' "autogenerate index detection" bug, when a MySQL table
Mike Bayer [Fri, 2 May 2014 20:04:43 +0000 (16:04 -0400)] 
- This releases' "autogenerate index detection" bug, when a MySQL table
includes an Index with the same name as a column, autogenerate reported
it as an "add" even though its not; this is because we ignore reflected
indexes of this nature due to MySQL creating them implicitly.  Indexes
that are named the same as a column are now ignored on
MySQL if we see that the backend is reporting that it already exists;
this indicates that we can still detect additions of these indexes
but not drops, as we cannot distinguish a backend index same-named
as the column as one that is user generated or mysql-generated.
fixes #202

11 years ago- index autogen is a big enough deal to warrant its own test suite
Mike Bayer [Fri, 2 May 2014 19:52:01 +0000 (15:52 -0400)] 
- index autogen is a big enough deal to warrant its own test suite

11 years ago- Added new feature :paramref:`.EnvironmentContext.configure.transaction_per_migration`,
Mike Bayer [Fri, 2 May 2014 19:46:00 +0000 (15:46 -0400)] 
- Added new feature :paramref:`.EnvironmentContext.configure.transaction_per_migration`,
which when True causes the BEGIN/COMMIT pair to incur for each migration
individually, rather than for the whole series of migrations.  This is
to assist with some database directives that need to be within individual
transactions, without the need to disable transactional DDL entirely.
fixes #201

11 years ago- Fixed bug where the ``include_object()`` filter would not receive
Mike Bayer [Wed, 30 Apr 2014 22:28:49 +0000 (18:28 -0400)] 
- Fixed bug where the ``include_object()`` filter would not receive
the original :class:`.Column` object when evaluating a database-only
column to be dropped; the object would not include the parent
:class:`.Table` nor other aspects of the column that are important
for generating the "downgrade" case where the column is recreated.
fixes #200

11 years ago- Fixed bug where :meth:`.EnvironmentContext.get_x_argument`
Mike Bayer [Fri, 4 Apr 2014 15:09:34 +0000 (11:09 -0400)] 
- Fixed bug where :meth:`.EnvironmentContext.get_x_argument`
would fail if the :class:`.Config` in use didn't actually
originate from a command line call. fixes #195

11 years agoFixed another bug regarding naming conventions, continuing
Mike Bayer [Tue, 1 Apr 2014 17:12:51 +0000 (13:12 -0400)] 
Fixed another bug regarding naming conventions, continuing
from :ticket:`183`, where add_index()
drop_index() directives would not correctly render the ``f()``
construct when the index contained a convention-driven name.
fixes #194 re: #183

11 years ago0.6.4 rel_0_6_4
Mike Bayer [Fri, 28 Mar 2014 23:39:49 +0000 (19:39 -0400)] 
0.6.4

11 years agouse correct __mro__ for new mssql _Exec classes, fix #192
Mike Bayer [Fri, 28 Mar 2014 18:58:45 +0000 (14:58 -0400)] 
use correct __mro__ for new mssql _Exec classes, fix #192

11 years ago- Added quoting to the table name when the special EXEC is run to
Mike Bayer [Tue, 25 Mar 2014 23:16:15 +0000 (19:16 -0400)] 
- Added quoting to the table name when the special EXEC is run to
drop any existing server defaults or constraints when the
:paramref:`.drop_column.mssql_drop_check` or
:paramref:`.drop_column.mssql_drop_default`
arguments are used.  fix #186

11 years ago- add an additional fix to the MySQL fixes for re: #103.
Mike Bayer [Sat, 22 Mar 2014 20:21:00 +0000 (16:21 -0400)] 
- add an additional fix to the MySQL fixes for re: #103.

