New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix np.MachAr
warning matching in test.
#9362
Conversation
Somehow numba#9347 changed the behavior of the filter. The invisible characters from termcolor highlight now matters. But comparing the warning message before and after that PR does not reveal any visible difference. Also added a simplefilter("ignore") to suppress any warning config from PYTHONWARNINGS in the python process.
To test this, make sure the environment has Run: |
BFID: |
Thanks for fixing this @sklam. I debugged this a bit, AFAICT the actual strings in the message are the same, what I think is changing is what happens internally as part of the matching process, now the Numba warnings inherit from the standard library warnings the way the message strings are normalized internally is different: |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks for the patch. I tested main
manually against the reproducer in #9362 (comment) and the test fails, application of this patch then makes it pass. The changes also make sense with respect to some debugging I did #9362 (comment).
CI passed and enough buildfarm config has passed (e.g just windows ones are taking a long time and we don't want to wait for it) |
Somehow #9347 changed the behavior of the filter.
The invisible characters from termcolor highlight now matters. But comparing the warning message before and after that PR does not reveal any visible difference.
Also added a
simplefilter("ignore")
to suppress any warning config fromPYTHONWARNINGS
in the python process.