Fix FunctionDef isinstance checks

[cdce8p]

Description

Fallout from pylint-dev/astroid#2115 - required for the next astroid update.
Fix isinstance checks now that nodes.FunctionDef doesn't inherit from nodes.Lambda.

Changes are compatible with the current astroid version, so don't need to wait to fix these.

[codecov]

Codecov Report

Merging #8607 (3bc915a) into main (3d036b7) will not change coverage.
The diff coverage is 100.00%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #8607   +/-   ##
=======================================
  Coverage   95.92%   95.92%           
=======================================
  Files         174      174           
  Lines       18415    18415           
=======================================
  Hits        17664    17664           
  Misses        751      751           

Impacted Files	Coverage Δ
pylint/checkers/typecheck.py	96.46% <100.00%> (ø)
pylint/checkers/utils.py	96.05% <100.00%> (ø)
pylint/checkers/variables.py	97.25% <100.00%> (ø)

[jacobtylerwalls]

Thanks! Just so all reviewers are aware, we will still need another round of cleanup after bumping to astroid 2.16, when we could remove things like this last line:

pylint/pylint/checkers/typecheck.py

Lines 676 to 678 in 3d036b7

	if (
	isinstance(scope, nodes.Lambda)
	and not isinstance(scope, nodes.FunctionDef)

Either then or now, we could tidy up some comments like this:

There's also a comment to clean up in `_determine_callable(), but we can do that later:

pylint/pylint/checkers/typecheck.py

Lines 614 to 615 in 3d036b7

	# Ordering is important, since BoundMethod is a subclass of UnboundMethod,
	# and Function inherits Lambda.

[DanielNoord]

Yeah, I think this change is very likely to cause a need to do a 3.0 and perhaps 4.0 quickly after it. Then again, still think it is a good change.

Thanks @cdce8p and @jacobtylerwalls for taking these first steps towards compatibility!

[jacobtylerwalls]

Yeah, I think this change is very likely to cause a need to do a 3.0 and perhaps 4.0 quickly after it. Then again, still think it is a good change.

Agree about 3.0. Let's slow the release cadence down for a little bit, and release both 3.0 in pylint and astroid when we're ready for python 3.12. We may not need a 4.0 then.

[DanielNoord]

Yeah, I think this change is very likely to cause a need to do a 3.0 and perhaps 4.0 quickly after it. Then again, still think it is a good change.

Agree about 3.0. Let's slow the release cadence down for a little bit, and release both 3.0 in pylint and astroid when we're ready for python 3.12. We may not need a 4.0 then.

Also fine with me, although we should definitely release 3.0a1 soon. I'd like to see behaviour in pylint as I'm not all that sure about the changes to some of the constructors. I think we're making definitive improvements but might have been too strict in some places.

[Pierre-Sassoulas]

+1, we can keep the release cadence by releasing alpha/betas and still wait to support python 3.12 for pylint 3.0. There's a lot to do in #3512 still, we'll probably release pylint 3.0 after python 3.12 is out in any case.

[github-actions]

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

This comment was generated for commit 3bc915a