11 years ago- use compat version of mock for py3
Mike Bayer [Mon, 17 Mar 2014 01:54:23 +0000 (21:54 -0400)] 
- use compat version of mock for py3

11 years ago- Added/fixed support for MySQL "SET DEFAULT" / "DROP DEFAULT" phrases,
Mike Bayer [Sun, 16 Mar 2014 23:26:21 +0000 (19:26 -0400)] 
- Added/fixed support for MySQL "SET DEFAULT" / "DROP DEFAULT" phrases,
which will now be rendered if only the server default is changing
or being dropped (e.g. specify None to alter_column() to indicate
"DROP DEFAULT").  Also added support for rendering MODIFY rather than
CHANGE when the column name isn't changing.
fixes #103

11 years ago- add test for drop default
Mike Bayer [Sun, 16 Mar 2014 22:59:26 +0000 (18:59 -0400)] 
- add test for drop default

11 years ago- Added support for the ``initially``, ``match`` keyword arguments
Mike Bayer [Fri, 14 Mar 2014 20:41:21 +0000 (16:41 -0400)] 
- Added support for the ``initially``, ``match`` keyword arguments
as well as dialect-specific keyword arguments to
:meth:`.Operations.create_foreign_key`.
fixes #190

11 years ago- Altered the support for "sourceless" migration files (e.g. only
Mike Bayer [Fri, 14 Mar 2014 14:47:55 +0000 (10:47 -0400)] 
- Altered the support for "sourceless" migration files (e.g. only
.pyc or .pyo present) so that the flag "sourceless=true" needs to
be in alembic.ini for this behavior to take effect.
fixes #163

11 years ago- ensure that target_metadata is also present in the offline migration context,
Mike Bayer [Fri, 14 Mar 2014 14:17:37 +0000 (10:17 -0400)] 
- ensure that target_metadata is also present in the offline migration context,
so that --sql mode gets access to naming conventions also.
existing environments need to add target_metadata to the offline migration
context manually.
fixes #189

11 years agoThe feature that keeps on giving, index/unique constraint autogenerate
Mike Bayer [Thu, 13 Mar 2014 17:52:11 +0000 (13:52 -0400)] 
The feature that keeps on giving, index/unique constraint autogenerate
detection, has even more fixes, this time to accommodate database dialects
that both don't yet report on unique constraints, but the backend
does report unique constraints as indexes.   The logic
Alembic uses to distinguish between "this is an index!" vs.
"this is a unique constraint that is also reported as an index!" has now
been further enhanced to not produce unwanted migrations when the dialect
is observed to not yet implement get_unique_constraints() (e.g. mssql).
Note that such a backend will no longer report index drops for unique
indexes, as these cannot be distinguished from an unreported unique
index.  fixes #185

11 years ago Extensive changes have been made to more fully support SQLAlchemy's new
Mike Bayer [Wed, 12 Mar 2014 21:34:01 +0000 (17:34 -0400)] 
  Extensive changes have been made to more fully support SQLAlchemy's new
      naming conventions feature.  Note that while SQLAlchemy has added this
      feature as of 0.9.2, some additional fixes in 0.9.4 are needed to
      resolve some of the issues:

      1. The :class:`.Operations` object now takes into account the naming
         conventions that are present on the :class:`.MetaData` object that's
         associated using :paramref:`~.EnvironmentContext.configure.target_metadata`.
         When :class:`.Operations` renders a constraint directive like
         ``ADD CONSTRAINT``, it now will make use of this naming convention
         when it produces its own temporary :class:`.MetaData` object.

      2. Note however that the autogenerate feature in most cases generates
         constraints like foreign keys and unique constraints with the
         final names intact; the only exception are the constraints implicit
         with a schema-type like Boolean or Enum.  In most of these cases,
         the naming convention feature will not take effect for these constraints
         and will instead use the given name as is, with one exception....

      3. Naming conventions which use the ``"%(constraint_name)s"`` token, that
         is, produce a new name that uses the original name as a component,
         will still be pulled into the naming convention converter and be
         converted.  The problem arises when autogenerate renders a constraint
         with it's already-generated name present in the migration file's source
         code, the name will be doubled up at render time due to the combination
         of #1 and #2.  So to work around this, autogenerate now renders these
         already-tokenized names using the new :meth:`.Operations.f` component.
         This component is only generated if **SQLAlchemy 0.9.4** or greater
         is in use.

