Fix contained items states

[DA-344]

Summary

Contained items (in Containers, ActionRows, etc.) had some attributes outdated or uncorrectly set as another value. This PR attempts to fix it.

Checklist

• If code changes were made then they have been tested.
  • I have updated the documentation to reflect the changes.
• This PR fixes an issue.
  • resolves [V2 components] Weird nesting behavior #10335
• This PR adds something new (e.g. new method or parameters).
• This PR is a breaking change (e.g. methods or parameters removed/renamed)
• This PR is not a code change (e.g. documentation, README, ...)

[laggron42]

Thanks a lot for the patch!

I have been writing test cases on my end to help covering the weird situations, and I'm still failing on your branch checking for ActionRow.parent which should be Container but is None (same example as the issue).
I also noticed that button.parent and self inside the button callback are of the same type (ActionRow), but not the same instance. It's probably due to all the deepcopy going on, I don't know how problematic that actually is.
My bad, my venv was not properly configured and did not use the right code 😓 The only case failing is self.parent.parent being None instead of Layout

Download this file, drop it in the tests folder, and run it with pytest -k test_ui_nested. I think it's worth pushing the test along the PR to avoid regression.

test_ui_nested.py

[laggron42]

add_item fails because it needs self._weight to exist

Suggested change

	for child in children:
	self.add_item(child)
	self._weight: int = sum(i.width for i in self._children)
	self._weight: int = sum(i.width for i in self._children)
	for child in children:
	self.add_item(child)

[DA-344]

A view will never be a parent, that is why parent is set to None when it is at layout level (and the reason why view exists).

[laggron42]

Oooh, that makes sense alright! Then all of my tests pass

[Rapptz]

I'm not sure how I feel about this fix honestly. It's not incorrect but it does feel a bit messy since there are now multiple places where the callback binding needs to be handled. I'm not a fan of that right now though I guess a better approach can be figured out later.

As proof of how finicky this stuff is to deal with now, you forgot to do this to Select components.