Hi folks,
I noticed that after setting this up, tests which don’t start with Test don’t get parsed any more. This includes notably benchmarks as well as a custom test runner we have (whose tests have a different naming scheme).
Is this behavior reasonable? IMO if the JSON output gives you a test you should parse it, regardless of the name. Right now I’m forced into awkward workarounds (that in turn break other parts of our tooling that expect the real name).
Thanks,
Tobias