fix: Fixing various bugs which appear with different libraries

[Masara]

Summary of Changes

• Fixed
  • various bugs to get the stub generator working with PyTorch, Scicit-Learn, Pandas and MatPlotLib.
  • parsing of result types.
  • a bug where already existing stubs content would be appended to the end of stub files again.
  • a bug for type parsing where subclasses of aliases couldn't be parsed.
  • a bug where some class attributes defined in the init function would wrongly be labeled as TypeVarType.
  • a bug where some class attributes would wrongly be labeled as TypeVarType.
  • a bug where type names would not be checked for naming convention and keywords.
  • a lot of bugs concerning the parsing of reexported files and directories.
• Added
  • parsing of different statement operator for types (such as "not", "is not", "is").
  • TypeAliasType parsing for parameters.
  • merging of text sections of docstrings, if a docstring has multiple.
• Replaced various raises with logger.infos.
• Furthermore I removed the unused EnumType and BoundaryType classes and added some short docstring descriptions to the code.

[github-actions[bot]]

🦙 MegaLinter status: ✅ SUCCESS

Descriptor	Linter	Files	Fixed	Errors	Elapsed time
✅ PYTHON	black	14	0	0	1.59s
✅ PYTHON	mypy	14		0	6.45s
✅ PYTHON	ruff	14	0	0	0.04s
✅ REPOSITORY	git_diff	yes		no	0.03s

See detailed report in MegaLinter reports _Set VALIDATE_ALL_CODEBASE: true in mega-linter.yml to validate all sources, not only the diff_

_MegaLinter is graciously provided by _

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 99.84%. Comparing base (0a8ef17) to head (fbe2a20).

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #168 +/- ## ========================================== + Coverage 99.57% 99.84% +0.26% ========================================== Files 26 26 Lines 2603 2618 +15 ========================================== + Hits 2592 2614 +22 + Misses 11 4 -7 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[Masara]

@lars-reimann Could you remind me how the stubs have to handle None as a return type? So for example these cases:

def fun1() -> None: ...

def fun2():
    return

def fun3():
    return None

Should the SDS-Stubs look like this for all three cases:

fun fun*()

? or should they all have a trailing -> result1: Nothing??

[Masara]

Answer to the question:

def fun1() -> None: ...

fun fun1()

def fun2():
    return

fun fun2()

def fun3():
    return None

fun fun3() -> result1: Nothing?

[Masara]

I decided to not change anything concerning the handling of "None" from the cases stated above, because it seems we already implemented everything concerning that topic already in the past. So the current state would be that every case creates the stubs without the -> result1: Nothing? part.

[Masara]

@lars-reimann This is ready for a review. I checked the stub changes on the other branch and everything seemed all right to me.