GH-145762: Fix test_ci_fuzz_stdlib when running the tests after building from the source archive#145778
GH-145762: Fix test_ci_fuzz_stdlib when running the tests after building from the source archive#145778chris-eibl wants to merge 1 commit intopython:mainfrom
Conversation
when running the tests after building from the source archive
| f = p | ||
| else: | ||
| self.fail(f"LIBRARY_FUZZER_PATHS contains an invalid entry: {p!r}") | ||
| continue |
There was a problem hiding this comment.
@StanFromIreland do we still want to log this, though? Or perhaps have a mechanism for skipping checks / checking against different sets of files in certain envs?
There was a problem hiding this comment.
Or perhaps have a mechanism for skipping checks / checking against different sets of files in certain envs?
As I noted in the issue, this will greatly complicate the test (and increase the maintenance burden, who knows what envs people use), for something that is not particularly important. As such, I think skipping is fine, logging them too, however, sounds good.
There was a problem hiding this comment.
"Simple prints" will get lost in the noise of the output, and be false-positives in the OPs case.
I don't think
print(f"Warning: LIBRARY_FUZZER_PATHS might contain an invalid entry: {p!r}")will do any good.
So I think this is fine to go. It was the previous behaviour, anyway. I just thought it is a low hanging fruit to crop when the else was removed in #145232 and we had spurious failures ...
3 participants
I assume skipping news is ok, because only the tests are affected.
test_toolsfailing overAssertionError: LIBRARY_FUZZER_PATHS contains an invalid entry: PosixPath('.github/workflows/reusable-cifuzz.yml')#145762