Martin Grotzke wrote:
> Hi Paul
>
> On Tue, 2008-06-10 at 14:05 +0200, Paul Sandoz wrote:
>> Hi Martin,
>>
>> The main issue we do not know at compile/deploy time what concrete
>> resources are returned by sub-resource locater methods.
> You're speaking of the cases when resource methods simply return a
> Response instead of a concrete type?
>
No a "sub-resource locator" not a "sub-resource method" :-) like the
sub-resource locator UsersResource.getUser.
This method could in the future also return a sub-type of UserResource,
say PreferredUserResource with additional URI space, say for a preferred
paying user.
I suppose if the return type of the sub-resource locator is recognized
as a resource we could do some processing on it even if it is not always
the concrete type returned.
> I thought that one should be able to specify several response
> representations in the javadoc so that application.wadl _can_ be
> generated at build time.
>
>> This is why we support the serving of WADL via the HTTP OPTIONS method
>> for each resource.
> Ok, sounds interesting, but I have no clue how this is realized :)
>
See:
com.sun.jersey.impl.wadl.WadlMethodFactory
Paul.
--
| ? + ? = To question
----------------\
Paul Sandoz
x38109
+33-4-76188109