Add brain for numpy core module einsumfunc.

[mbyrnepr2]

Refs pylint-dev/pylint#5821

Steps

• For new features or bug fixes, add a ChangeLog entry describing what your PR does.
• Write a good description on what the PR does.

Description

Type of Changes

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

Related Issue

[coveralls]

Pull Request Test Coverage Report for Build 2621848802

• 8 of 8 (100.0%) changed or added relevant lines in 1 file are covered.
• No unchanged relevant lines lost coverage.
• Overall coverage increased (+0.006%) to 92.124%

---

Totals
Change from base Build 2613849741:	0.006%
Covered Lines:	9439
Relevant Lines:	10246

---

💛 - Coveralls

[DanielNoord]

Not sure why this landed in PR limbo, but I thought I'd start with a review.

More general question: why are we putting al numpy brains into separate files? Seems like a bit of a waste, right? Or am I missing something?

[mbyrnepr2]

Not sure why this landed in PR limbo, but I thought I'd start with a review.

More general question: why are we putting al numpy brains into separate files? Seems like a bit of a waste, right? Or am I missing something?

Thanks!

I don't know the story behind this but to my mind it seems more maintainable to have them as separate modules, especially if they increase in size over time. Perhaps we could create a separate sub-package astroid/brain/numpy/ to collect them in one place?

Or perhaps your point is more related to performance @DanielNoord?

[DanielNoord]

I don't know the story behind this but to my mind it seems more maintainable to have them as separate modules, especially if they increase in size over time.

Yeah I think a refactor might be beneficial as I'm not sure this is actually more maintainable. For example, the brain you added could have been added to https://github.com/PyCQA/astroid/blob/main/astroid/brain/brain_numpy_core_fromnumeric.py.
That would have removed 17 unnecessary lines from this diff in the brain file alone.

Same probably goes for test files.

[mbyrnepr2]

Hi @DanielNoord I think you have good ideas. As a way forward may I suggest you create a new issue to capture the ideas behind your refactoring suggestion? I'm sure there are knowledgeable folks who are aware of the reasoning behind the existing structure who can provide input & help move that forward.

The current issue can move forward independently & according to the existing structure; if the refactoring issue comes to fruition then it wouldn't be much extra work to incorporate einsumfunc brain into the new system.

What do you think?

[DanielNoord]

I'm fine with creating a new issue, but I'm not sure if there is anybody around anymore that knows the original rationale.

Why don't approach this from what we would want ourselves. If you think keeping these separate is better that's fine with me, but if you think it's better to merge them then let's do so now!

[mbyrnepr2]

Mainly I think for now I'd like to follow the existing structure in order to stick to the scope of fixing the Pylint issue; hopefully I have time to finish that much this week.
Regarding the refactor itself - I'm in two minds because I think the structure now is helpful with the 1:1 of brain to numpy module mapping; but perhaps I am missing something.

[DanielNoord]

Pushed some final changes. Thanks @mbyrnepr2 😄