3 ms·
Maybe there's a place for some sort of a smart test runner that would execute assertions multiple times with random input, trying to achieve failure. If all att
by goblin89 15y ago
Maybe there's a place for some sort of a smart test runner that would execute assertions multiple times with random input, trying to achieve failure. If all attempts succeed then it'd mark test as questionable.
- AngryParsley 15y agoHave you heard of fuzzing?[1] 1. http://en.wikipedia.org/wiki/Fuzz_testing http://en.wikipedia.org/wiki/Fuzz_testing
- goblin89 15y agoI haven't, thanks for pointing out. Seems like a closely related concept, although appears to be used mostly in security testing context. Also, at least for Python most tools are a bit too complex—I imagined something more like a TestCase subclass (not even a test runner) that augments certain assert methods and possibly provides new ones. Anyway, on a second thought, adopting TDD looks like a better solution.
- ryanbraganza 15y agoThere's this: http://ruby.sadi.st/Heckle.html http://ruby.sadi.st/Heckle.html It changes your class, then runs your tests to make sure they fail. I haven't tried it myself.