OK, that was a bug where it breaks down a save dependency into a hierarchical graph of smaller dependencies, when you have a circular mapper going...it was forgetting about all the other dependent operations on the original save dependency. my three current tests for self-referring mappers didnt take this case into account....so your test is working now if you update. rdb.Column('selfref_id', rdb.Integer, rdb.ForeignKey('t1')), to rdb.Column('selfref_id', rdb.Integer, rdb.ForeignKey('t1.id')), since it otherwise looks for a column of the same name, i.e. creates a reference to itself. the 'primaryjoin' criterion you have in the mapper is overriding it anyway. On Jan 19, 2006, at 8:13 PM, [EMAIL PROTECTED] wrote:
|
- [Sqlalchemy-users] self-ref .commit() problem daishi
- Re: [Sqlalchemy-users] self-ref .commit() problem Michael Bayer
- Re: [Sqlalchemy-users] self-ref .commit() problem Michael Bayer
- Re: [Sqlalchemy-users] self-ref .commit() problem daishi
- Re: [Sqlalchemy-users] self-ref .commit() probl... Michael Bayer
- Re: [Sqlalchemy-users] self-ref .commit() p... daishi
- Re: [Sqlalchemy-users] self-ref .commi... Michael Bayer
- Re: [Sqlalchemy-users] self-ref .c... daishi
- Re: [Sqlalchemy-users] self-re... Michael Bayer
- Re: [Sqlalchemy-users] self-re... daishi
- Re: [Sqlalchemy-users] ID sequ... daishi
- Re: [Sqlalchemy-users] self-re... Michael Bayer
- Re: [Sqlalchemy-users] self-ref .commit() probl... Michael Bayer

