> You get the picture, also, it's usually best practice to limit your > assertions to one per test case.
In general, I agree, but I also find it's often useful to have multiple assertions in one test case, where the test case maps to a business function, and you need information about the whole business function. My example is for SOAP API functions. I have one test method for each API function available. Each test method contains a number of assertions about that function. That buys me two things: first, a failure tells me that the entire API function is suspicious and needs investigation. Second, a failure for one API function does not stop the tests for the other functions from running, and I don't have to build try/catch logic into individual assertions. _______________________________________________ Wtr-general mailing list [email protected] http://rubyforge.org/mailman/listinfo/wtr-general
