[sqlalchemy] Re: SQLAlchemy experts wanted

2008-05-18 Thread Michael Bayer
On May 16, 2008, at 3:23 PM, Jim R. Wilson wrote: Hi all, SQLAlchemy is a great project and a growing niche. As it becomes even more popular, there will be increasing demand for experts in the field. I am compiling a contact list of SQLAlchemy experts who may be interested in

[sqlalchemy] Re: composite primary key/postgres

2008-05-18 Thread Arlo Belshee
Here's the TSQL for a unique index: CREATE UNIQUE NONCLUSTERED INDEX IX_UQ_Sample ON Sample ( first ASC, other ASC, something ASC) I defined Sample as: CREATE TABLE Sample( first int NOT NULL, something int NULL, other bit NULL) I don't know how you'd get SqlAlchemy to

[sqlalchemy] Re: composite primary key/postgres

2008-05-18 Thread Michael Bayer
On May 18, 2008, at 11:34 AM, Arlo Belshee wrote: Here's the TSQL for a unique index: CREATE UNIQUE NONCLUSTERED INDEX IX_UQ_Sample ON Sample ( first ASC, other ASC, something ASC) I defined Sample as: CREATE TABLE Sample( first int NOT NULL, something int NULL, other

[sqlalchemy] Re: retrieving mapped objects from within a session

2008-05-18 Thread Moshe C.
Got back to this issue after a while. The SessionExtension objects allows me to hook on to a session and get notified of various events. My question is different: Given a session, before commit, how can I query it to know what is going to happen at commit. My intention is to derive from that, a

[sqlalchemy] Re: retrieving mapped objects from within a session

2008-05-18 Thread Michael Bayer
On May 18, 2008, at 2:09 PM, Moshe C. wrote: Got back to this issue after a while. The SessionExtension objects allows me to hook on to a session and get notified of various events. My question is different: Given a session, before commit, how can I query it to know what is going to

[sqlalchemy] transactional sessions not transactional?

2008-05-18 Thread Matthew Dennis
The following test case of mine fails on PG 8.3 and SA 0.4.3 Basically, create two sessions, make some changes in the first and obverse they are visible before commit/rollback in the second (and via connectionless execution directly on the engine), but become unvisible after rollback. The first