Use inference to determine if **kwargs is missing a named parameter

[jacobtylerwalls]

I would expect these two lines to raise the same set of messages. We don't use inference to analyze whether **kwargs contains the missing named parameter x. We could.

import copy
import os

copy.copy(**{"y": os.environ})  # [unexpected-keyword-arg]
copy.copy(y=os.environ)  # [no-value-for-parameter, unexpected-keyword-arg]

False negative for no-value-for-parameter.

[Pierre-Sassoulas]

We did that recently for #8719, we could design a generic solution for checker that analyses args / kwargs,

[jacobtylerwalls]

Yep, the aftermath of that PR is what led me to #8775 and then #8774 and then this :-)

[jacobtylerwalls]

I think a broader refactor is a separate issue, though. This issue could be as simple as just removing this assumption:

pylint/pylint/checkers/typecheck.py

Lines 1616 to 1622 in 7eaac57

	# 3. Match the **kwargs, if any.
	if node.kwargs:
	for i, [(name, _defval), _assigned] in enumerate(parameters):
	# Assume that *kwargs provides values for all remaining
	# unassigned named parameters.
	if name is not None:
	parameters[i] = (parameters[i][0], True)