Then it would incorrectly default to the same SHA. This PR fixes that behavior, it also detects this failure mode and raises an error if it happens somehow (perhaps the user accidentally puts it in twice somehow).
In addition there's a nice litte testing output that advertises which sha and description you're testing before you execute the tests:
When using perf:library there was a bug that would result in the same sha being tested twice for instance if you ran:
Then it would incorrectly default to the same SHA. This PR fixes that behavior, it also detects this failure mode and raises an error if it happens somehow (perhaps the user accidentally puts it in twice somehow).
In addition there's a nice litte testing output that advertises which sha and description you're testing before you execute the tests: