> > I have a table that has two > > parents, > > that is a foreign key to two tables. One of those tables isn't needed > > by the application, but it needed for other applications that use > > these > > same SQLAlchemy models. > > > The following example shows first a default parent/child relationship > > with the resulting query calls. The second example shows an example > > using passive_deletes=True in the relationship from child to parent, > > which I understand doesn't make too much sense but causes my desired > > behavior. This, also, has the resulting query calls. > > > Is there a better way for me to get the behavior I desire, or is this > > a > > bug? > > Its a bug. Many-to-ones are generally cheap since they're usually already > loaded, but in this case the load isn't needed and this is actually a recent > regression as of 0.6.6, so 0.6.5 won't exhibit this behavior...of course > 0.6.5 has the previous issue that was fixed here but its likely not as > common. This is ticket #2049 and a fix will be available shortly. > > Also you might want to consider linking the "parent" and "children" > relationships via the "back_populates" attribute, or otherwise map them at > once using relationship + backref. Otherwise the ORM treats mutations in > each attribute as separate which could lead to doubling of operations. If > you're only mutating one side (or none) then it doesn't matter much. > > >
Will this bug fix cause issues with models that have a relationship with itself (where the parent of the object is another object of the same class)? I applied the patch you came up with and my tests fail on this type of relationship. -- 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.
