[DO NOT MERGE] Add list length min-max aggregate

[mhk197]

Adds an aggregate function ListLengthMinMax to calculate the minimum and maximum length of a List-like array. This is just the current MinMax aggregate composed with the ListLength scalar function.

This will allow us to do several beneficial things if incorporated as a default stat in upcoming pr)

• Easily detect if variable sized lists have a single shared size and write with upcoming FixedSizeList layout
• Include as zone map stat to allow for pruning list predicates like list_length (obviously) and others, since we can identify zones with all-empty lists and prune them (e.g. list_sum > k)

What changes are included in this PR?

• Add ListLengthMinMax for List and FixedSizeList inputs.
• Return nullable min_length and max_length fields, ignoring null outer lists while allowing valid empty lists to contribute zero.
• Reuse the numeric min_max implementation over the List sizes child with outer-list validity attached.
• Short-circuit FixedSizeList inputs from their dtype and outer validity, saturating after the first non-null list.
• Register the aggregate in the default aggregate-function session.

NOTE: Should come up with framework to do aggregate on derived column instead of just implementing this one-off

[codspeed-hq]

Merging this PR will degrade performance by 19.98%

⚡ 1 improved benchmark
❌ 2 regressed benchmarks
✅ 1844 untouched benchmarks
🆕 2 new benchmarks
⏩ 46 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	copy_nullable[65536]	578.8 µs	1,033.4 µs	-43.99%
❌	Simulation	copy_non_nullable[65536]	799 µs	1,112 µs	-28.15%
⚡	Simulation	decode_varbin[(1000, 4)]	34.3 µs	26.9 µs	+27.34%
🆕	Simulation	add_decimal_i128_nullable	N/A	3.9 ms	N/A
🆕	Simulation	add_decimal_i64_nonnull	N/A	1.9 ms	N/A

Tip

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

---

_{Comparing mk/list-shape-stats (614492d) with develop (3a098b4)²}

Footnotes

1. 46 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. ↩

2. No successful run was found on develop (46b0b2b) during the generation of this report, so 3a098b4 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[github-actions]

Polar Signals Profiling Results

Latest Run

Status	Commit	Job	Attempt	Link
🟢 Done	45e424f		1	Explore Profiling Data

Previous Runs (2)

Status	Commit	Job	Attempt	Link
🟢 Done	4c3a191		1	Explore Profiling Data
🟢 Done	dbba71d		1	Explore Profiling Data

---

Powered by Polar Signals Cloud

[github-actions]

Benchmarks: Vortex queries 📖

Verdict: Likely regression (medium confidence)
Attributed Vortex impact: +11.9%
Engines: DataFusion Likely regression (+21.4%, environment too noisy confidence) · DuckDB No clear signal (+3.2%, low confidence)
Vortex (geomean): 1.094x ➖
Parquet (geomean): 1.035x ➖
Shifts: Parquet (control) +3.5% · Median polish +5.1%

How to read Verdict and Engines

• Verdict: Overall PR-level signal after subtracting baseline drift estimated from Parquet control rows. It can be Likely improvement, Likely regression, or No clear signal.
• Engines: Per-engine attribution. DataFusion is compared against DataFusion/Parquet controls; DuckDB is compared against DuckDB/Parquet controls. This answers whether each engine improved or regressed independently.
• Confidence: Based on directional consistency, share of rows above the noise floor, and control-run noise.

---

datafusion / vortex-file-compressed (1.276x ❌, 0↑ 2↓)

name	PR 45e424f (ns)	base 57962f4 (ns)	ratio (PR/base)
vortex_q00/datafusion:vortex-file-compressed 🚨	9906785	8197567	1.21
vortex_q01/datafusion:vortex-file-compressed 🚨	6582691	4885078	1.35

datafusion / parquet (1.051x ➖, 0↑ 0↓)

name	PR 45e424f (ns)	base 57962f4 (ns)	ratio (PR/base)
vortex_q00/datafusion:parquet	20931221	20638806	1.01
vortex_q01/datafusion:parquet	4990970	4580085	1.09

duckdb / vortex-file-compressed (1.051x ➖, 0↑ 0↓)

name	PR 45e424f (ns)	base 57962f4 (ns)	ratio (PR/base)
vortex_q00/duckdb:vortex-file-compressed	10561746	10061599	1.05
vortex_q01/duckdb:vortex-file-compressed	6483576	6167184	1.05

duckdb / parquet (1.018x ➖, 0↑ 0↓)

name	PR 45e424f (ns)	base 57962f4 (ns)	ratio (PR/base)
vortex_q00/duckdb:parquet	23699642	23195009	1.02
vortex_q01/duckdb:parquet	9476549	9340266	1.01

---

No file size changes detected.