feat(python): Add infer_schema parameter to read_csv / scan_csv

[Julian-J-S]

fix #10342

[codecov]

Codecov Report

Attention: Patch coverage is 50.00000% with 2 lines in your changes missing coverage. Please review.

Project coverage is 80.68%. Comparing base (6816707) to head (4cf80ea).
Report is 5 commits behind head on main.

Files	Patch %	Lines
py-polars/polars/io/csv/functions.py	50.00%	1 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17617      +/-   ##
==========================================
+ Coverage   80.65%   80.68%   +0.03%     
==========================================
  Files        1484     1484              
  Lines      195511   195406     -105     
  Branches     2781     2781              
==========================================
- Hits       157685   157661      -24     
+ Misses      37315    37234      -81     
  Partials      511      511              

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

[ritchie46]

Thanks. I think we need to do the same for scan_csv.

[mcrumiller]

Random question, but what happens if someone uses infer_schema=True along with infer_schema_length=0? Same with infer_schema=False / infer_schema_length=n where n>0. Maybe a quick check with a raised exception? Something like infer_schema_length cannot be 0 when infer_schema is True?

[ritchie46]

0 is a valid length. That doesn't change. When you set infer_schema=False, we ignore the length.