Move c_coords out of kwargs

[melonora]

This PR implements one of the tasks in #750 to improve the handling of channel names. It moves the c_coords parameter out of kwargs and also adds a validation for a mismatch in length between c_coords and the length of the c dimension.

[codecov]

Codecov Report

Attention: Patch coverage is 88.88889% with 1 line in your changes missing coverage. Please review.

Project coverage is 91.84%. Comparing base (42f7b6a) to head (088dec2).
Report is 68 commits behind head on main.

Files with missing lines	Patch %	Lines
src/spatialdata/models/models.py	88.88%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #779      +/-   ##
==========================================
- Coverage   91.85%   91.84%   -0.01%     
==========================================
  Files          45       45              
  Lines        6885     6893       +8     
==========================================
+ Hits         6324     6331       +7     
- Misses        561      562       +1     

Files with missing lines	Coverage Δ
src/spatialdata/datasets.py	100.00% <ø> (ø)
src/spatialdata/models/models.py	87.76% <88.88%> (-0.01%)	⬇️

[melonora]

still need to see how we can have c_coords only in docstring when dealing with images and not labels.

[LucaMarconato]

still need to see how we can have c_coords only in docstring when dealing with images and not labels.

Does this work now?

Btw, looks great, pre-approving. Only missing thing, please add some tests (testing that it works for images and testing that it gives an error for labels, and for images when there is a mismatch). Then feel free to merge, thanks!

[melonora]

Main problem is that the docstring must be a literal so you can't use f strings or anything.