      Therefore it is highly recommended that an upgrade to Alembic 0.6.4
      be accompanied by an upgrade of SQLAlchemy 0.9.4, if the new naming
      conventions feature is used.

fixes #183

11 years ago- Suppressed IOErrors which can raise when program output pipe
Mike Bayer [Tue, 11 Mar 2014 00:30:03 +0000 (20:30 -0400)] 
- Suppressed IOErrors which can raise when program output pipe
is closed under a program like ``head``; however this only
works on Python 2.  On Python 3, there is not yet a known way to
suppress the BrokenPipeError warnings without prematurely terminating
the program via signals. fixes #160.
Added comments to http://bugs.python.org/issue11380 to see what the
status is on py3k.

11 years agogot the docs totally wrong on that
Mike Bayer [Sat, 8 Mar 2014 05:19:18 +0000 (00:19 -0500)] 
got the docs totally wrong on that

11 years ago- Fixed bug where :meth:`.Operations.bulk_insert` would not function
Mike Bayer [Sat, 8 Mar 2014 05:15:46 +0000 (00:15 -0500)] 
- Fixed bug where :meth:`.Operations.bulk_insert` would not function
properly when :meth:`.Operations.inline_literal` values were used,
either in --sql or non-sql mode.    The values will now render
directly in --sql mode.  For compatibility with "online" mode,
a new flag :paramref:`~.Operations.inline_literal.multiparams`
can be set to False which will cause each parameter set to be
compiled and executed with individual INSERT statements.
fixes #179

11 years ago- add new section, cookbook. First example, add conditonal sections to migrations
Mike Bayer [Tue, 4 Mar 2014 23:24:15 +0000 (18:24 -0500)] 
- add new section, cookbook.  First example, add conditonal sections to migrations
within script.py.mako.   fixes #177

11 years ago- add py.test config, see if jenkins no longer pulls in sqlalchemy conftest files...
Mike Bayer [Tue, 4 Mar 2014 06:26:45 +0000 (01:26 -0500)] 
- add py.test config, see if jenkins no longer pulls in sqlalchemy conftest files...

11 years ago- Fixed a failure of the system that allows "legacy keyword arguments"
Mike Bayer [Wed, 19 Feb 2014 01:44:16 +0000 (20:44 -0500)] 
- Fixed a failure of the system that allows "legacy keyword arguments"
to be understood, which arose as of a change in Python 3.4 regarding
decorators.  A workaround is applied that allows the code to work
across Python 3 versions. #175

11 years ago- 0.6.4
Mike Bayer [Mon, 17 Feb 2014 20:58:16 +0000 (15:58 -0500)] 
- 0.6.4
- changelog for pullreq bitbucket:20

11 years agoReturn the script object from command.revision
Robbie Coomber [Mon, 17 Feb 2014 15:17:05 +0000 (15:17 +0000)] 
Return the script object from command.revision

11 years agoRemoved unreachable raise statement
Jayson Reis [Thu, 13 Feb 2014 16:32:29 +0000 (16:32 +0000)] 
Removed unreachable raise statement

11 years ago- remove this rel_0_6_3
Mike Bayer [Mon, 3 Feb 2014 00:55:05 +0000 (19:55 -0500)] 
- remove this

11 years ago- 0.6.3
Mike Bayer [Mon, 3 Feb 2014 00:52:04 +0000 (19:52 -0500)] 
- 0.6.3
- 2014

11 years ago- add new section "the importance of naming constraints"
Mike Bayer [Sun, 2 Feb 2014 22:48:15 +0000 (17:48 -0500)] 
- add new section "the importance of naming constraints"
- add a note that a new branching approach is planned, remove note about
"splice"

11 years ago- Added a workaround for when we call ``fcntl.ioctl()`` to get at
Mike Bayer [Sun, 2 Feb 2014 21:42:29 +0000 (16:42 -0500)] 
- Added a workaround for when we call ``fcntl.ioctl()`` to get at
``TERMWIDTH``; if the function returns zero, as is reported to occur
in some pseudo-ttys, the message wrapping system is disabled in the
same way as if ``ioctl()`` failed. #172

