This is exactly one of the situations where mocking shines; it can not only assert whether a method was called but it can also return fixed values (that you might normally get through a web service or somesuch). So for testing your calculation you’d have your object return fixed values (and in a different test suite you’d test that those object actually retrieve/calculate their values correctly.

Does that make sense?

Erik

I know this is kind of an old post, but I’ve been evaluating Telerik OA as well now and I like it too, but the one issue I’ve had is just this. Your example is a good one, but what if you needed to perform a calculation on multiple objects it’s much different than just asserting that a method was called or not. I’ll look at Rhino mocks as you mentioned maybe it can do this as well, but I wanted to see if you have gone done this path already.

Thanks,

Shane

I’m afraid I don’t quite get what your exact issue is. If you could provide some example code, that’d be great. My email is eburger at reversealchemy dot net.

Cheers,
Erik

I’ve also been using Rhino Mocks to avoid dependencies and focus on fine grained tests. Unfortunately, I’m a bit stuck when it comes to stubbing a [Persistent] class.

For example, if you had:

[Persistent]
public abstract class Person
{
…
}

Because it is abstract, it would be nice to call:

mockRepository.Stub()

but I always get an exception relating to
Telerik.OpenAccess.PersistentAttribute.get_IdentityClassName()

Any thoughts?
(A work-around is to create a derived class without the [Persistent] property and stub that, but I don’t want to do that for every domain class!)

Pete.

There is a pretty large amount of mocking frameworks available, all with their own strengths and weaknesses. The most popular for C# (as far as I know) are NMock, Rhino Mocks and TypeMock.

A good list can be found on Wikipedia, along with some interesting articles.

Hope this helps,

Erik