Fin Springs wrote: > > I have been using: > > SELECT NULL FROM sqlite_master WHERE type='table' AND lower(name)=? > > to determine whether a table exists and looking at the number of rows > returned (I'm using sqlite3_get_table through an API). I get one row > back if the table exists and no rows when it doesn't. There wouldn't be > multiple rows to LIMIT in this case though. > > Is that bad?
It isn't good or bad, just different. Your application has to perform the extra logic of counting the returned rows to determine existence. If you use a subselect and exists, SQLite will effectively do that for your application and return a boolean value directly. select exists (SELECT * FROM sqlite_master WHERE type='table' AND lower(name)=?) This will always return one row with one column with a value that is either 0 or 1. HTH _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
