Replace ToCanonical trait usages with execute

[robert3005]

This also removes ArrayBuilder::extend_from_array and moves that logic
in to ArrayRef::append_to_builder

Plumb ExecutionCtx through more methods instead of creating fresh one.

Fixes: #3235

Signed-off-by: Robert Kruszewski github@robertk.io

[codspeed-hq]

Merging this PR will not alter performance

⚠️ 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.

⚡ 7 improved benchmarks
❌ 10 regressed benchmarks
✅ 1578 untouched benchmarks
⏩ 4 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	extend_from_array_zctl[(1000, 8)]	238.9 µs	404.6 µs	-40.96%
❌	Simulation	extend_from_array_non_zctl_overlapping[(1000, 8)]	271.3 µs	436.3 µs	-37.81%
❌	Simulation	extend_from_array_non_zctl_overlapping[(1000, 32)]	735.7 µs	900.5 µs	-18.31%
❌	Simulation	chunked_varbinview_into_canonical[(1000, 10)]	169.1 µs	205.6 µs	-17.73%
❌	Simulation	slice_empty_vortex	339.4 ns	397.8 ns	-14.66%
❌	Simulation	extend_from_array_zctl[(1000, 64)]	1 ms	1.2 ms	-13.91%
❌	Simulation	i32_small_overlapping	39.9 µs	44.8 µs	-11.08%
❌	Simulation	extend_from_array_zctl[(10000, 8)]	1.9 ms	2.2 ms	-10.77%
❌	Simulation	extend_from_array_non_zctl_overlapping[(10000, 8)]	2.2 ms	2.5 ms	-10.62%
❌	Simulation	compact_sliced[(4096, 90)]	750 ns	838.1 ns	-10.51%
⚡	Simulation	chunked_bool_canonical_into[(1000, 10)]	26.3 µs	16 µs	+64.78%
⚡	Simulation	bench_many_codes_few_values[1024]	529.3 µs	364.1 µs	+45.4%
⚡	Simulation	chunked_varbinview_canonical_into[(100, 100)]	259.7 µs	223.6 µs	+16.14%
⚡	Simulation	bitwise_not_vortex_buffer_mut[128]	244.4 ns	215.3 ns	+13.55%
⚡	Simulation	chunked_varbinview_into_canonical[(100, 100)]	306.7 µs	270.8 µs	+13.27%
⚡	Simulation	rebuild_naive	109 µs	98.5 µs	+10.69%
⚡	Simulation	bitwise_not_vortex_buffer_mut[1024]	304.7 ns	275.6 ns	+10.58%

Tip

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

---

_{Comparing claude/nice-hypatia-g68zgt (947a966) with develop (bf2be52)}

Footnotes

1. 4 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. ↩

[lwwmanning]

builder_with_capacity(DType::List, ...) returns a ListViewBuilder<u64, u64>, so generic callers that allocate the canonical builder and then append a ListArray now hit this branch and error out instead of appending. For example, zip/case_when build the default list builder and append sliced branches through append_to_builder. This should either append into the canonical ListViewBuilder or fall back through execute::<ListViewArray> before appending.

[robert3005]

This part needs reworking. Needs to do a different downcast here

[lwwmanning]

len can now be as high as 2048, but random_strictly_sorted_ends still randomly chooses PType::U8 and later does u8::try_from(e).vortex_expect(...) after forcing the last end toward this target. Ordinary fuzzer input can therefore panic during Arbitrary instead of returning arbitrary::Error. Please choose an end type that can represent the generated max end, cap the target by the chosen type, or return an error instead of vortex_expect.