seems that A0.dbid = A1.dbid = Ax.dbid are not equivalent, in regards to the relation/get().
On Thursday 02 October 2008 22:48:23 Michael Bayer wrote: > On Oct 2, 2008, at 4:05 PM, [EMAIL PROTECTED] wrote: > > On Thursday 02 October 2008 22:26:06 Michael Bayer wrote: > >> On Oct 2, 2008, at 11:57 AM, [EMAIL PROTECTED] wrote: > >>> hi > >>> lets say, A points to B. > >>> i do a query against A about which things link to some instance > >>> of B. then, > >>> 1) if i specify eagerload, i get many times same object-data > >>> being piped from the server; ok but i dont want it > >>> 2) if i dont specify eagerload (lazy), access to each A fires a > >>> lazyload-query for the (same) B! Although the object is in > >>> memory already. is this expected behaviour or? the relevant > >>> dbid is already present, hence it can be checked against > >>> session-caches etc?? what should i do here? > >> > >> if you load an A that wasnt previously present, its A.bs > >> collection is empty. > >> The individual B objects might be in the session but the > >> information that they are associated with your "A" is not > >> present, unless we scanned through the Session and matched up > >> foreign keys. This latter approach basically turns Session into > >> its own relational engine which is why we dont even try to get > >> into that. So the lazy query for the"B"s proceeds, but we do > >> reuse the B's already present in the session so there is far > >> less overhead associated with fetching the rows. > >> > >> The approach taken by some ORMs Ive seen is to link the A.bs > >> load to load the .bs for *all* As. That seems pretty "dirty" to > >> me since it means we have to keep track of Queries and such. > > > > nah, it's a simple *toone singular reference, not a collection. > > can't something be done for it? hack something somewhere? > > a simple many-to-one will use query.get(x) for the lazyload, and no > SQL will be issued if B is already present in the session. > > > --~--~---------~--~----~------------~-------~--~----~ 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 [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sqlalchemy?hl=en -~----------~----~----~----~------~----~------~--~---
