Am 21.11.2009 um 04:48 schrieb Phil Longstaff: > sqlite> explain query plan SELECT DISTINCT t.* FROM transactions AS > t, splits > AS s WHERE s.tx_guid=t.guid AND > s.account_guid='e3ea8186deb3a9c160ab3b9409ea618f'; > 0|0|TABLE transactions AS t WITH INDEX > sqlite_autoindex_transactions_1 ORDER > BY > 1|1|TABLE splits AS s > sqlite> > > Hmmm... There are indices. I recently modified the splits table, > removing > "NOT NULL" from a column. I then copied the data into the new table > using: > > CREATE TABLE splits_new (...) > INSERT INTO splits_new SELECT FROM splits > DROP TABLE splits > ALTER TABLE splits_new RENAME TO splits > > How can I force the indices to be regenerated?
You dropped the table splits, which implicitly also dropped any indices associated with that table. You'll have to recreate the indices after recreating a table using the standard 'CREATE INDEX ...' SQL statement. </jum> _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
