Index optimized for small data

[yihozhang]

4/5 of the indices built for some large benchmarks only have one element. It is expensive to build a ColumnIndex for them.

This PR is currently bottlenecked on scanning and materializing the rows corresponding to the Subset.

[codecov-commenter]

Codecov Report

❌ Patch coverage is 94.54545% with 12 lines in your changes missing coverage. Please review.
✅ Project coverage is 86.94%. Comparing base (0a5967e) to head (948e30c).
⚠️ Report is 2 commits behind head on yihozhang-decomp.

Files with missing lines	Patch %	Lines
core-relations/src/row_buffer/mod.rs	90.59%	11 Missing ⚠️
core-relations/src/free_join/execute.rs	99.00%	1 Missing ⚠️

Additional details and impacted files

@@                 Coverage Diff                  @@
##           yihozhang-decomp     #840      +/-   ##
====================================================
+ Coverage             86.90%   86.94%   +0.04%     
====================================================
  Files                    85       85              
  Lines                 24326    24475     +149     
====================================================
+ Hits                  21140    21280     +140     
- Misses                 3186     3195       +9     

☔ 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.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[codspeed-hq]

Merging this PR will improve performance by 24.19%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 9 improved benchmarks
✅ 35 untouched benchmarks
⏩ 190 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	WallTime	tests[eggcc-extraction]	1.8 s	1.7 s	+6.04%
⚡	WallTime	tests[hardboiled_conv1d_32]	1.7 s	1.4 s	+16.98%
⚡	WallTime	tests[hardboiled_conv1d_128]	6.7 s	5.5 s	+21.38%
⚡	WallTime	tests[herbie]	272.5 ms	254.9 ms	+6.89%
⚡	Simulation	tests[math-microbenchmark]	8.9 s	8.5 s	+5.1%
⚡	WallTime	tests[proof_testing_typecheck]	311.2 ms	293.6 ms	+5.99%
⚡	WallTime	tests[python_array_optimize]	6.9 s	6.5 s	+5.67%
⚡	Simulation	tests[hardboiled_conv1d_32]	1.8 s	1.7 s	+8.02%
⚡	Simulation	tests[hardboiled_conv1d_128]	8.9 s	7.2 s	+24.19%

---

_{Comparing yihozhang-decomp-small-index (948e30c) with yihozhang-decomp (22d06c1)}

Footnotes

1. 190 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