11 years ago - add new tests to ensure the naming convention thing works on this end
Mike Bayer [Sun, 2 Feb 2014 16:33:55 +0000 (11:33 -0500)] 
 - add new tests to ensure the naming convention thing works on this end
as expected

11 years agomore doc fixes and updates
Mike Bayer [Sat, 1 Feb 2014 17:56:00 +0000 (12:56 -0500)] 
more doc fixes and updates

11 years agomove this seealso
Mike Bayer [Sat, 1 Feb 2014 17:37:39 +0000 (12:37 -0500)] 
move this seealso

11 years ago- use latest extension versions, trying to get rtd to build
Mike Bayer [Sat, 1 Feb 2014 17:32:56 +0000 (12:32 -0500)] 
- use latest extension versions, trying to get rtd to build

11 years ago- Added new argument
Mike Bayer [Sat, 1 Feb 2014 17:29:03 +0000 (12:29 -0500)] 
- Added new argument
:paramref:`.EnvironmentContext.configure.user_module_prefix`.
This prefix is applied when autogenerate renders a user-defined type,
which here is defined as any type that is from a module outside of the
``sqlalchemy.`` hierarchy.   This prefix defaults to ``None``, in
which case the :paramref:`.EnvironmentContext.configure.sqlalchemy_module_prefix`
is used, thus preserving the current behavior. #171
- added new documentation sections regarding rendering of types
- improved CSS so that deprecations/version changes are highlighted
- cleanup of some configure paramter docs

11 years ago- Added support for autogenerate covering the use case where :class:`.Table`
Mike Bayer [Fri, 31 Jan 2014 01:29:29 +0000 (20:29 -0500)] 
- Added support for autogenerate covering the use case where :class:`.Table`
objects specified in the metadata have an explicit ``schema`` attribute
whose name matches that of the connection's default schema
(e.g. "public" for Postgresql).  Previously, it was assumed that "schema"
was ``None`` when it matched the "default" schema, now the comparison
adjusts for this. #170

11 years ago- changelog
Mike Bayer [Tue, 28 Jan 2014 17:31:08 +0000 (12:31 -0500)] 
- changelog
- reorg tests a bit
- add test coverage for "include_schemas" in compare_metadata

11 years agoMerge pull request #9 from malor/compare_metadata_include_object
mike bayer [Tue, 28 Jan 2014 17:15:15 +0000 (09:15 -0800)] 
Merge pull request #9 from malor/compare_metadata_include_object

Handle include_* arguments in compare_metadata()

11 years agoHandle include_* arguments in compare_metadata()
Roman Podoliaka [Tue, 28 Jan 2014 07:08:58 +0000 (09:08 +0200)] 
Handle include_* arguments in compare_metadata()

include_object, include_symbol and include_schemas are very useful,
when you need to specify what objects you want to be compared.
Modify compare_metadata() public API function, so that it takes those
arguments into account and pass them to _produce_net_changes().

11 years ago- Calling :func:`.bulk_insert` with an empty list will not emit any
Mike Bayer [Fri, 24 Jan 2014 15:43:43 +0000 (10:43 -0500)] 
- Calling :func:`.bulk_insert` with an empty list will not emit any
commands on the current connection.  This was already the case with
``--sql`` mode, so is now the case with "online" mode.

11 years ago- changelog for pullreq 17
Mike Bayer [Sun, 19 Jan 2014 23:33:45 +0000 (18:33 -0500)] 
- changelog for pullreq 17
- unit tests for schema-level unique/index detection
- cleanup a bit for the get indexes/uniques if only table present

11 years ago- fix up the "build" paths
Mike Bayer [Sun, 19 Jan 2014 23:33:28 +0000 (18:33 -0500)] 
- fix up the "build" paths

11 years agoMerge https://bitbucket.org/dtheodor/alembic/branch/master into m
Mike Bayer [Sun, 19 Jan 2014 23:10:13 +0000 (18:10 -0500)] 
Merge https://bitbucket.org/dtheodor/alembic/branch/master into m

11 years ago- use a valid argument name here as SQLA 0.9.2 will now start validating these
Mike Bayer [Sun, 19 Jan 2014 21:35:18 +0000 (16:35 -0500)] 
- use a valid argument name here as SQLA 0.9.2 will now start validating these

11 years agoAutogenerate index and unique's comparison takes schema into account
Dimitris Theodorou [Sun, 19 Jan 2014 14:49:14 +0000 (15:49 +0100)] 
Autogenerate index and unique's comparison takes schema into account

The inspector now will recover the indexes and unique constraints of
the table in the correct schema. Previously it was not taking schema
into account, resulting in always re-creating the existing uniques
and indexes.

