Recursive selects: http://web2py.com/book/default/chapter/06#Recursive-selects. One query per record, so a join is more efficient if processing lots of records at once.
Anthony On Saturday, September 24, 2011 1:59:21 AM UTC-4, encompass wrote: > > I thought I needed a join to print .name here... > def create_connections(): > #get all the user's tags for the about me > attribute = db((db.users_tags.event == session.current_event) & > (db.users_tags.intent == "attribute"))._select(db.users_tags.tag) > #get all the user's tags for the looking for > looking_for = db((db.users_tags.event == session.current_event) & > (db.users_tags.intent == "looking") & (db.users_tags.user == > db.auth_user.id))._select(db.users_tags.tag) > interesting_users = db((db.auth_user.id == db.users_tags.user) & > (db.users_tags.event == session.current_event) & (db.users_tags.user != > auth.user_id) & (db.users_tags.intent == "attribute") & > (db.users_tags.tag.belongs(looking_for))).select() > print interesting_users > for item in interesting_users: > print item.auth_user.nickname > print item.users_tags.event.name > Is this some new feature or is it like django where it is lazy and > creates another query from that line. > BR, > Jason Brower > >
