currently not but it is planned to be configurable (at the very minimum, cacheable).
Your method BTW works only if all the table is shown, doesn't take into account a possible condition passed to the grid (such as db.table.user_id == auth.user_id) or keyword queries. I'll start working on that as soon as possible, shouldn't be hard..... Talking about "extensibility", what you'd like to pass as a parameter ? I'm thinking to allow: - a callable (to which query and keywords are passed) --> you can pass the result of your custom query in this case, just watch out for the shortcomings explained earlier - an integer (which will trigger the "cache this count" for n seconds) Would that be enough? On Wednesday, April 17, 2013 12:58:27 PM UTC+2, Johann Spies wrote: > > By large datasets I mean a database of which one of the tables contain > more than 42 million records. > > Using SQLFORM.grid on such a dataset is very slow because the process > wants to count the total number of records. > > I am using Postgresql. I can get an estimated size of the table by > replacing count(*) with > > # SELECT reltuples::integer FROM pg_class WHERE oid = > 'isi.ritem'::regclass; > reltuples > ----------- > 42183232 > (1 row) > > And I get the result in a fraction of a second. When working on such a > scale the exact number of rows (count is also not always accurate on an > active database) is not a necessity. > > Is there a way I can replace the count(*) in the grid with the query > illustrated above? > > Regards > Johann > -- > Because experiencing your loyal love is better than life itself, > my lips will praise you. (Psalm 63:3) > -- --- You received this message because you are subscribed to the Google Groups "web2py-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to web2py+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
