Re: [Sqlalchemy-users] Namespace possible bug.

2006-08-06 Thread Michael Bayer
im not really following migrate so much but you can get in touch with its author Evan. you can check for tables directly by calling engine.dialect.has_table (tablename) (which needs to have a "schema" keyword argument and appripriate SQL added to each DB module, more or less finishing #233..

Re: [Sqlalchemy-users] Namespace possible bug.

2006-08-06 Thread Tzahi Fadida
On Sunday 06 August 2006 06:04, Michael Bayer wrote: > ah yeah the first would be fixed by ticket #233, i.e. a TODO. Well, it is very difficult to work with schemas without fixing 233, because even if i will write a ddl manipulator using plain sql statements i still have to detect this. Is ther

Re: [Sqlalchemy-users] Namespace possible bug.

2006-08-05 Thread Michael Bayer
ah yeah the first would be fixed by ticket #233, i.e. a TODO. for the second, we currently dont have any support for CREATE SCHEMA, (nor do we have things like ALTER, CREATE VIEW, stored procedure management, etc). the create/drop/reflection capabilities of SA are meant for small, self-pack

[Sqlalchemy-users] Namespace possible bug.

2006-08-05 Thread Tzahi Fadida
I am using postgresql. When i use the schema keyword in Table(), if there is already a table with the same name in the "public"/default schema then metadata.create_all() assumes the schema (other than public) and table already exists. I.e. instead of checking in the requested schema it uses "public