fix[vortex-array]: don't use input list field type in preferred conversion

[asubiotto]

This would cause a panic in arrow when attempting to create a list when the preferred conversion resulted in a different element array type due to the different types.

This commit also simplifies the input arguments so that preferred type conversion never has access to the canonical arrow type.

Ran into this while testing #4464

[codspeed-hq]

CodSpeed Performance Report

Merging #4481 will degrade performances by 12.57%

_{Comparing asubiotto/listpref (77f5590) with develop (bdb7cac)¹}

Summary

⚡ 22 improvements
❌ 1 regressions
✅ 1328 untouched benchmarks

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Benchmarks breakdown

	Benchmark	BASE	HEAD	Change
⚡	decode_primitives[f32, (1000, 2)]	21.2 µs	19.3 µs	+10.03%
⚡	decode_primitives[f32, (1000, 4)]	21 µs	19.1 µs	+10.11%
⚡	decode_primitives[u8, (1000, 128)]	21.5 µs	19.5 µs	+10.39%
⚡	decode_primitives[u8, (1000, 32)]	21.4 µs	19.4 µs	+10.44%
⚡	decode_primitives[u8, (1000, 4)]	21.4 µs	19.4 µs	+10.44%
⚡	decode_primitives[u8, (1000, 512)]	21.5 µs	19.5 µs	+10.38%
⚡	decode_primitives[u8, (1000, 8)]	21.4 µs	19.4 µs	+10.44%
❌	new_bp_prim_test_between[i16, 2048]	36.6 µs	41.8 µs	-12.57%
⚡	new_raw_prim_test_between[i32, 2048]	31.4 µs	25.3 µs	+23.9%
⚡	take_indices[(1000, 16, 0.005)]	19.2 µs	17 µs	+13.31%
⚡	take_indices[(1000, 16, 0.01)]	18.9 µs	16.6 µs	+13.58%
⚡	take_indices[(1000, 16, 0.03)]	19.8 µs	17.5 µs	+12.88%
⚡	take_indices[(1000, 256, 0.005)]	18.5 µs	16.3 µs	+13.9%
⚡	take_indices[(1000, 256, 0.01)]	18.6 µs	16.3 µs	+13.83%
⚡	take_indices[(1000, 256, 0.03)]	20 µs	17.7 µs	+12.75%
⚡	take_indices[(1000, 4, 0.01)]	20.9 µs	18.6 µs	+12.13%
⚡	take_indices[(1000, 4, 0.03)]	22 µs	19.7 µs	+11.47%
⚡	take_indices[(10000, 16, 0.005)]	21.5 µs	19.2 µs	+11.77%
⚡	take_indices[(10000, 16, 0.01)]	24.3 µs	22 µs	+10.26%
⚡	take_indices[(10000, 256, 0.005)]	20.6 µs	18.3 µs	+12.32%
...	...	...	...	...

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

Footnotes

1. No successful run was found on develop (473e486) during the generation of this report, so bdb7cac was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩

[codecov]

Codecov Report

❌ Patch coverage is 95.45455% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 88.93%. Comparing base (fb77ecd) to head (77f5590).
⚠️ Report is 8 commits behind head on develop.

Files with missing lines	Patch %	Lines
...rtex-array/src/arrow/compute/to_arrow/canonical.rs	95.45%	1 Missing ⚠️

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.