I have this problem, which may be a bug in Rhino Mocks 3.5:
stubObj = MockRepository.GenerateStub(IObject);
stubObj.Stub(a=>a.Get()).Return (Guid.Empty); //1.stub
stubObj.Stub(a=>a.Get()).Return (Guid.NewGuid()); //2.stub, should overwrite the first one?
this:
var value = stubObj.Get();
returns Guid.Empty
, is this the correct behavior?
Stub: a dummy piece of code that lets the test run, but you don't care what happens to it. Substitutes for real working code. Mock: a dummy piece of code that you verify is called correctly as part of the test. Substitutes for real working code.
Rhino. Mocks supports three basic types of mock objects: Strict Mock. A strict mock requires you to provide alternate implementations for each method/property that is used on the mock.
If you want to return the empty guid a known number of times, the only thing you have to do is to tell RhinoMocks how many times to repeat it, e.g. the following test passes:
[Test]
public void MultipleStubsTest()
{
var testMock = MockRepository.GenerateMock<ITest>();
testMock.Stub(x => x.Get()).Return(Guid.Empty).Repeat.Once();
testMock.Stub(x => x.Get()).Return(Guid.NewGuid());
Assert.AreEqual(Guid.Empty, testMock.Get());
Assert.AreNotEqual(Guid.Empty, testMock.Get());
}
if you don't know how many times Get() will be called before the guid should change, than you can always use .Do() and code it there (please let me know if you need more details).
You have just programmed the Stub object for two separate calls. If you call stubObj.Get again, you should get what Guid.NewGuid
generated. You can prepare your fake object for any number of invocations of different kinds. For this reason, it doesnt make sense to expect the last .Stub
call for a given invocation to replace previous .Stub
bings of that call.
In your test code, which should be short and neat, there should never be a case where you need to 'undo' such programming of the mock in the way you seem to want to do.
If what needs to be returned is a conditional thing which varies depending on other bits of your test code across multiple calls to this block of code, the last thing you want is magic happening to make readers have to figure out what you meant. If it's conditional, you should make it clear.
And then, when you've made it clear, refactor it out as you should not have Conditional Logic in Tests (see xUnit Test Patterns)
That's how you would do it. It passes on my machine.
[TestMethod]
public void Test()
{
stubObj = MockRepository.GenerateMock<IGuidTest>();
stubObj.Stub(a => a.Get()).Repeat.Times(1).Return(Guid.Empty);
stubObj.Stub(a => a.Get()).Repeat.Times(1).Return(Guid.NewGuid());
Assert.AreEqual(Guid.Empty, stubObj.Get());
Assert.AreNotEqual(Guid.Empty, stubObj.Get());
}
public IGuidTest stubObj;
public interface IGuidTest
{
Guid Get();
}
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With