11 years agoSchema fixes on rendering of alembic op. directives
Dimitris Theodorou [Sun, 19 Jan 2014 14:40:03 +0000 (15:40 +0100)] 
Schema fixes on rendering of alembic op. directives

Added schema parameter in _drop_index, _drop_constraint
Removed prepending of the schema name in _add_index table
parameter.

11 years ago- More fixes to index autodetection; indexes created with expressions
Mike Bayer [Sun, 12 Jan 2014 02:55:17 +0000 (21:55 -0500)] 
- More fixes to index autodetection; indexes created with expressions
like DESC or functional indexes will no longer cause AttributeError
exceptions when attempting to compare the columns. #164

11 years agoMerged in vrutsky/alembic/vrutsky/remove-space-before-exclamation-mark-wow-1389275979...
Mike Bayer [Thu, 9 Jan 2014 19:33:54 +0000 (14:33 -0500)] 
Merged in vrutsky/alembic/vrutsky/remove-space-before-exclamation-mark-wow-1389275979894 (pull request #16)

Remove space before exclamation mark: "Wow that rocked !"

11 years agoRemove space before exclamation mark: "Wow that rocked !"
vrutsky [Thu, 9 Jan 2014 14:02:36 +0000 (14:02 +0000)] 
Remove space before exclamation mark: "Wow that rocked !"

11 years ago- more fixes for #163, make sure env.py load figures out to load env.pyc or
Mike Bayer [Tue, 31 Dec 2013 20:19:40 +0000 (15:19 -0500)] 
- more fixes for #163, make sure env.py load figures out to load env.pyc or
env.pyo as well

11 years ago- The :class:`.ScriptDirectory` system that loads migration files
Mike Bayer [Tue, 31 Dec 2013 20:01:50 +0000 (15:01 -0500)] 
- The :class:`.ScriptDirectory` system that loads migration files
from a  ``versions/`` directory now supports so-called
"sourceless" operation,  where the ``.py`` files are not present
and instead ``.pyc`` or ``.pyo`` files are directly present where
the ``.py`` files should be.  Note that while Python 3.3 has a
new system of locating ``.pyc``/``.pyo`` files within a directory
called ``__pycache__`` (e.g. PEP-3147), PEP-3147 maintains
support for the "source-less imports" use case, where the
``.pyc``/``.pyo`` are in present in the "old" location, e.g. next
to the ``.py`` file; this is the usage that's supported even when
running Python3.3.  #163

11 years ago0.1.8
Mike Bayer [Mon, 30 Dec 2013 18:46:21 +0000 (13:46 -0500)] 
0.1.8

11 years ago0.1.7
Mike Bayer [Mon, 30 Dec 2013 18:18:32 +0000 (13:18 -0500)] 
0.1.7

11 years ago0.1.6
Mike Bayer [Mon, 30 Dec 2013 18:11:44 +0000 (13:11 -0500)] 
0.1.6

11 years ago0.1.5
Mike Bayer [Mon, 30 Dec 2013 17:59:58 +0000 (12:59 -0500)] 
0.1.5

11 years ago0.1.4
Mike Bayer [Mon, 30 Dec 2013 17:55:30 +0000 (12:55 -0500)] 
0.1.4

11 years agoup
Mike Bayer [Mon, 30 Dec 2013 17:51:21 +0000 (12:51 -0500)] 
up

11 years ago0.1.1
Mike Bayer [Mon, 30 Dec 2013 17:48:31 +0000 (12:48 -0500)] 
0.1.1

11 years agotry out paramlinks
Mike Bayer [Mon, 30 Dec 2013 17:23:46 +0000 (12:23 -0500)] 
try out paramlinks

