| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
values are unicodes if they come from the base class as a string.
Get rid of the 'global id'. Now use just Message-ID. Rename
X-List-ID-Hash to X-Message-ID-Hash. Do not take Date header into
account when calculating this hash.
Because of the above change, the assumption is that there will be no
Message-ID collisions. Therefore, get rid of IMessageStore
.get_message(), .get_messages_by_message_id() and
.get_messages_by_hash(). Instead, it's now .get_message_by_id() and
.get_message_by_hash() both of which return the message object or
None.
Message.hash -> Message.message_id_hash
When storing a message in the message store, the final path component
has the entire hash, not just the leftover parts after directory
prefix splitting.
MessageStore.delete_message() deletes the file too.
Doctests clean up message store messages though the message store
instead of directly off the filesystem.
|
| |
|
|
|
| |
subdirectory and updating all relevant imports. Move of the circular
import problems have been eliminated in the process.
|
| |
|
|
|
|
| |
tests, we're making very good progress. Just the tough ones are left. This
change did modify the the schema a bit, for better naming and typing.
E.g. 'type' -> 'request_type' and using a RawStr for a hash type.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
enables a few interesting things:
1. It makes it easier to do our "pillars of storage" idea, where list data and
messages could live in one database, but user information live in a
separate database.
2. It reduces the number of moving parts. SQLAlchemy and Elixir can both go
away in favor of just one database layer.
3. No more Unicode/string mush hell. Somewhere along the way the upgrade to
SQLAlchemy 0.4 and Elixir 0.4 made the strings coming out the database
sometimes Unicode and sometimes 8-bit. This was totally unpredictable.
Storm asserts that if a property is declared Unicode, it comes in and goes
out as Unicode.
4. 'flush' is gone.
One cost of this is that Storm does not yet currently support schema
generation. So I cheat by dumping the trunk's SQLite schema and using that as
a starting place for the Storm-based schema. I hope that Storm will
eventually address this.
Other related changes include:
- SQLALCHEMY_ENGINE_URL is renamed to DEFAULT_DATABASE_URL. This may still
get changed.
Things I still want to fix:
- Ickyness with clearing the databases.
- Really implement multiple stores with better management of the Store
instances.
- Fix all the circular import nasties.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
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).
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Mailman/app/moderator.py: Most of the application level interface
provided by ListAdmin is moved here now, including the ability to hold
messages, subscriptions, and unsubscriptions, and to handle message
(defer, discard, reject, accept). More work needed.
Some untested conversion of API in Mailman/Cgi/admindb.py, confirm.py,
bin/checkdbs.py.
messagestore.py: Don't use or require the Date: header in the global
message ID calculation. As described on the mailing list, we're only
going to use the Message-ID header.
IListRequests: added count_of() and of_type() methods.
|
|
|
for the archive or not is left to be seen.
Define an interface, test, and implementation of a basic message store using
globally unique identifiers of the form: archive/hash/seqno
- archive is the base url of the archive, e.g. http://archives.example.com.
This is available in the List-Archive header.
- hash is the base32 encoded sha1 hash of the message's Message-ID and Date
headers, which it must have. This is available in the X-List-ID-Hash
header.
- seqno is a sequence number specific to the archive which will uniquely
identify the message should there be a Message-ID/Date collision. this is
available in the X-List-Sequence-Number header.
Added a MESSAGES_DIR variable to the config.
Added a .message_store attribute to the config.
|
