Better dataset_id inference and minor fullres image fix

[quentinblampey]

When dataset_id is None, we now try to see if dataset_id="" works, and, if not, try to infer the dataset_id.
Previously, we directly inferred a (non-empty) dataset_id, which was therefore not working when dataset_id="".

I also made a small fix regarding the fullres image:
We had two variables fullres_image_file_paths and fullres_image_paths, causing an index error in load_image(path=fullres_image_file_paths[0], ...) when fullres_image_paths is used but not fullres_image_file_paths.

FYI @LucaMarconato @grst

[codecov-commenter]

Codecov Report

❌ Patch coverage is 55.55556% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 55.07%. Comparing base (d5fa3f2) to head (3c7425e).

Files with missing lines	Patch %	Lines
src/spatialdata_io/readers/visium_hd.py	55.55%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #352      +/-   ##
==========================================
- Coverage   55.16%   55.07%   -0.09%     
==========================================
  Files          26       26              
  Lines        2810     2809       -1     
==========================================
- Hits         1550     1547       -3     
- Misses       1260     1262       +2     

Files with missing lines	Coverage Δ
src/spatialdata_io/readers/visium_hd.py	75.08% <55.55%> (-0.76%)	⬇️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[LucaMarconato]

Thanks! I made some minor changes:

• removed the assert and added a raise
• added a docstring
• changed a bit the logic since tests were failing since dataset_id (used to name the elements) was becoming "" in the new code when it was not "". Now if dataset_id is not None and the slice file with no prefix is detected, the prefix becomes "", but dataset_id remains intact.