11 years ago- 0.6.2 release rel_0_6_2
Mike Bayer [Fri, 27 Dec 2013 19:25:05 +0000 (14:25 -0500)] 
- 0.6.2 release

11 years ago- clean out test_autogenerate
Mike Bayer [Fri, 27 Dec 2013 17:37:07 +0000 (12:37 -0500)] 
- clean out test_autogenerate
- start pinning some tests to 0.9, don't need to pin to 0.7 anymore
- Autogenerate for ``op.create_table()`` will not include a
  ``PrimaryKeyConstraint()`` that has no columns.
- don't need a full string test for boolean/check constraint autogen

11 years ago- move out "render" types of tests into new test_autogen_render
Mike Bayer [Fri, 27 Dec 2013 17:36:04 +0000 (12:36 -0500)] 
- move out "render" types of tests into new test_autogen_render

11 years ago- Fixed bug in the not-internally-used :meth:`.ScriptDirectory.get_base`
Mike Bayer [Fri, 27 Dec 2013 15:25:51 +0000 (10:25 -0500)] 
- Fixed bug in the not-internally-used :meth:`.ScriptDirectory.get_base`
method which would fail if called on an empty versions directory.

11 years ago- more for #157:
Mike Bayer [Fri, 27 Dec 2013 06:51:43 +0000 (01:51 -0500)] 
- more for #157:
- the ``op.create_table()`` directive will be auto-generated with
  the ``UniqueConstraint`` objects inline, but will not double them
  up with a separate ``create_unique_constraint()`` call, which may
  have been occurring.  Indexes still get rendered as distinct
  ``op.create_index()`` calls even when the corresponding table was
  created in the same script.
- the inline ``UniqueConstraint`` within ``op.create_table()`` includes
  all the options like ``deferrable``, ``initially``, etc.  Previously
  these weren't rendering.
- fixed the index tests to make sure the connection closes after each test
- _render_unique_constraint() and _add_unique_constraint() both call into
a common function now
- call _alembic_autogenerate_prefix within the add_index/drop_index renders

11 years ago- repair the autogen index test which apparently wasn't running for Postgresql
Mike Bayer [Fri, 20 Dec 2013 20:29:57 +0000 (15:29 -0500)] 
- repair the autogen index test which apparently wasn't running for Postgresql
due to a hardcoded sqlite_db()
- get autogen index test cleaned up, working on Postgresql and MySQL
- rework the autogen index/unique constraint logic to consider both sets of
objects together, and work up all new rules for deduping, comparison.
Add new rules specific to MySQL and SQLite and expand the flexibility surrounding
the "dupe unique index/unique constraint" thing to work around MySQL doing it
in both directions, handles #157
- add a lot more tests and try to emphasize no false positives

11 years agosince msg() right-trims, move the space to the subsequent message
Mike Bayer [Wed, 18 Dec 2013 16:58:30 +0000 (11:58 -0500)] 
since msg() right-trims, move the space to the subsequent message

11 years agoMerged in s0undt3ch/alembic (pull request #14)
Mike Bayer [Wed, 18 Dec 2013 16:55:27 +0000 (11:55 -0500)] 
Merged in s0undt3ch/alembic (pull request #14)

Just a space wrapping `...`

11 years agoJust a space wrapping `...`.
Pedro Algarvio [Wed, 18 Dec 2013 01:57:02 +0000 (01:57 +0000)] 
Just a space wrapping `...`.

This eases, for example, selecting the path when the text us double clicked on a console.

11 years agofix incorrect var for log message
Mike Bayer [Thu, 12 Dec 2013 00:28:50 +0000 (19:28 -0500)] 
fix incorrect var for log message

11 years ago- Fixed an issue with unique constraint autogenerate detection where
Mike Bayer [Wed, 11 Dec 2013 22:00:06 +0000 (17:00 -0500)] 
- Fixed an issue with unique constraint autogenerate detection where
a named ``UniqueConstraint`` on both sides with column changes would
render with the "add" operation before the "drop", requiring the
user to reverse the order manually.
- reorganize the index/unique autogenerate test into individual test cases;
ideally the whole test suite would be broken out like this for those big
tests

