fix[doc]: Render long method names on multiple lines (#4591)

[ghuls]

Render long method names on multiple lines, so no sidewards
scrolling is needed to see them.

[codecov-commenter]

Codecov Report

Merging #4604 (6fb2ac2) into master (8bc22e0) will decrease coverage by 0.02%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #4604      +/-   ##
==========================================
- Coverage   79.41%   79.38%   -0.03%     
==========================================
  Files         499      499              
  Lines       77923    77923              
==========================================
- Hits        61885    61863      -22     
- Misses      16038    16060      +22     

Impacted Files	Coverage Δ
polars/polars-core/src/chunked_array/ops/zip.rs	64.64% <0.00%> (-12.13%)	⬇️
...polars-core/src/chunked_array/iterator/par/list.rs	95.23% <0.00%> (-4.77%)	⬇️
...s/polars-core/src/chunked_array/ops/take/traits.rs	64.51% <0.00%> (-4.31%)	⬇️
.../polars-lazy/src/physical_plan/expressions/cast.rs	79.38% <0.00%> (-3.10%)	⬇️
polars/polars-core/src/chunked_array/cast.rs	85.41% <0.00%> (-0.70%)	⬇️
polars/polars-core/src/series/from.rs	77.12% <0.00%> (-0.59%)	⬇️
...s/polars-core/src/chunked_array/ops/unique/rank.rs	95.50% <0.00%> (-0.35%)	⬇️
polars/polars-time/src/windows/groupby.rs	95.20% <0.00%> (-0.30%)	⬇️
polars/polars-core/src/utils/mod.rs	87.46% <0.00%> (-0.30%)	⬇️
polars/polars-core/src/vector_hasher.rs	73.71% <0.00%> (-0.26%)	⬇️
... and 10 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[stinodego]

Do you have a screenshot of what this looks like? I'm worried that vertical scrolling would be even worse when these names are wrapped.

[ghuls]

It is not the most beautiful thing, but as long as internal names are displayed, it is nicer to be able to see the actual name than just the same prefix.
On my screen it only affects, polars.internal.series.*, polars.internal.dataframe.* and polars.internal.lazyframe.*

[ghuls]

[alexander-beedie]

Cunning workaround :)
Was surprised to discover that there is no sensible Sphinx directive for stripping unnecessary module path!

[stinodego]

Cunning workaround :) Was surprised to discover that there is no sensible Sphinx directive for stripping unnecessary module path!

There isn't? That's disappointing. There has to be a better way!

For now I think this line wrapping workaround is definitely an improvement.

[alexander-beedie]

There isn't? That's disappointing. There has to be a better way!
For now I think this line wrapping workaround is definitely an improvement.

If there is I couldn't readily find it (I also notice that pandas, which uses the same Sphinx theme, suffers from the same problem). There was an outside chance that add_module_names = False in conf.py might help, but sadly not.

A slightly more granular restructuring of the top-level reference toctrees might help though. I'll experiment later.