attr -> pydantic

[Stoops-ML]

Replace attrs with pydantic.

• Type checking of every argument of every class
• Simplifies dumping to json string
• Bumps version to 2.0.0 as there are breaking changes
• Supports Python 3.11 upwards due to using type hints Self and StrEnum. StrEnum is very useful as all classes in enums.py can be used with auto() to return the exact member. Self isn't critical and less clean alternatives can be found for replacing StrEnum in order to support Pythons <3.11.
• Tests have been modified minimally to account for small changes in output (e.g. spaces in how pydantic performs the model dump). All tests pass in this current PR for Pythons 3.11+.

@astrojuanlu Let me know your thoughts.

[codecov]

Codecov Report

Attention: Patch coverage is 99.30556% with 5 lines in your changes missing coverage. Please review.

Project coverage is 99.44%. Comparing base (56632e3) to head (1b4ead7).
Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
src/czml3/types.py	98.16%	4 Missing ⚠️
src/czml3/enums.py	96.42%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #154      +/-   ##
==========================================
+ Coverage   99.28%   99.44%   +0.16%     
==========================================
  Files          12       11       -1     
  Lines         837      898      +61     
==========================================
+ Hits          831      893      +62     
+ Misses          6        5       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[astrojuanlu]

Oh I love this idea! I think it makes a ton of sense. Cannot look into the code just yet but ping me again when the CI is green

[Stoops-ML]

What's your decision on typing? If we're using Self and StrEnum then we're limited to Python 3.11+ (see my original post).

[astrojuanlu]

If we're using Self and StrEnum then we're limited to Python 3.11+

Aren't there backports in typing_extensions or similar?

[Stoops-ML]

Python 3.7 support has been dropped as Pydantic>=2.10.0 doesn't support it.

@astrojuanlu

[astrojuanlu]

Good for me, 3.7 is EOL anyway. So is 3.8, by the way. Let's drop both.

I see you replaced | with Union everywhere, is that necessary?

[Stoops-ML]

Python 3.8 dropped.

pydantic doesn't work with | on Python 3.9 even with from __future__ import annotations specified. Someone noted the problem on SO here.

Want to drop Python 3.9 or implement Union?

[astrojuanlu]

Hm interesting. Well, let's drop Python 3.9 then!

[Stoops-ML]

All checks passed!

[astrojuanlu]

LGTM!