Hello. I have a question about SQLite join-mechanisms. Let me explain with an example.
I have the follow table. CREATE TABLE a( word_id INTEGER, doc_id INTEGER ); CREATE INDEX "a_index1" ON "a"( "doc_id" ASC ); CREATE INDEX "a_index" ON a ( "word_id" DESC ); And how can SQLite do an optimized join of the follow query SELECT a.doc_id FROM a, a as a1, a as a2 WHERE a.doc_id = a1.doc_id and a1.doc_id = a2.doc_id and a.word_id = 1 and a1.word_id = 2 and a2.word_id = 4 LIMIT 0,20 What the example do is find the doc_id that has word_id 1,2 and 4. And suppose that there is: - 40000 docs with word_id 1 - 50000 docs with word_id 2 - 353500 docs with word_id 4 What i am searching is a optimized way to join, i mean, the algorithm. And I think SQLite has a great performance with joins. Thank to all. -- Cesar Rodas http://www.cesarodas.com/ Mobile Phone: 595 961 974165 Phone: 595 21 645590 [EMAIL PROTECTED] [EMAIL PROTECTED]
