On Jun 14, 2006, at 8:39 PM, Jonathan LaCour wrote: > I am still having to do a lot more checking than I would like because > the 'relation' constructor takes in a class or mapper, and I have to > wait for the class or mapper to *exist* before creating it. Any > chance > of being able to pass a class name there instead? Then I could remove > this code entirely, and create relations as they are detected in the > metaclass. This would *greatly* simplify ActiveMapper. >
well im just a little concerned that it gives people another way to break their code.....now they can have two classes of the same name, which will have to be in some global dictionary somewhere to match the strings up with the classes/mappers, then itll either silently fail when the second entry blows away the first entry, or we have to make it throw an error. in the world of explicit mappers, having it use strings for class names is not very useful (though obviously more useful for SQLOb...i mean ActiveMapper :) ). i dunno i guess if i stick it in and just dont document it maybe.... _______________________________________________ Sqlalchemy-users mailing list Sqlalchemy-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/sqlalchemy-users
