On Wed, Mar 4, 2009 at 11:40 AM, John Machin <sjmac...@lexicon.net> wrote:
> On 4/03/2009 2:12 PM, Peng Huang wrote: > > Hi Igor Tandetnik, > > > > Thanks for your quick reply. > > > > Your solution works. But in some cases, each y%d may has two or three > > choices. So the SQL will become very complex, we need ( 2 * 2 * 2 * 2) > sub > > where statements. Does SQLite have some build-in features to optimize > those > > kinds of SQL statements? Or do you have other suggestions to optimize the > > database of SQL statements? > > > > For example: > > "SELECT * FROM py_phrase WHERE ylen = ? and (y0 = ? or y0 = ?) and (y1 = > ? > > or y1 =? ) and (y2 = ? or y2 = ?) and (y3 = ? or y3 = ? or y3 = ?) ORDERY > BY > > user_freq DESC, freq DESC" > > > > (y3 = ? or y3 = ? or y3 = ?) > is *logically* equivalent to > (y3 IN (?, ?, ?)) > > Does it run at the same speed? Its improvement is not obviously. :( _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
