If you have synonym for rowid (i.e. column INTEGER PRIMARY KEY) or some other unique columns combination then you can do something like this:
INSERT OR REPLACE INTO table_1 (rowid, field_a, field_b, field_c, field_d) SELECT table_1.rowid, table_2.field_a, table_2.field_b, table_2.field_c, table_2.field_d FROM table_1, table_2 WHERE … Pavel On Thu, Aug 20, 2009 at 6:01 AM, Gerald Ebner<geraldo.eb...@gmail.com> wrote: > Dear all, > > it seems that sqlite does not allow update statements of this kind: > > UPDATE table_1 SET (field_a, field_b, field_c, field_d) = ( > SELECT field_a, field_b, field_c, field_d FROM table_2 WHERE … > ) > WHERE .... > > I succeeded only with > > UPDATE table_1 SET > field_a = (SELECT field_a FROM table_2 WHERE … ), > field_b = (SELECT field_b FROM table_2 WHERE … ), > field_c = (SELECT field_c FROM table_2 WHERE … ), > field_d = (SELECT field_d FROM table_2 WHERE … ) > WHERE .... > > and that's may be not th optimal way, dealing with tables of a dozen of > fields .... > > thanks in advance > Geraldo > > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users > _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users