I think the inheritance introspector fix does come into play here.  Please verify that you have this fix in your source:
 
http://fisheye3.cenqua.com/changelog/ibatis/?cs=453298
 
I have a test working that, I think, is set up similarly to your situation.  If you want, I'll send yo the source offline.
 
Jeff Butler


 
On 10/17/06, Tom Duffey <[EMAIL PROTECTED]> wrote:

On Oct 17, 2006, at 11:25 AM, Tom Duffey wrote:


On Oct 17, 2006, at 11:22 AM, Jeff Butler wrote:

Never mind - the change was made prior to that.  BTW - the fisheye link on the iBATIS homepage is VERY usefull when trying to see what shanges have been committed to SVN.
 
I'll take a look in my own environment.

 

 
Actually I just realized that I only deployed ibatis-sqlmap-2.jar from trunk into my application so if the change was in ibatis-common then I didn't pick it up.  I'm rebuilding now and will deploy everything from trunk and let you know if it helps.

 
Same NPE after deploying latest ibatis-common and ibatis-sqlmap jars from trunk.  To recap, my type handler implementation is working fine because specifying it explicitly in the inline parameter map works.  The only difference I see in my setup compared to others is I define my enums in implements, e.g.,

 
public interface Something {
public enum Type { NORMAL, WEIRD }
...
}

 
and iBATIS maps to things that implement Something, e.g.,

 
public class WonderfulThing implements Something {
private Type type;
getType()
setType()
...
}

 
and without the explicit type handler mapping for that attribute iBATIS generates the NPE on insert/update.  Declaring SomethingType externally also works but I'd rather not do that.

Reply via email to