perf: Make is_in row-group pruning precise on null-containing haystacks

[azimafroozeh]

Follow-up to #27475. The helper try_extract_is_in_haystack used to give up on haystacks with nulls under default nulls_equal=false. Now it drops them (like pyarrow does); under nulls_equal=true, if a null was dropped, we add null_count(col) == 0 to the skip predicate so row groups with nulls aren't skipped.

Also fixes a hidden bug: col_has_no_nulls = col_nc.has_no_nulls(arena) computes a column-wide value that's always true against the stats table, so the check did nothing. Replaced with col_nc.eq(idx_zero, arena) so it checks per row group.

[nameexhaustion]

Looks good, just 1 nit https://github.com/pola-rs/polars/pull/27495/changes#r3180804874

[codecov]

Codecov Report

❌ Patch coverage is 87.50000% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 81.14%. Comparing base (62f38ef) to head (5a9c1d4).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
...rs-plan/src/plans/aexpr/predicates/skip_batches.rs	80.95%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #27495      +/-   ##
==========================================
+ Coverage   80.54%   81.14%   +0.60%     
==========================================
  Files        1842     1842              
  Lines      254718   254720       +2     
  Branches     3181     3181              
==========================================
+ Hits       205173   206705    +1532     
+ Misses      48722    47192    -1530     
  Partials      823      823              

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

🚀 New features to boost your workflow:

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

.