Jim, just did a test with your recommendation, however that ended up using too much memory, since it is selecting the entire list in on go and keeping the memory for a significant amount of time, for the embedded system it is running on. Other tasks started having problems.
Travis -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Jim Wilcoxson Sent: March-16-11 12:51 PM To: General Discussion of SQLite Database Subject: Re: [sqlite] Optimizing list retrieval with a FTS3 table On Wed, Mar 16, 2011 at 3:15 PM, Travis Orr <[email protected]> wrote: > I am currently working on a project that requires retrieving a list of > all the rows from a FTS3 table. The ordering of the results varies by > search criteria. Since this is for an embedded project the list results > are passed in chunks to another module to give the appearance of faster > operations. Somewhere you have some state information so that you know what offset to use. Store the SQLite cursor with that state information, and use it to fetch the next 2000 rows on each call, ie, only do the query once. Jim -- HashBackup: easy onsite and offsite Unix backup http://www.hashbackup.com _______________________________________________ sqlite-users mailing list [email protected] http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list [email protected] http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
