Sebastian Rittau
Sebastian Rittau
There's a merge conflict in the allowlist now.
I've added two suggestions that will hopefully help.
Looking at the code in mypy_test, it seems that we are not actually handling sub-modules at all when it comes to the `VERSIONS` file. I will send a PR.
The mypy errors are gone now. I have some ideas about fixing the pyright test, but I need to find some time in the next few days.
I think the easiest solution for now is to add `stdlib/importlib/resources/_functional.pyi` to the exclude list of `pyrightconfig.stricter.json` (with an appropriate comment).
Regarding "deprecations", not general type warnings, I'll repeat my points from the other discussion with one addition: ---- There are multiple reasons why deprecations are better than just removing functions:...
From reading the discussion, I think the consensus is to be fairly liberal when using `@deprecated`. I think it was mainly @AlexWaygood dissenting? One open question that wasn't discussed is...
See above, I've amended the suggestion.
Here's your example modified a bit to use star imports: ```python from hashlib import * from sys import path as md5 # error: Incompatible import of "md5" (imported name has...
It seems that due to the new union element, pyright is not able to statically determine the type of the empty dict in the tests anymore. I don't know know...