The test should be fairly simple. Go ahead and check it in without review.
-Tom
Michael Bouschen wrote:
>Hi Tom,
>
>
>
>>Hi Michael,
>>
>> The changes look good.
>>
>>
>
>Thanks!
>
>
>
>> Are you adding a test to the entity persistence tests?
>>
>>
>
>That makes sense, I will add a test tommorrow. Are you interested to
>review the test before I check in?
>
>Regards Michael
>
>
>
>>-Tom
>>
>>Michael Bouschen wrote:
>>
>>
>>
>>>Hi,
>>>
>>>sorry, I forgot the attachemnt.
>>>
>>>Regards Michael
>>>
>>>
>>>
>>>
>>>
>>>>Hi Tom,
>>>>
>>>>welcome back!
>>>>
>>>>attached you find my changes fixing the NPE in ReportQuery thrown
>>>>for an EJB QL query using an aggregate function as an argument of a
>>>>constructor
>>>>expression:
>>>> SELECT NEW custom.CityCount(c.city, COUNT(c)) FROM Customer c
>>>> GROUP BY c.city
>>>>
>>>>In the parse tree node ConstructorNode I switch back to calling the
>>>>ReportQuery method returnObjectsFromConstructor taking Class
>>>>instances for the constructor arguments (instead of mappings). I
>>>>understood that with your checkin from March 3 ("delayed ejbql query
>>>>parsing till login") this should work for constructor expressions in
>>>>named queries. Please have a look.
>>>>
>>>>The corresponding glassfish issue is:
>>>>https://glassfish.dev.java.net/issues/show_bug.cgi?id=310
>>>>
>>>>Thanks!
>>>>
>>>>Regards Michael
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>>
>>>
>
>
>
>
--
Tom Ware
Principal Software Engineer
Oracle Canada Inc.
Direct: (613) 783-4598
Email: tom.ware_at_oracle.com