Add __init__ to the ObjectModel and return BoundMethods

[DanielNoord]

Steps

• Write a good description on what the PR does.

Description

Type of Changes

	Type
✓	🔨 Refactoring

Related Issue

Blocked by #1686
Precedes #1519

@jacobtylerwalls Please only consider the last commit. This is the next step in "completing" our ObjectModel. Since all objects have an __init__ that normally returns None we can define attr___init__ on ObjectModel.

[coveralls]

Pull Request Test Coverage Report for Build 2622711369

• 21 of 21 (100.0%) changed or added relevant lines in 1 file are covered.
• 4 unchanged lines in 3 files lost coverage.
• Overall coverage decreased (-0.02%) to 92.127%

---

Files with Coverage Reduction	New Missed Lines	%
astroid/interpreter/objectmodel.py	1	92.62%
astroid/nodes/node_ng.py	1	92.81%
astroid/bases.py	2	87.35%

Totals
Change from base Build 2622607134:	-0.02%
Covered Lines:	9443
Relevant Lines:	10250

---

💛 - Coveralls

[DanielNoord]

After testing on pylint I found that there were still issues with the previous implementation.

I think now I (finally) fixed those.

I was mistaken to think that the other attr___ methods are correct. Rather than returning the value directly these methods should return the BoundMethod of the method that is going to be called. The functools brain uses a somewhat similar trick:
https://github.com/PyCQA/astroid/blob/58750a6f982c20fd7224ac396c95ee8adbbfa34f/astroid/brain/brain_functools.py#L52-L55

We will infer the result of these BoundMethods succesfully by calling infer_call_result on them, but for some cases we need the BoundMethod itself. For example for the following code A().__init__ #@. That shouldn't be a nodes.Const, but the BoundMethod that will return nodes.Const.

[jacobtylerwalls]

Looks good!

[jacobtylerwalls]

This feels like a new feature to some degree. What do you think about a short changelog?

[jacobtylerwalls]

🎉