[sqlalchemy] Re: Polymorphic collections / ticket #500

2007-03-12 Thread Rick Morrison
Yeah, disregard this, I found an easier way that doesn't fight SA. For the record, the use case is a polymorphic collection with the occasional exception. Exceptions were one-offs for which I didn't want to provide derived mappers because they were well, rare one-offs. Those exceptions were to be

[sqlalchemy] Re: Polymorphic collections / ticket #500

2007-03-07 Thread Michael Bayer
im not sure if i understand the use case youre describing ? I also stuck the previously mentioned flag in the trunk, the exception message will tell you about it. On Mar 6, 2007, at 11:08 PM, Rick Morrison wrote: > Mike: > > I think I've seen a few requests here on the list over the past

[sqlalchemy] Re: Polymorphic collections / ticket #500

2007-03-06 Thread Rick Morrison
Mike: I think I've seen a few requests here on the list over the past months for some kind of "default" or "unspecified" or other catch-all for the occasional one-off exception to what are otherwise polymorphic collections. I think it's actually a fairly common use-case. I'll try some of Simon's

[sqlalchemy] Re: Polymorphic collections / ticket #500

2007-03-06 Thread Michael Bayer
yeah, i didnt like adding this particular error message, but since it leads to a bigger problem later i sort of had to. i am sure a lot of people are going to hit this one. so i guess ill put a flag in for this onei think people shoud me made aware that they are allowing a potentially

[sqlalchemy] Re: Polymorphic collections / ticket #500

2007-03-06 Thread King Simon-NFHD78
I wanted to do something like this in the past, and in the end, rather than using polymorphic mappers it made more sense to create a MapperExtension which overrides create_instance. In create_instance you can examine your 'typ' column to decide what class to create, selecting one of your Manager/De