Add Compaction encoding for normalizing arrays to compact form

[joseph-isaacs]

This is more of an idea for now

Summary

This PR introduces the Compaction encoding, a transient array wrapper that normalizes its child into a compact canonical form when executed. This is similar to the existing Slice encoding in that it's non-serializable and exists primarily to drive execution.

The Compaction encoding handles structural normalization of various array types:

• ListView arrays are rebuilt to be zero-copy convertible to Arrow-style ListArray (overlapping views deduplicated, leading/trailing garbage trimmed)
• VarBinView arrays have their data buffers garbage collected
• Dict arrays are either decoded to flat canonical form or garbage collected in place (dead values removed, codes remapped) based on a cost heuristic
• Struct fields are recursively compacted

[codspeed-hq]

Merging this PR will not alter performance

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

⚡ 7 improved benchmarks
❌ 4 regressed benchmarks
✅ 1496 untouched benchmarks

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)]	162 µs	198.2 µs	-18.26%
❌	Simulation	compare[15]	120.4 µs	146.5 µs	-17.8%
❌	Simulation	compare[14]	118 µs	142.2 µs	-17%
❌	Simulation	compare[13]	116.1 µs	138.5 µs	-16.14%
⚡	Simulation	varbinview_zip_block_mask	3.7 ms	2.9 ms	+27.51%
⚡	Simulation	bitwise_not_vortex_buffer_mut[128]	275.3 ns	216.9 ns	+26.89%
⚡	Simulation	bitwise_not_vortex_buffer_mut[1024]	336.9 ns	278.6 ns	+20.94%
⚡	Simulation	bitwise_not_vortex_buffer_mut[2048]	400.6 ns	342.2 ns	+17.05%
⚡	Simulation	varbinview_zip_fragmented_mask	6.9 ms	6.1 ms	+13.07%
⚡	Simulation	chunked_varbinview_canonical_into[(100, 100)]	309.6 µs	274.3 µs	+12.88%
⚡	Simulation	compare[5]	77.5 µs	70.1 µs	+10.5%

Tip

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

---

_{Comparing claude/quirky-carson-1NaxF (1b3a817) with develop (d97d2bd)}