Hi Brian,
The best thing to do is to open a JIRA [1] for an issue like this.
Providing a junit test case is an extra bonus that will help a developer
narrow in on the problem immediately, and we can add it to our regression
bucket to prevent this from happening again.  Thanks!

Kevin

[1]  http://openjpa.apache.org/found-a-bug.html

On Wed, May 18, 2011 at 11:09 AM, Brian Gebala <[email protected]> wrote:

> Good morning all -
>
> I just updated to v2.1.0 from v2.0.1 and noticed that a query using TYPE()
> <> X is no longer working. I have an abstract base class and three
> subclasses. My query asks for all but one of the subclasses, like this:
>
> SELECT d FROM AbstractClass d WHERE d._state = :state AND TYPE(d) <>
> SubClassThree
>
> When I run this query with v2.1.0, the SQL is generated as if the query
> said
> "TYPE(d) = SubClassThree", resulting in a match on
> the SubClassThree discriminator. I found a work-around using NOT IN:
>
> SELECT d FROM AbstractClass d WHERE d._state = :state AND TYPE(d) NOT IN
> (SubClassThree)
>
> If anyone would like additional information, please let me know.
>
> Thanks,
> Brian
>

Reply via email to