Fix false positive for positional-only-arguments-expected

[mbyrnepr2]

Type of Changes

	Type
✓	🐛 Bug fix
	✨ New feature
	🔨 Refactoring
	📜 Docs

Description

Fix false positive for positional-only-arguments-expected when a
function contains both a positional-only parameter that has a default value, and **kwargs.

Closes #8555

[codecov]

Codecov Report

Merging #8556 (6e4d89f) into main (82cee37) will increase coverage by 0.00%.
The diff coverage is 100.00%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #8556   +/-   ##
=======================================
  Coverage   95.90%   95.91%           
=======================================
  Files         174      174           
  Lines       18353    18370   +17     
=======================================
+ Hits        17602    17619   +17     
  Misses        751      751           

Impacted Files	Coverage Δ
pylint/checkers/method_args.py	100.00% <100.00%> (ø)
pylint/checkers/variables.py	97.17% <100.00%> (+0.01%)	⬆️

... and 3 files with indirect coverage changes

[github-actions]

🤖 According to the primer, this change has no effect on the checked open source code. 🤖🎉

This comment was generated for commit 6e4d89f

[jacobtylerwalls]

Nice edge case! I find this profoundly weird, and honestly, after reading PEP 570 a couple times, I feel like this could even be reported to CPython as a bug. But even if it's not a bug, it should be a pylint warning on the function def, not a pylint error on the call.

However, I did find this regression, based on the example in PEP 570 ("Semantic Corner Case"). Would you be able to fix and add a test case?

def foo(name, /, **kwds):
    return 'name' in kwds

foo(name=3)

Gives:

TypeError: foo() missing 1 required positional argument: 'name'

main emits:

corner.py:4:0: E3102: `foo()` got some positional-only arguments passed as keyword arguments: 'name' (positional-only-arguments-expected)

PR emits nothing.

[jacobtylerwalls]

Opened #8558 to track emitting a different message for the original reported case.

[mbyrnepr2]

However, I did find this regression, based on the example in PEP 570 ("Semantic Corner Case"). Would you be able to fix and add a test case?

I don't think this is a regression. The TypeError you've identified is missing 1 required positional argument whereas this checker deals with the TypeError got some positional-only arguments passed as keyword arguments.
This is similar behaviour to the current fix in the sense that name=3 is collected by **kwds and hence the name argument is indeed missing. Here would be a working example:

def foo(name, /, **kwds):
    print(kwds)
    return 'name' in kwds

>>> foo(1, name=3)
>>> {'name': 3}

[jacobtylerwalls]

Ah, I see, makes sense.