I dont understand what the feature would be. If you need to load many
types of objects and just construct them together in some custom way,
just load them all and construct as needed, as in :
for x, y, z in Session.query(X, Y, Z).join(....):
yield SomeGoofyObject(x, y, z)
On May 1, 2009, at 6:49 PM, David Gardner wrote:
> Yeah, I guess that is the problem. I'm working on one of those
> "everything" reports, and different bits of information are at
> different levels in the hierarchy, which are not directly related to
> each other.
>
> So my joins are somewhat arbitrary joins, without a real foreign key
> or a relation. I'm wondering if this is something that could get
> added to SQLAlchemy?
> Would this be difficult?
>
> Michael Bayer wrote:
>>
>> oh. contains_eager is only used for relation()s, so you'd have to
>> name what relation gets the collection, as in
>> contains_eager(Asset.something, alias=scn_alias).
>>
>> On May 1, 2009, at 5:17 PM, David Gardner wrote:
>>
>>> The problem I'm running into is what to pass into
>>> contains_eager(). In my table I'm mapping a file system,
>>> Asset.path is a unique constraint, so my outerjoin() will either
>>> find a single row or none.
>>>
>>> When I do:
>>> --------------------
>>> scn_alias=aliased(Asset)
>>> p='testshow/eps/201/s22/t04'
>>> take = session.query(Asset).filter(Asset.path==p).\
>>> outerjoin((scn_alias,scn_alias.path=="%s/anim/scn"%p)).\
>>> options(contains_eager(Asset,alias=scn_alias)).one()
>>
>>
>>
>>
>>
>>>
>>> -----------------------
>>> I get:
>>> Traceback (most recent call last):
>>> File "<stdin>", line 3, in <module>
>>> File "/usr/lib/pymodules/python2.5/sqlalchemy/orm/query.py",
>>> line 602, in options
>>> return self.__options(False, *args)
>>> File "<string>", line 1, in <lambda>
>>> File "/usr/lib/pymodules/python2.5/sqlalchemy/orm/query.py",
>>> line 52, in generate
>>> fn(self, *args[1:], **kw)
>>> File "/usr/lib/pymodules/python2.5/sqlalchemy/orm/query.py",
>>> line 619, in __options
>>> opt.process_query(self)
>>> File "/usr/lib/pymodules/python2.5/sqlalchemy/orm/
>>> interfaces.py", line 655, in process_query
>>> self._process(query, True)
>>> File "/usr/lib/pymodules/python2.5/sqlalchemy/orm/
>>> interfaces.py", line 661, in _process
>>> paths = self.__get_paths(query, raiseerr)
>>> File "/usr/lib/pymodules/python2.5/sqlalchemy/orm/
>>> interfaces.py", line 719, in __get_paths
>>> raise sa_exc.ArgumentError("mapper option expects string key
>>> or list of attributes")
>>> sqlalchemy.exc.ArgumentError: mapper option expects string key or
>>> list of attributes
>>>
>>>
>>>
>>> Michael Bayer wrote:
>>>>
>>>> On May 1, 2009, at 2:55 PM, David Gardner wrote:
>>>>
>>>>
>>>>> Along those same lines, one of the things I would like to do is to
>>>>> do an arbitrary join, my table is hierarchical and often I want to
>>>>> do a join against a a node, along with another node two levels
>>>>> deep
>>>>> below it, and since its a tree structure eager loading all of the
>>>>> children, and grandchildren would be costly, when I only care
>>>>> about
>>>>> two of them.
>>>>>
>>>>> So I have this:
>>>>>
>>>>> scn_alias=aliased(Asset)
>>>>> p='testshow/eps/201/s22/t04'
>>>>>
>>>>> take = session.query(Asset).filter(Asset.path==p).\
>>>>> outerjoin((scn_alias,scn_alias.path=="%s/anim/
>>>>> scn"%p),aliased=True).one()
>>>>>
>>>>> This produces the correct SQL, but since scn_alias isn't directly
>>>>> mapped as a relation to the row I am querying, I can't figure out
>>>>> the correct arguments to pass to contains_eager().
>>>>>
>>>> contains_eager just needs to know what columns in the selectable it
>>>> needs to draw from. so again here, drop the aliased=True since
>>>> scn_alias is already an alias, then pass scn_alias as the "alias"
>>>> keyword to contains_eager(). Also your outerjoin should probably
>>>> be
>>>> qualified against the parent Asset object otherwise you're in
>>>> danger
>>>> of cartesian products.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> David Gardner
>>> Pipeline Tools Programmer, "Sid the Science Kid"
>>> Jim Henson Creature Shop
>>> [email protected]
>>>
>>>
>>>
>>>
>>
>>
>>
>
>
> --
> David Gardner
> Pipeline Tools Programmer, "Sid the Science Kid"
> Jim Henson Creature Shop
> [email protected]
>
> >
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/sqlalchemy?hl=en
-~----------~----~----~----~------~----~------~--~---
