Ah, belongs() + a clever select SQL.
Thanks.
Al
On Jun 29, 7:04 pm, mdipierro <mdipie...@cs.depaul.edu> wrote:
> If I understand:
>
> for row in db(db.link_table.table1==r.id)
> (db.table2.id==db.linked_table.table2).select(db.table2.id,db.table2.other):
> print row.id, row.other
>
> if no double links else
>
> sub=db(db.link_table.table1==r.id)._select(db.linked_table.table2)
> for row in
> db(db.table2.id.belongs(sub)).select(db.table2.id,db.table2.other):
> print row.id, row.other
>
> linked_view = db(db.table1.id == db.link_table.table1) & (db.table2.id
>
> > == db.linked_table.table2))
>
> > for row in linked_view(db.table1.id == r.id).select([fields from
> > table2 I want])
>
> On 29 Giu, 17:54, Al <rayalans...@yahoo.com> wrote:
>
>
>
> > Given a many-to-many relationship between table1 and table2 and a row
> > in table1, I am trying to figure out the fastest way to select all the
> > rows in table2 that are associated with the row. (I'm still trying to
> > pick up web2py, so thanks for everyone's patience if I've missed the
> > blatently obvious.)
>
> > From the web2py book documentation, the best I can figure out is
> > something like this:
> > db.define_table('link_table',
> > Field('table1', db.table1),
> > Field('table2', db.table2)
> > )
>
> > # Pick the row from table 1
> > r = db(db.table1.name == "Example").select()[0]
>
> > linked_view = db(db.table1.id == db.link_table.table1) & (db.table2.id
> > == db.linked_table.table2))
>
> > for row in linked_view(db.table1.id == r.id).select([fields from
> > table2 I want])
> > # Do something clever for the matching rows I've found
> > pass
>
> > But I keep thinking there ought to be something cleverer that I can
> > do, especially given that r has a set containing the rows of
> > link_table that contain the entries of table2 that I want. But I can't
> > figure out how to leverage that set into a shorter query.
>
> > Thanks,
> > Al- Hide quoted text -
>
> - Show quoted text -
