Hmmm... I can see we really need good documentation around this. By simple nature of a join, you MUST use groupBy. Otherwise you'll end up with too many parent objects. You can use queryForList or queryForObject with the solution, as long as the latter only returns a single parent object.
Regardless, you can use the resultMap and groupBy independently as required. You just combine them to eliminate repeating groups for a complex object graph. Cheers, Clinton On Mon, 17 Jan 2005 16:16:42 -0500, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote: > > I'm having trouble implementing the 2.0.9 N+1 solution. The example posted > on the wiki deals with a groupBy relationship. Is this the only situation > where the fix applies? I have a simple nested 1:M relationship that I would > love to used this new feature with. Presently I am using the old method of > using a "select " parameter in the result element of my resultMap. I have > replaced this parameter with a "resultMap" parameter. Of course I have > changed the SQL to join the necessary tables. Must I change my query to a > QueryForList and build my result manually or should I be able to use the > same queryForObject that I was using before? I am getting an > executeQueryForObject returned too many results exception. > _______________________________________ > Tim Hammar > Senior Programmer Analyst > McNichols Company > 5501 W. Gray Street, Tampa, FL 33609-1007 > P.O. Box 30300, Tampa, FL 33630-3300 > Local: 813-282-3828, ext. 2419 > Fax: 813-287-8387 > [EMAIL PROTECTED] > www.mcnichols.com