11 years ago- Added new argument ``mssql_drop_foreign_key`` to
Mike Bayer [Sat, 7 Dec 2013 00:05:53 +0000 (19:05 -0500)] 
- Added new argument ``mssql_drop_foreign_key`` to
:meth:`.Operations.drop_column`.  Like ``mssql_drop_default``
and ``mssql_drop_check``, will do an inline lookup for a
single foreign key which applies to this column, and drop it.
For a column with more than one FK, you'd still need to explicitly
use :meth:`.Operations.drop_constraint` given the name,
even though only MSSQL has this limitation in the first place.

11 years ago- test adjustments to account for SQLAlchemy pullreq #7, MSSQL DROP INDEX syntax
Mike Bayer [Fri, 6 Dec 2013 21:37:38 +0000 (16:37 -0500)] 
- test adjustments to account for SQLAlchemy pullreq #7, MSSQL DROP INDEX syntax

11 years ago- 0.6.2
Mike Bayer [Thu, 5 Dec 2013 23:26:43 +0000 (18:26 -0500)] 
- 0.6.2
- changelog for pullreq 13

11 years agoMerged in dharland/alembic/patch-mssql-final-go (pull request #13)
Mike Bayer [Thu, 5 Dec 2013 23:24:52 +0000 (18:24 -0500)] 
Merged in dharland/alembic/patch-mssql-final-go (pull request #13)

MSSQL Batch generator should ensure a COMMIT is followed by the batch separator

11 years agoadjust for latest ENUM formatting
Mike Bayer [Sat, 30 Nov 2013 21:57:00 +0000 (16:57 -0500)] 
adjust for latest ENUM formatting

11 years agoEnsure a commit is followed by a GO
donkopotamus [Thu, 28 Nov 2013 03:30:32 +0000 (16:30 +1300)] 
Ensure a commit is followed by a GO

11 years agoMerged zzzeek/alembic into master
Derek Harland [Thu, 28 Nov 2013 02:43:42 +0000 (15:43 +1300)] 
Merged zzzeek/alembic into master

11 years ago- 0.6.1 release rel_0_6_1
Mike Bayer [Wed, 27 Nov 2013 17:24:52 +0000 (12:24 -0500)] 
- 0.6.1 release
- add wheel format

11 years ago- Fixed bug where :func:`.op.alter_column` in the MySQL dialect
Mike Bayer [Wed, 27 Nov 2013 00:14:27 +0000 (19:14 -0500)] 
- Fixed bug where :func:`.op.alter_column` in the MySQL dialect
would fail to apply quotes to column names that had mixed casing
or spaces. #152

11 years ago- rework mysql tests to have suites
Mike Bayer [Wed, 27 Nov 2013 00:05:03 +0000 (19:05 -0500)] 
- rework mysql tests to have suites
- add a test to verify pullreq #5 - check if a TIMESTAMP column with
CURRENT_TIMESTAMP as default successfully compares for equality/inequality.
test passes without the pullreq, so no failure case has been demonstrated.

11 years agocleanup
Mike Bayer [Tue, 26 Nov 2013 23:48:11 +0000 (18:48 -0500)] 
cleanup

11 years ago- make slug length default to 40, but also make it configurable.
Mike Bayer [Tue, 26 Nov 2013 19:08:46 +0000 (14:08 -0500)] 
- make slug length default to 40, but also make it configurable.

11 years ago- changelog
Mike Bayer [Tue, 26 Nov 2013 18:53:44 +0000 (13:53 -0500)] 
- changelog
- use newer changelog pullreq format

11 years agoMerged in Frozenball/alembic/longer-filenames (pull request #12)
Mike Bayer [Tue, 26 Nov 2013 18:44:56 +0000 (13:44 -0500)] 
Merged in Frozenball/alembic/longer-filenames (pull request #12)

Change filename length restriction to 60 characters & word truncation

11 years agoMerge branch 'fix_python2_sql_output'
Mike Bayer [Thu, 21 Nov 2013 23:49:47 +0000 (18:49 -0500)] 
Merge branch 'fix_python2_sql_output'

Conflicts:
docs/build/changelog.rst

11 years agoFixes to Py3k in-place compatibity regarding output encoding and related;
Mike Bayer [Thu, 21 Nov 2013 23:46:35 +0000 (18:46 -0500)] 
Fixes to Py3k in-place compatibity regarding output encoding and related;
the use of the new io.* package introduced some incompatibilities on Py2k.
These should be resolved, due to the introduction of new adapter types
for translating from io.* to Py2k file types, StringIO types.
Thanks to Javier Santacruz for help with this.

