Refactor/put start and end on

[GoThrones]

Overview: What does this pull request change?

This PR refactors the method "put_start_and_end_on" in mobject.py and opengl_mobject.py:

Replaces self.points = np.array(start) with self.shift(np.asarray(start) - current_start) in the case of a mobject whose start and end point are the same, in both mobject.py and opengl_mobject.py

Adds a warning that warns the user that the mobject which is calling this method is a loop, and it will be shifted.

Some variable names have been changed for better reading.

Motivation and Explanation: Why and how do your changes improve the library?

The put_start_and_end_on method in both mobject.py and opengl_mobject.py previously used self.points = np.array(start) when the mobject had its start and end as the same point. This collapsed all of the mobject's points into a single Point3D, destroying its geometry entirely. For example, calling this method on a Circle or Square would reduce it to a single point, making it cease to exist visually.
This PR fixes the issue by replacing that line with self.shift(np.asarray(start) - current_start), which simply repositions the mobject to the start point while fully preserving its geometry.

Reviewer Checklist

• The PR title is descriptive enough for the changelog, and the PR is labeled correctly
• If applicable: newly added non-private functions and classes have a docstring including a short summary and a PARAMETERS section
• If applicable: newly added functions and classes are tested

[chopan050]

Thanks for your contribution! Shifting the Mobject in this case makes sense, and I appreciate both expanding the variable names and making the behavior consistent in Mobject and OpenGLMobject.

I left some change requests:

[GoThrones]

I've made necessary changes in the code as suggested by you.

[chopan050]

Thanks for applying the requested changes!

Now most pipelines do not fail because of that test. However, precommit is still failing for a different reason: the current use of warnings.warn(). Please address this final suggestion in order to be able to merge this:

[chopan050]

precommit demands that the stacklevel parameter is explicitly set when emitting a warning. It suggests a value of 2, which is the best choice in this case:

Suggested change

	warnings.warn(
	"put_start_and_end_on has been called on a closed loop or zero-length mobject. "
	f"{type(self).__name__} will be shifted to start point instead."
	)
	warnings.warn(
	"put_start_and_end_on has been called on a closed loop or zero-length mobject. "
	f"{type(self).__name__} will be shifted to start point instead.",
	stacklevel=2,
	)

[chopan050]

Suggested change

	warnings.warn(
	"put_start_and_end_on has been called on a closed loop or zero-length mobject. "
	f"{type(self).__name__} will be shifted to start point instead."
	)
	warnings.warn(
	"put_start_and_end_on has been called on a closed loop or zero-length mobject. "
	f"{type(self).__name__} will be shifted to start point instead.",
	stacklevel=2,
	)