Implement compare in vortex instead of falling back to arrow

[robert3005]

Instead of going through arrow compare logic implement it in vortex

Lets us leverage optimised bitbuffer optimistaions and ultimately remove arrow
dependency from vortex-array

[codspeed-hq]

Merging this PR will improve performance by 32.73%

⚡ 79 improved benchmarks
❌ 2 regressed benchmarks
✅ 1526 untouched benchmarks
🆕 10 new benchmarks
⏩ 42 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_varbinview_canonical_into[(1000, 10)]	154.7 µs	191 µs	-19.02%
❌	Simulation	slice_empty_vortex	310.3 ns	368.6 ns	-15.83%
⚡	Simulation	compare_bool	458 µs	40.2 µs	×11
⚡	Simulation	case_when_nary_10_conditions[1000]	241.8 µs	148.4 µs	+62.91%
⚡	Simulation	baseline_lt[4, 1024]	66.9 µs	42.3 µs	+58.25%
⚡	Simulation	baseline_eq[4, 1024]	65.3 µs	42.1 µs	+55.06%
⚡	Simulation	baseline_eq[16, 1024]	65.6 µs	42.4 µs	+54.47%
⚡	Simulation	baseline_lt[16, 1024]	66.1 µs	42.9 µs	+54.13%
⚡	Simulation	bench_compare_sliced_dict_varbinview[(1000, 10000)]	110.6 µs	71.9 µs	+53.76%
⚡	Simulation	bench_compare_sliced_dict_primitive[(1000, 10000)]	80.8 µs	53.5 µs	+51.09%
⚡	Simulation	bench_compare_sliced_dict_varbinview[(2000, 10000)]	136.7 µs	93.5 µs	+46.27%
⚡	Simulation	bench_compare_sliced_dict_varbinview[(2500, 10000)]	150.6 µs	104.3 µs	+44.39%
⚡	Simulation	case_when_nary_equality_lookup[1000]	174 µs	120.9 µs	+43.87%
⚡	Simulation	bench_compare_sliced_dict_primitive[(2000, 10000)]	84.8 µs	59.1 µs	+43.66%
⚡	Simulation	case_when_all_false[1000]	75.3 µs	52.7 µs	+42.93%
⚡	Simulation	bench_compare_sliced_dict_primitive[(2500, 10000)]	87.8 µs	61.6 µs	+42.55%
⚡	Simulation	bench_compare_sliced_dict_varbinview[(3333, 10000)]	173.8 µs	122.6 µs	+41.7%
⚡	Simulation	canonicalize_compare[(1000, 4, 4)]	116.7 µs	82.5 µs	+41.58%
⚡	Simulation	canonicalize_compare[(1000, 4, 8)]	118.7 µs	84.1 µs	+41.23%
⚡	Simulation	case_when_all_true[1000]	76.6 µs	54.5 µs	+40.38%
...	...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing rk/vortexcompare (e54a66b) with develop (a6e3d3e)}

Footnotes

1. 42 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. ↩