6580
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6579
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6578
|
|
Convert to the Storm Python ORM <storm.canonical.com>. There were several reasons for this, but most importantly, the changes from SQLAlchemy/Elixir 0.3 to 0.4 were substantial and caused a lot of work. This work unfortunately did not result in a working branch due to very strange and inconsistent behavior with Unicode columns. Sometimes such columns would return Unicode, sometimes 8-bit strings, with no rhyme or reason. I gave up debugging this after many hours of head scratching.
Oh yeah, no more flush!
Storm enforces Unicode columns, which is very nice, though requires us to add lots of 'u's in places we didn't have them before. Ultimately, this is a good thing so that the core of Mailman will be Unicode consistent.
One thing I still want to clean up after this, is the function-scoped imports in the model code. Part of the reason for the separate model classes was to avoid this, but for now, we'll live with it. Storm's architecture requires us to maintain a database-table-class registry for simple clearing after tests in Database._reset(). This is made fairly simple by Storm allowing us to use our own metaclass for model classes.
Storm does require that we write our own SQL files, which is a downside, but I think our schema will be easy enough that this won't be a huge burden. Plus we have a head-start <wink>.
Another cool thing about Storm is the explicit use of stores for objects. This should eventually allow me to flesh out my idea of storage pillars for 1) lists, 2) users, 3) messages.
Some other changes:
- pylint and pyflakes cleanups - SQLALCHEMY_ENGINE_URL -> DEFAULT_DATABASE_URL - Don't import-* from Version in Defaults.py - Add interface method to Mailman.Message.Message so that __getitem__() and get_all() always return Unicode headers, even when the underlying objects are strings. This should generally be safe as headers are required by RFC to be within the ASCII range. - Fix bin/arch.py to use proper initialization.
|
Barry Warsaw |
17 years ago
|
 |
|
6577
|
|
|
Mark Sapiro |
17 years ago
|
 |
|
6576
|
|
|
Mark Sapiro |
17 years ago
|
 |
|
6575
|
|
|
Mark Sapiro |
17 years ago
|
 |
|
6574
|
|
|
Mark Sapiro |
17 years ago
|
 |
|
6573
|
|
|
Mark Sapiro |
17 years ago
|
 |
|
6572
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6571
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6570
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6569
|
|
Much progress, though not perfect, on migrating to SQLAlchemy 0.4 and Elixir 0.4. Lots of things changes, which broke lots of our code. There are still a couple of failures in the test suite that I don't understand. It seems that for pending.txt and requests.txt, sometimes strings come back from the database as 8-bit strings and other times as unicodes. It's impossible to make these tests work both separately and together.
users.txt is also failing intermittently. Lots of different behavior between running the full test suite all together and running individual tests. Sigh.
Note also that actually, Elixir 0.4.0 doesn't work for us. There's a bug in that version that prevented zope.interfaces and Elixir working together. Get the latest 0.4.0 from source to fix this.
Other changes include:
- Remove Mailman/lockfile.py. While I haven't totally eliminated locking, I have released the lockfile as a separate Python package called locknix, which Mailman 3.0 now depends on.
- Renamed Mailman/interfaces/messagestore.py and added an IMessage interface.
- bin/testall raises turns on SQLALCHEMY_ECHO when the verbosity is above 3 (that's three -v's because the default verbosity is 1).
- add_domain() in config files now allows url_host to be optional. If not given, it defaults to email_host.
- Added a non-public interface IDatabase._reset() used by the test suite to zap the database between doctests. Added an implementation in the model which just runs through all rows in all entities, deleting them.
- [I]Pending renamed to [I]Pended
- Don't allow Pendings.add() to infloop.
- In the model's User impelementations, we don't need to append or remove the address when linking and unlinking. By setting the address.user attribute, SQLAlchemy appears to do the right thing, though I'm not 100% sure of that (see the above mentioned failures).
|
Barry Warsaw |
17 years ago
|
 |
|
6568
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6567
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6566
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6565
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6564
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6563
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6562
|
|
|
Barry Warsaw |
17 years ago
|
 |
|
6561
|
|
|
Barry Warsaw |
17 years ago
|
 |
|