Or, How to test your Presenter classes

A couple of colleagues and I recently finished up a project for a client that involved a lot of C# code. Our unit testing tools of choice were NUnit and Rhino Mocks. The NUnit choice was a pretty easy one (although there are other platforms out there), but we spent a little more time choosing a mocking library.

Our first choice was NMock, primarily because we had experience using it on previous projects and we knew it could get the job done. Before too long, though, we switched to using Rhino Mocks. The primary reason for the change (if I am remembering correctly) was because of its superior event handling capabilities, and we grew to prefer its explicit use of the record-playback-verify mocking model.

Rhino Mocks was one of the first libraries to directly support event registration and event raising by mocks (at least, it was the first we had used). This was a big advantage over having to add a SubscribeEvent() method to our view and model interfaces and having to use syntactically obscure paradigms to capture and fire the event raisers. With Rhino Mocks we could add public events directly to our interfaces and (fairly) explicitly capture the event raisers.

Read the rest of this entry