As I had mentioned, support for some form of dynamic response was the last feature I had wanted to get into the stub module, so I am certainly not opposed to this. I had simply thought to eat dog food by using [functor] interfaces, but that's not a big deal. I have not yet reviewed your latest work, but I really don't want to lose the fluent and typesafe API that exists currently.
Matt On Jul 27, 2013 11:17 AM, "James Carman" <ja...@carmanconsulting.com> wrote: > I have created > > https://issues.apache.org/jira/browse/PROXY-20 > > to track the progress of this issue. I have already checked in some > code into the branch. > > > On Sat, Jul 27, 2013 at 9:54 AM, James Carman > <ja...@carmanconsulting.com> wrote: > > On Sat, Jul 27, 2013 at 9:44 AM, Romain Manni-Bucau > > <rmannibu...@gmail.com> wrote: > >> Hi > >> > >> Isnt it a particular kind of interceptor/handler > (CompositeInterceptor)? So > >> does it need so much details? > > > > Well, the idea behind "stubbing" is that we would be specifying > > behavior for very specific method invocation cases (such as only a > > single method or only when the parameters match certain values). If > > we introduce the abstractions I'm suggesting, we can handle these > > cases and many more. I don't know if we need the Response concept, > > since the interface exactly matches the Interceptor interface. > > Perhaps we call the new class ConditionalInterceptor and you register > > InvocationMatcher/Interceptor pairs just so we don't have to re-invent > > the wheel. > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org > For additional commands, e-mail: dev-h...@commons.apache.org > >
