[not-an-iterable] Add regression test for uninferable instances

[jacobtylerwalls]

Type of Changes

	Type
✓	🔨 Refactoring

Description

While investigating #9253, I found that the regression test for suppressing not-an-iterable for Uninferable instances was not testing what it intended. Inference had improved over eight years, and we were testing an Instance instead, not Uninferable.

Refs #9253

[codecov]

Codecov Report

Merging #9264 (fc83c51) into main (b468e5b) will not change coverage.
Report is 1 commits behind head on main.
The diff coverage is n/a.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #9264   +/-   ##
=======================================
  Coverage   95.81%   95.81%           
=======================================
  Files         173      173           
  Lines       18761    18761           
=======================================
  Hits        17975    17975           
  Misses        786      786           

[Pierre-Sassoulas]

Are we inferring successfully that something is Iterable from a module we can't import because the namekf the class is 'Iterable' ? This would fail on deceiving code, right ? (but on the other hand it's probably working very well on Real World code)

[jacobtylerwalls]

From the comment, I took the motivation for the test to be testing an Uninferable. We weren't getting that. We were getting <Instance of blahblah.MyClass>, and testing that. We can keep the existing test if you like, I thought it was duplicative with the other tests in this module. The failed import is a red herring; it has nothing to do with the test, since we get <Instance of blahblah.MyClass> whether the import succeeds or not.

[jacobtylerwalls]

Oh! The bases being able to be inferred actually does matter. I'll just keep both tests (and improve the comment).``

[Pierre-Sassoulas]

Thanks for investigating 👍

[jacobtylerwalls]

Rebased for the docs check fix

[Pierre-Sassoulas]

Nice github enhancement I just noticed: rebase + force push don't trigger the need for a new approving review !

[jacobtylerwalls]

🙏 https://github.com/orgs/community/discussions/12876#discussioncomment-7445850