> >> - check whether to revive the doPsSelect methods or whether all
selects
> > can
> >> use prepared statements
> >
> > Re-added BasePeer.doPSSelect() for now.
>
> This method has some weaknesses IIRC. It dos not work correctly with
> selects containing joins. I'd prefer to use prepared statements for all
> selects if at all possible and I don't think we need separate methods.

So would I. Much less room for SQL-injection attacks. Will take a bit of
work, but if nobody objects, I'll have a go at it.

> >> - there was come code in doDelete which handled cascading deletes (at
> > leats
> >> theoretically, never saw it working) tris needs to be reimplemented
> >
> > Not re-added. The code can definitely not have worked for composite
foreign
> > keys because there is no way to tell a composite foreign key from the
> > database map.
>
> Still I'm pretty sure that deletes containing joins do not work
> correctly. So there should be another solution some day.

What do you mean by "deletes containing joins"? A delete which where clause
contains a join or a delete which deletes from multiple tables? I'm pretty
sure the former works now as the table name is now specified in the
interface (no magic table-to-delete-from detection from the criteria any
more); not sure whether the latter can work in SQL.

Would you mind to create a simple test case to show what you mean ?

> >> - Maybe one can clean up LargeSelect from the reflection stuff therein
> >
> > Not while the Peers have static methods.
> >> ...
>
> I beg to differ. LargeSelect just tries to call the methods
> addSelectColumns() and populateObject() on the given builder class. You
> can provide any class you want here. I'd prefer to use some kind of
> Builder-interface for this, however.

Agreed. But does one not still need to have non-static methods for
implementing an interface.

   Thomas


---------------------------------------------------------------------
To unsubscribe, e-mail: torque-dev-unsubscr...@db.apache.org
For additional commands, e-mail: torque-dev-h...@db.apache.org

Reply via email to