Ok, I have a many_to_many relationship between two tables (A, B) implemented by a join table (J). J has two columns: a_id and b_id. Currently, I use a query like this: <code> recs = A.find :all, :select => 'A.*', :joins => ', B, J', :conditions => 'A.id = J.a_id AND J.b_id = B.id AND B.somefield = somevalue ' </code> That works well, but is there a more Rails like way to do it? Also, is there a more efficient way to do it? Notice in my RoR code, I'm joining 3 tables together in a single SQL statement. Would it be faster to select ids in an SQL call joining only 2 tables, then calling A.find :all, ids? In code: <code> ids_array = "SELECT J.a_id FROM J, B WHERE J.b_id = B.id AND B.somefield = somevalue" recs = A.find :all, ids_array </code> How do I efficiently execute that query and get an array of ids in return? Thanks for the help.
on 2007-01-19 16:30
on 2007-09-25 22:30
You have two options. has_and_belongs_to_many :foo Each of the two models that's gonna be joined has a has_and_belongs_to_many :other_model in it. Say your models are called Book and Author - the table would then be named authors_books (plural and in alphabetic order). The authors_books table would have two columns: book_id and author_id. No primary key. has_many :foo, :through => :bar A bit to intricate to explain. Look at this pastie: http://pastie.caboo.se/34094 On Jan 18, 9:15 pm, "Christopher J. Bottaro" <cjbott...@gmail.com>