perf: Optimize strings slices

[ritchie46]

Didn't do every length combinations, but the most occurring are handled.

This doesn't do a memcopy of the string anymore, but updates the views. Depending on the size of the slice this has increasing performance improvements:

length,    improvement
4               1.17x
19             1.42

And this will scale linearly.

[codecov]

Codecov Report

Attention: Patch coverage is 96.79487% with 5 lines in your changes missing coverage. Please review.

Project coverage is 80.51%. Comparing base (03cbfd5) to head (8810811).
Report is 2 commits behind head on main.

Files	Patch %	Lines
.../polars-ops/src/chunked_array/strings/substring.rs	96.12%	5 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17996      +/-   ##
==========================================
+ Coverage   80.49%   80.51%   +0.02%     
==========================================
  Files        1496     1496              
  Lines      196639   196728      +89     
  Branches     2817     2817              
==========================================
+ Hits       158278   158395     +117     
+ Misses      37840    37812      -28     
  Partials      521      521              

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