The "force" parameter in SQLAlchemy IdentifierPreparer.quote()
has been a no-op since 0.9 in
031ef08078,
which was six years ago. In SQLAlchemy 1.3 this parameter
will be removed entirely. Bump requirements to 0.9 series
here and remove usage of the "force" flag.
Change-Id: I4492df2e7d2075fefbf13d6782de11f7d402f6b8
The 2.0.0 is breaking in that it removes the use of warnerrors in
build_sphinux.
sqlalchemy-migrate isn't using that feature, so it shoudln't break.
The cap on pbr is preventing other OpenStack projects that would like to
use pbr 2.0.0 (and sphinx 1.5.1) from doing so as it breaks
co-installability with sqlalchemy-migrate
Change-Id: I1c907201c717fe42caca24831985a119f2a1738b
Related-Bug: 1668848
There are two changes which have to go together to pass the gate
tests:
1. Update pbr and mock requirements from global-requirements
mock 1.2 supports py26 again so make that the minimum version. The
same change is being made in g-r with:
Ic6b9e18eaec9c81bbbbc57129e024904be928e09
Sync up with latest pbr in global-requirements while we're at it.
Closes-Bug: #1474925
2. Fix the importpath module to work with python >= 3.3 where the
__import__ built-in is raising an ImportError on a temporary file
that is added to the system path.
Closes-Bug: #1475339
Change-Id: Ie98938ba75f3983094dd540b7d26a7ec46be4f6e
This reverts commit e4d0e5be8d
Now that flake8-related dependencies have been updated in the
dependent change the pep8 job is fixed.
Change-Id: Idfa6a18836d7ce02dfaa5d9da1a51c98ad987f51
This reverts commit 35832555c5
The non-voting pep8 job was failing due a VersionConflict
with pbr, so this shouldn't have been merged.
Change-Id: I4917b92121cac524fd89575f30d72d7319cbe20c
The cap of pbr causes issues now that pbr 1.2 is basically needed for
a lot of projects to do the right thing with requirements ranges. The
sqla capping is preventing new versions of sqla to be used in the
OpenStack gates, and shouldn't be capped by a library.
Change-Id: I5fc142eb8c9d616db2ed9b2f3e4e4d1147e131ff
Lift the requirements to support SQLAlchemy 1.0. Two tests
were calling upon revised APIs and required adjustment.
Change-Id: Ic91a91bb3c915027b522eace302f2ed074233294
Some drivers [f.e. MySQL Connector] do not like multiple statements
being passed to .execute(). They require either passing multi=True
parameter to .execute() that is not DB-API 2.0 defined, or executing
those statements one by one.
For that patch, I've chosen the second option to stay away from driver
specific hacks.
Also removed SQLite hack that seems to be related to the same multiple
statements issue.
blueprint enable-mysql-connector
Change-Id: Ic6d53ed1fef8aee9471f3540f06b39cd5ee4ef82
migrate can't just take a global requirements sync because it
needs to be tested against multiple versions of SQLA to assure
compatibility. A recent change had the effect of only testing
migrate against SQLA 0.7, which is definitely *not* what we
want to be doing.
this reverts that change, and leaves very specific comments to
hopefully prevent this from happening in the future.
Change-Id: Icb4e136f0de6caa224019bb955341c4b67c5e1a1
This patch adds the initial support for DB2 10.5 to migrate. It
includes:
1. The dialect implementation for DB2.
2. The dialect registration with the visitor.
3. Code to parse the engine name in version.py.
4. A new dependency on ibm_db_sa in test-requirements.txt.
5. A connection string in test_db.cfg for ibm_db_sa.
Part of blueprint add-db2-support
Co-authored-by: Sheng Bo Hou <sbhou@cn.ibm.com>
Co-authored-by: Thuy Christenson <thuy@us.ibm.com>
Co-authored-by: Rahul Priyadarshi <rahul.priyadarshi@in.ibm.com>
Change-Id: I745ec615487b1b06c5d1a09ea316f376d66ee4c0
pbr updated itself to remove the d2to1 dependency. One of the main
reasons for this is a direct upstream dependency on distribute, which
causes the distribute/setuptools merge bug to get triggered, which
breaks people's system if they install with pip < 1.4.
Remove d2to1 references, and also bump the required version requested of
pbr to a version late enough to no longer depend on d2to1.
Remove the version specifier from setup.py, because setup_requires has
no way to trigger a version requirement upgrade. The flat version should
be sufficient for the setup.py interface in perpetuity.
Remove the setup-hook from setup.cfg, as it is no longer needed.
Change-Id: I40d3507c77802d13fd894aa16714b2ed6aa7d59f
Mike Bayer