execute_until returns Match rather than ArrayRef

[mhk197]

Reworks ArrayRef::execute_until so it returns the matcher's view directly instead of discarding the match result and handing back an ArrayRef that callers must re-downcast.

pub fn execute_until<'a, M: Matcher>(
    &'a mut self,
    ctx: &mut ExecutionCtx,
) -> VortexResult<M::Match<'a>>

• Takes &mut self: it mem::replaces self with a cheap alloc-free Array::<Null>::new(0) placeholder, runs the existing loop (now a private execute_until_done(self, ctx, done)), writes the executed array back into *self, and returns M::try_match(self). The placeholder keeps the loop's in-place Arc::get_mut optimizations intact.
• Behavioral change: errors if execution converges to a canonical form that M does not match, rather than silently returning that canonical array.
• Adds OrCanonical<M> (+ OrCanonicalMatch { Matched(M::Match) | Canonical(CanonicalView) }) to restore the old non-failing behavior with a typed fallback, for callers that want it.

[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.

⚠️ 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

⚡ 11 improved benchmarks
❌ 12 regressed benchmarks
✅ 1566 untouched benchmarks
⏩ 4 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	chunked_bool_canonical_into[(1000, 10)]	16 µs	26.4 µs	-39.45%
❌	Simulation	chunked_varbinview_into_canonical[(1000, 10)]	169 µs	206.2 µs	-18.05%
❌	Simulation	chunked_opt_bool_canonical_into[(100, 100)]	140.2 µs	163.9 µs	-14.44%
❌	Simulation	decode_primitives[f32, (1000, 2)]	17.2 µs	20 µs	-14.32%
❌	Simulation	decode_primitives[f32, (1000, 4)]	17.2 µs	20 µs	-13.93%
❌	Simulation	decode_primitives[f32, (1000, 8)]	17.2 µs	19.6 µs	-12.29%
❌	Simulation	decode_primitives[f32, (1000, 32)]	17.2 µs	19.6 µs	-12.25%
❌	Simulation	chunked_varbinview_opt_canonical_into[(10, 1000)]	1.7 ms	1.9 ms	-11.59%
❌	Simulation	chunked_opt_bool_into_canonical[(100, 100)]	188.1 µs	211.3 µs	-10.99%
❌	Simulation	decode_primitives[i64, (1000, 32)]	20.3 µs	22.5 µs	-10.15%
❌	Simulation	chunked_opt_bool_canonical_into[(1000, 10)]	29.7 µs	33 µs	-10.15%
❌	Simulation	eq_i64_constant	288.6 µs	321 µs	-10.08%
⚡	Simulation	varbinview_large	392.9 µs	134 µs	×2.9
⚡	Simulation	copy_nullable[65536]	1.4 ms	1 ms	+31.88%
⚡	Simulation	bitwise_not_vortex_buffer_mut[128]	244.4 ns	186.1 ns	+31.34%
⚡	Simulation	bitwise_not_vortex_buffer_mut[1024]	304.7 ns	246.4 ns	+23.68%
⚡	Simulation	copy_non_nullable[65536]	1,089.1 µs	909.7 µs	+19.72%
⚡	Simulation	bitwise_not_vortex_buffer_mut[2048]	398.6 ns	340.3 ns	+17.14%
⚡	Simulation	chunked_varbinview_into_canonical[(100, 100)]	306.5 µs	272.5 µs	+12.45%
⚡	Simulation	encode_varbin[(1000, 32)]	165 µs	146.9 µs	+12.34%
...	...	...	...	...	...

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

Tip

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

---

_{Comparing mk/execute-until-match (bc13cc5) with develop (9567467)}

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. ↩