11 years agoFixed the output wrapping for Alembic message output, so that
Mike Bayer [Thu, 21 Nov 2013 22:47:24 +0000 (17:47 -0500)] 
Fixed the output wrapping for Alembic message output, so that
we either get the terminal width for "pretty printing" with
indentation, or if not we just output the text as is; in any
case the text won't be wrapped too short. #135

11 years agoChange filename length restriction to 60 characters & word truncation
Teemu Kokkonen [Wed, 20 Nov 2013 13:52:17 +0000 (15:52 +0200)] 
Change filename length restriction to 60 characters & word truncation

Example of a longer filename:
- 3e12cdb9209d_alter_user_table_telephone_and_mobile_columns.py

Example of a filename that has been truncated:
- 12894284029g_create_partial_index_for_article_telephone_column_and_.py

11 years ago- make these comparisons handle the indexes in either order
Mike Bayer [Sun, 17 Nov 2013 19:48:09 +0000 (14:48 -0500)] 
- make these comparisons handle the indexes in either order

11 years ago- remember to add the SQLA prefix
Mike Bayer [Sun, 17 Nov 2013 17:57:46 +0000 (12:57 -0500)] 
- remember to add the SQLA prefix
- fix the SQLA 0.7 case a bit

11 years agowrap text() around it
Mike Bayer [Wed, 13 Nov 2013 16:35:42 +0000 (11:35 -0500)] 
wrap text() around it

11 years agomove _get_index_column_names into autogenerate
Mike Bayer [Wed, 13 Nov 2013 15:08:23 +0000 (10:08 -0500)] 
move _get_index_column_names into autogenerate

11 years ago- support SQLAlchemy ticket 2742 by ensuring that expressions in DDL are rendered
Mike Bayer [Wed, 13 Nov 2013 15:03:55 +0000 (10:03 -0500)] 
- support SQLAlchemy ticket 2742 by ensuring that expressions in DDL are rendered
(at least in indexes for now) using literal_binds=True, resolve #155

11 years ago- also need requirements file for rtd
Mike Bayer [Sat, 9 Nov 2013 18:23:53 +0000 (13:23 -0500)] 
- also need requirements file for rtd

11 years ago- use pypi-based changelog
Mike Bayer [Sat, 9 Nov 2013 18:02:09 +0000 (13:02 -0500)] 
- use pypi-based changelog

11 years agoFixes stdout --sql output in python2
Javier Santacruz [Thu, 17 Oct 2013 15:24:28 +0000 (17:24 +0200)] 
Fixes stdout --sql output in python2

When output_encoding is set, wraps the output buffer into a io.TextIOWrapper class
This means that the output_buffer has to be a io.IOBase instance
in order to work along with the TextIOWrapper

Handles wrapping when the buffer is Python2 stdtout, which has 'file' type
Adds test for this situation

11 years agoMerge pull request #6 from ijl/issue_141
mike bayer [Tue, 15 Oct 2013 17:41:02 +0000 (10:41 -0700)] 
Merge pull request #6 from ijl/issue_141

Fix #141: SQLA 0.7 compatibility for index column names

11 years agoMerge pull request #7 from ijl/issue_149
mike bayer [Tue, 15 Oct 2013 17:30:45 +0000 (10:30 -0700)] 
Merge pull request #7 from ijl/issue_149

Fix #149: bulk_insert example does not import op like other examples

11 years agoFix #149: bulk_insert example does not import op like other examples
ijl [Tue, 15 Oct 2013 15:40:49 +0000 (11:40 -0400)] 
Fix #149: bulk_insert example does not import op like other examples