]> git.ipfire.org Git - thirdparty/sqlalchemy/alembic.git/commit
- will call this 0.6
authorMike Bayer <mike_mp@zzzcomputing.com>
Sat, 13 Apr 2013 00:10:56 +0000 (20:10 -0400)
committerMike Bayer <mike_mp@zzzcomputing.com>
Sat, 13 Apr 2013 00:10:56 +0000 (20:10 -0400)
commitd77b1d838d20d9b57154ce9bbb0c2e5a24c48cb0
tree33289812bacd8693094dbcb37df18ff7bf00340d
parent2f7b645df2a9c5d9d9601baa0ada4bd0f4c366b3
- will call this 0.6
- cleanup unused symbols in compat
- move any remaining conditional import switches out to compat
- pretty sure anything we use for a buffer will have flush()
- I can see that using the new io.* stuff means codecs isn't too
smooth here, but removed the double-wrap check in migration.py
which seems to be due to the bytesio setup in the test suite, simplified
this so that an encoding-writing test just writes to BytesIO
- removed all "from __future__ import with_statement"
- i think when we load_source(), that file might not be ascii
so we need to open as bytes
- make encoding an explcit argument in test suite write_script()
- sort the columns in autogenerate so that the autogen tests run
consistently on python3.3, seems to have a very random set ordering,
also loosen up an ordered check in test_op
15 files changed:
alembic/__init__.py
alembic/autogenerate.py
alembic/compat.py
alembic/config.py
alembic/ddl/impl.py
alembic/migration.py
alembic/script.py
alembic/util.py
tests/__init__.py
tests/test_mssql.py
tests/test_op.py
tests/test_oracle.py
tests/test_postgresql.py
tests/test_sql_script.py
tests/test_versioning.py