Compute writer chunk hints through the aggregate cache

[connortsui20]

Superseded by #10269, which groups this change with the related statistics work. The original commits and branch are preserved.

Original PR description

Tracking Issue: #10177

Summary

The file writer needs both a cached answer for each chunk and the partial state needed to combine chunks. A finalized answer cannot always supply that state: two individually sorted chunks can be out of order at their shared boundary. Publishing through a bound computation keeps both results tied to the input that produced them.

Changes

Reuse the writer's chunk accumulators through the aggregate cache API. Preserve fused extrema, cached Min/Max recovery, overflow state, and the existing omission of null chunk hints. For review, start with the bound computation contract, then its core accumulator dispatch, followed by the writer callers.

API Changes

Add AggregationsRef::compute_into. It resets and computes a matching core accumulator, caches an exact non-null final, and leaves the real partial state available for merging. compute_result continues to cache exact nulls.

Stack

Depends on #10265 in migration stack #10229.

[codspeed]

Merging this PR will improve performance by 36.17%

⚠️ 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.

⚠️ 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

⚡ 3 improved benchmarks
✅ 2100 untouched benchmarks
⏩ 503 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	bench_compare_sliced_dict_primitive[(2500, 10000)]	83 µs	42.8 µs	+94.17%
⚡	Simulation	chunked_opt_bool_canonical_into[(1000, 10)]	69.9 µs	59.3 µs	+17.81%
⚡	Simulation	compress_fsst[(500, 4, 8)]	216.9 µs	196.5 µs	+10.37%

Tip

Curious why performance improved? Comment @codspeedbot explain why performance improved on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing ct/stats-writer-cache-computation (aef451a) with ct/stats-remaining-consumers (019c756)}

Footnotes

1. 503 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. ↩