How are you pointing the unit test project against the mutated assemblies?

Jul 28, 2013 at 8:15 PM
I'm assuming that after generating the mutant(s), you'll then have to point the unit test suite against each mutated assembly in turn.

How is it that you are swapping out the default assembly, and pointing the unit test project against the mutated assemblies in turn?

I couldn't quite figure this out from the call.

Nov 18, 2013 at 3:39 PM
Sorry We missed this James.

I haven't worked on this for a while, but I'll do my best to answer.

It works (iirc) as follows:
  • Each mutated assembly is generated and placed in a temp directory with all the other assemblies and files that would be there in a regular test run
  • This assembly is similarly named to the original.
  • The unit testing framework is then run against the test assembly, with the new mutated assembly in place of the original
    • only the tests that affect the mutated code are run
  • Any non-failures are reported.
There are quite a few mutated assemblies that get generated out of this process, and the generation/run process is parallelised as much as possible.

I hope that goes some way to answering your question.