feat(vortex-geo): add area scalar function

[HarukiMoriarty]

Summary

• add a native GeoArea scalar function backed by geo::Area::unsigned_area
• support every native geometry dtype, including zero-area point and line families and rectangular boxes
• preserve input nullability and propagate null rows through unary columnar execution
• give every geo scalar function a local operand validator that checks exact arity before dtype validation
• parameterize repeated Area, Envelope, and MakeLine cases with rstest for clearer per-dtype failures and broader operand-order coverage

Semantics

This follows DuckDB ST_Area semantics over the native Vortex geometry types: point and line families return zero, polygon holes are subtracted, multipolygon component areas are summed, rectangles use width times height, and higher coordinate dimensions are ignored. WKB remains excluded because it is a storage representation rather than a native decoded geometry.

[codspeed-hq]

Merging this PR will improve performance by 43.62%

⚡ 10 improved benchmarks
✅ 1633 untouched benchmarks
🆕 32 new benchmarks
⏩ 307 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	take_chunked_fsl_sorted[32, 64]	1,290.7 µs	559.8 µs	×2.3
⚡	Simulation	take_chunked_fsl_sorted[16, 64]	916.4 µs	533.5 µs	+71.8%
⚡	Simulation	take_fsl_chunked_elements_sorted[32, 64]	1,075.4 µs	702.7 µs	+53.04%
⚡	Simulation	take_fsl_chunked_elements_random[32, 64]	1,084.6 µs	742.1 µs	+46.15%
⚡	Simulation	take_chunked_fsl_sorted[8, 64]	722.5 µs	507.4 µs	+42.4%
⚡	Simulation	take_fsl_chunked_elements_sorted[16, 64]	662.4 µs	502.4 µs	+31.85%
⚡	Simulation	take_fsl_chunked_elements_random[16, 64]	680.3 µs	520 µs	+30.83%
⚡	Simulation	take_fsl_chunked_elements_random[8, 64]	492.1 µs	407.6 µs	+20.73%
⚡	Simulation	take_fsl_chunked_elements_sorted[8, 64]	464.6 µs	385.3 µs	+20.58%
⚡	Simulation	take_chunked_fsl_random[32, 64]	1.3 ms	1.1 ms	+17.81%
🆕	Simulation	take[core/shuffled/primitive/nonnull/chunks=1024/indices=10000]	N/A	16.3 ms	N/A
🆕	Simulation	take[core/shuffled/primitive/nonnull/chunks=1024/indices=100000]	N/A	48.9 ms	N/A
🆕	Simulation	take[core/shuffled/primitive/nonnull/chunks=16/indices=10000]	N/A	4 ms	N/A
🆕	Simulation	take[core/shuffled/primitive/nonnull/chunks=16/indices=100000]	N/A	38.2 ms	N/A
🆕	Simulation	take[core/sorted/primitive/nonnull/chunks=1024/indices=10000]	N/A	14.3 ms	N/A
🆕	Simulation	take[core/sorted/primitive/nonnull/chunks=1024/indices=100000]	N/A	31 ms	N/A
🆕	Simulation	take[core/sorted/primitive/nonnull/chunks=16/indices=10000]	N/A	2.7 ms	N/A
🆕	Simulation	take[core/sorted/primitive/nonnull/chunks=16/indices=100000]	N/A	29.5 ms	N/A
🆕	Simulation	take[duplicates/duplicate90/primitive/nonnull/chunks=16/indices=100000]	N/A	34.4 ms	N/A
🆕	Simulation	take[duplicates/duplicate99/primitive/nonnull/chunks=16/indices=100000]	N/A	33.6 ms	N/A
...	...	...	...	...	...

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

Tip

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

---

_{Comparing nemo/geo-area (40f4666) with nemo/geo-native-bbox-precheck (4abcde8)²}

Footnotes

1. 307 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 nemo/geo-native-bbox-precheck (98e2031) during the generation of this report, so 157838c was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