feat(python): Add name parameter to GroupBy.len method

[alexander-beedie]

Closes #15222.

Seems like a simple/reasonable addition? While adding tests for this, I also found/fixed a bug where a column name is legitimately the empty string, and we reference it from schema_overrides; was getting incorrectly renamed as "column_0".

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 81.29%. Comparing base (103bb66) to head (33349e2).
Report is 27 commits behind head on main.

❗ Current head 33349e2 differs from pull request most recent head 73f16aa. Consider uploading reports for the commit 73f16aa to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #15235      +/-   ##
==========================================
+ Coverage   81.25%   81.29%   +0.04%     
==========================================
  Files        1354     1356       +2     
  Lines      175390   175692     +302     
  Branches     2518     2525       +7     
==========================================
+ Hits       142505   142830     +325     
+ Misses      32404    32379      -25     
- Partials      481      483       +2     

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

[stinodego]

I had to think on this one a bit, but I think it's a valid addition. The default name can clash, so if you're grouping by a column named len, you cannot use the GroupBy.len method.

Since this is supposed to be a syntactic sugar / ease of use function, it might as well be easy to use... or you might argue the other way and say it should be simple, and for more complex use cases, do not use the syntactic sugar.

Anyway, I think it should be added. If you can address the 2 comments this can be merged!

[stinodego]

Great, thanks!