[misc & build] replace isort pydocstyle and black with ruff

[felixdittrich92]

This PR:

• remove isort black and pydocstyle it's handled with ruff now
• improve and fix some docstrings
• fix for newest mypy Version
• fix docs: JSON export -> export

Any feedback is welcome 🤗
@odulcy-mindee this would require again to update the required ci job settings :)

[codecov]

Codecov Report

Attention: 1 lines in your changes are missing coverage. Please review.

Comparison is base (e00cc79) 95.78% compared to head (1339101) 95.78%.

Files	Patch %	Lines
doctr/datasets/detection.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1379      +/-   ##
==========================================
- Coverage   95.78%   95.78%   -0.01%     
==========================================
  Files         155      155              
  Lines        6952     6948       -4     
==========================================
- Hits         6659     6655       -4     
  Misses        293      293              

Flag	Coverage Δ
unittests	95.78% <97.77%> (-0.01%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

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

[odulcy-mindee]

I'm fine to use ruff but I'm not comfortable about dropping Python 3.8 version as it may have a wider impact.

[felixdittrich92]

I'm fine to use ruff but I'm not comfortable about dropping Python 3.8 version as it may have a wider impact.

I completely agree with you py3.8 end-of-life is in the fall of next year...however TF has already moved to >=py3.9 :/
ref: https://pypi.org/project/tensorflow/

[odulcy-mindee]

I completely agree with you py3.8 end-of-life is in the fall of next year...however TF has already moved to >=py3.9 :/

Yeah, I saw that after tf 2.14 :-/

[felixT2K]

I completely agree with you py3.8 end-of-life is in the fall of next year...however TF has already moved to >=py3.9 :/

Yeah, I saw that after tf 2.14 :-/

And 2.16 will break everything because 2.15 will be the last release with keras v2 in 2.16 (hopefully 3.0) it will integrate keras v3 (keras-core) and break the whole save and load logic (the way how we save and load tf checkpoints in doctr will not be anymore supported with keras 😅 )