feat(vortex-geo): add make-line and length scalar functions

[HarukiMoriarty]

Stack

This PR is stacked on #9215, which introduces the shared unary and binary geometry scalar execution layer. Review that PR first; this diff contains only the ST_MakeLine / ST_Length feature work and its InterleaveArray dependency.

Rationale for this change

Vortex's native geometry extension types should support constructing two-point line strings and measuring line-string length without an Arrow or WKB round trip.

What changes are included in this PR?

• Add generic primitive InterleaveArray execution, used to assemble MakeLine coordinate buffers from paired point columns.
• Add vortex.geo.make_line for pairs of native Point arrays, paired row-by-row.
• Promote mixed XY, XYZ, XYM, and XYZM inputs and fill missing ordinates with zero.
• Preserve constant execution when both MakeLine operands are constant.
• Add vortex.geo.length with planar Euclidean semantics, restricted to native LineString inputs.
• Build MakeLine results directly as native columnar LineString storage.
• Add coverage for primitive interleave execution, operand ordering, constants, null propagation, mixed dimensions, CRS handling, and invalid operand types.

What APIs are changed? Are there any user-facing changes?

The Vortex geometry session now registers and exports GeoMakeLine and GeoLength.

[codspeed-hq]

Merging this PR will improve performance by 19.96%

⚠️ 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

⚡ 2 improved benchmarks
✅ 1893 untouched benchmarks
🆕 6 new benchmarks
⏩ 43 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	decode_varbin[(1000, 2)]	78.6 µs	62 µs	+26.61%
⚡	Simulation	compress_fsst[(1000, 64, 8)]	1.2 ms	1 ms	+13.66%
🆕	Simulation	nullable_two_vertex_lines	N/A	1 ms	N/A
🆕	Simulation	sixteen_vertex_lines	N/A	1.9 ms	N/A
🆕	Simulation	two_vertex_lines	N/A	668.3 µs	N/A
🆕	Simulation	column_x_column	N/A	272 µs	N/A
🆕	Simulation	column_x_constant	N/A	647.3 µs	N/A
🆕	Simulation	nullable_columns	N/A	292.5 µs	N/A

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-native-bbox-precheck (72a4ae1) with nemo/geo-scalar-execution (09f16ff)}

Footnotes

1. 43 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. ↩

[connortsui20]

@claude review this PR, and look for potential regressions as well as inconsistencies with other code that is not in this PR

[github-actions]

Claude encountered an error after 2s —— View job

---

I'll analyze this and get back to you.