>
> that raises the question of whether we support enum-mapped
> properties as the discriminators, and if so, what's the proper
> syntax for specifying the enum constant in the qualifier field?
We should. This comes down to adding enum literals support in the
"exp" package. Maybe:
xyz = org.example.EnumClass.ENUM_NAME
This is also seems to be the JPA (EJBQL) syntax for enum literals.
Andrus
On Apr 24, 2009, at 9:51 AM, Robert Zeigler wrote:
> This was with a SelectQuery already.
>
> Just verified that you're right: this is a problem with my qualifiers.
>
> The column type is varchar. I was trying to map the object property
> as an enum. But I wasn't sure how to specify the qualifier, so I
> tried as:
>
> entryType = 'ENUM_CONST_NAME'
>
> Which failed. When I changed the java type of entryType to string,
> cayenne properly instantiated the subclasses.
> So I suppose that raises the question of whether we support enum-
> mapped properties as the discriminators, and if so, what's the
> proper syntax for specifying the enum constant in the qualifier field?
>
> Robert
>
> On Apr 24, 2009, at 4/241:23 AM , Andrus Adamchik wrote:
>
>>
>> On Apr 24, 2009, at 5:35 AM, Robert Zeigler wrote:
>>
>>> No, it's a bit more complicated. The template is one portion of
>>> the fix, but even if the entity is marked abstract (in both the
>>> mapping and the java class), cayenne will try to instantiate it in
>>> the aforementioned query on superclass case. So a complete
>>> solution will require that case to resolve and instantiate the
>>> subclass types, rather than attempting to instantiate the
>>> superclass type.
>>>
>>> Robert
>>
>> Hi Robert,
>>
>> The whole logic for instantiating an object of a given class is
>> based on the combination of entity qualifiers in the hierarchy
>> (including on the base class). So unless this is a bug specific to
>> EJBQLQuery (could you try the same thing with SelectQuery?), maybe
>> the condition on a given result set row yielded a base class somehow?
>>
>> Could you post the stack trace and entity qualifiers?
>>
>> Thanks,
>> Andrus
>>
>
>
This archive was generated by hypermail 2.0.0 : Fri Apr 24 2009 - 03:03:47 EDT