Port PiecewiseSequence run-take consumers

[danking]

Summary

• add Extension TakeReduce so take can push through extension storage during reduce
• fast-path List.take(PiecewiseSequence) by rebuilding list offsets and taking contiguous child element ranges when multipliers is a constant unsigned 1
• restore the large-list ListView rebuild path using PiecewiseSequence child indices with multiplier 1 instead of list-by-list rebuild

Validation

• cargo fmt -p vortex-array
• cargo test -p vortex-array piecewise_sequence
• cargo test -p vortex-array piecewise_sequence_take
• cargo test -p vortex-array test_rebuild_flatten_large_lists_with_piecewise_indices
• cargo check -p vortex-array
• cargo check -p vortex-array --benches
• git diff --check

[codspeed-hq]

Merging this PR will regress 2 benchmarks

⚡ 4 improved benchmarks
❌ 2 regressed benchmarks
✅ 1792 untouched benchmarks
⏩ 52 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	copy_nullable[65536]	1 ms	1.4 ms	-24.24%
❌	Simulation	copy_non_nullable[65536]	908.9 µs	1,089.5 µs	-16.57%
⚡	Simulation	struct_large	431.9 µs	39.1 µs	×11
⚡	Simulation	fsl_large	615.7 µs	89.2 µs	×6.9
⚡	Simulation	list_i32_large	172.8 µs	107.8 µs	+60.32%
⚡	Simulation	i32_large	277.8 µs	238 µs	+16.71%

Tip

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

---

_{Comparing codex/piecewise-sequential-run-take-consumers (490e9db) with codex/piecewise-sequential-take (c94c4ef)}

Footnotes

1. 52 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. ↩