Let me check my understanding of what you're suggesting. The result
would look something like:
crates2apples = select([ crate_table.c.id,
apple_table.c.id,
apple_table.c.description,
apple_table.c.stem_length ],
from_obj = <join here>)
crates2apples_mapper = mapper(Apples, crates2apples, non_primary=True)
mapper(Crates, crate_table, properties={
'apples': relation(Apples, <actually, I have no clue what would go here>),
})
And, just to be clear, I could eagerload the Crates.apples relation
under this configuration?
Thanks,
Alan
On Sat, Dec 27, 2008 at 4:01 PM, Michael Bayer <mike...@zzzcomputing.com> wrote:
>
> you probably dont want to bother with "secondary" since it doesnt get
> you anything here and you should likely pare down the size of your
> join condition to something more minimal. you might want to also
> make life easier and make the relation to a non_primary mapper for
> Apples which maps to a simple select() statement, instead of shoving
> everything into a huge relation(). You don't need to establish
> use_labels as that only has to do with multiple tables in a single
> select() not conflicting on the column namespace. You also should be
> using 0.5 not only because it handles things like this better but also
> any bugs we fix are likely to be in that series.
>
>
> On Dec 27, 2008, at 5:24 PM, Alan Shields wrote:
>
>>
>> (wow, way to go tab key! Sorry, folks)
>>
>> I'm having a bit of trouble and I think it may be a deficiency in my
>> understanding of SqlAlchemy.
>>
>> Let's say I have:
>>
>> Table(Apples)
>> Table(Crates)
>>
>> apples_by_crate = \
>> select([ crate_table.c.id.label("crate_id"),
>> apple_table.c.id.label("apple_id"),
>> ],
>> from_obj = <long list of tables and conditions here>)
>> apples_by_crate_a =
>> apples_by_crate.correlate(None).alias('_apples_by_crate')
>>
>>
>> mapper(Crates,
>> 'apples':relation(Apples, viewonly=True,
>> secondary=apples_by_crate,
>> primaryjoin= apple_table.c.id
>> == apples_by_crate_a.c.apple_id,
>> secondaryjoin=
>> apples_by_crate_a.c.crate_id == crate_table.c.id,
>> foreignkeys= [
>> apples_by_crate_a.c.apple_id, crate_table.c.id ]))
>>
>> This works fine...until I try to eagerload apples. Bad Things Happen.
>> If it's a confounding factor, know that
>> apples_by_crate contains another subquery that also has a column
>> "crate_id".
>>
>> My impression right now is that I need to .use_labels for my
>> subqueries so I don't have conflicting names, but I'm starting to
>> wonder if I'm just doing it wrong.
>>
>> Any advice?
>>
>> Alan
>>
>> >
>
>
> >
>
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To post to this group, send email to sqlalchemy@googlegroups.com
To unsubscribe from this group, send email to
sqlalchemy+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sqlalchemy?hl=en
-~----------~----~----~----~------~----~------~--~---
