feat(array): add probe_scalar to OperationsVTable with a ProbeState shell

[joseph-isaacs]

The mechanical half of the scalar-probe API, split out so the probe work itself (stacked on this) can be reviewed on its own.

What

• OperationsVTable gains type ProbeState: Default + 'static, the state an encoding may keep across repeated reads, and probe_scalar(state: &mut ProbeState<'_, V>, index, ctx), defaulting to scalar_at. No encoding changes behaviour.
• ProbeState<'_, V> for now only carries the ArrayView. The erased dispatch routes one-off reads through it.
• Every encoding declares type ProbeState = () (44 files, one line each).
• scalar_at stays until encodings migrate.

Checks

• cargo +nightly fmt --all -- --check: clean
• cargo clippy -p vortex-array --all-targets --all-features: clean
• cargo nextest run -p vortex-array --cargo-profile bench --no-fail-fast: passes apart from the two listview should_panic tests that cannot pass in the bench profile (pre-existing)
• Doctests for vortex-array: pass
• cargo build --workspace (excluding CUDA crates): passes

🤖 Generated with Claude Code

[joseph-isaacs]

Folded into #9890 as its first commit.

[codspeed]

Merging this PR will degrade performance by 0.16%

⚠️ 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

⚡ 3 improved benchmarks
❌ 2 regressed benchmarks
✅ 2193 untouched benchmarks
⏩ 218 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	random_i8[0.5]	71.3 µs	94.7 µs	-24.68%
❌	Simulation	decompress[u64, (4000, 1024)]	71.5 µs	86.9 µs	-17.71%
⚡	Simulation	random_i16[0.8]	96.4 µs	77.8 µs	+23.93%
⚡	WallTime	words_gather_scalar_avx2[65536]	9.4 µs	8.2 µs	+13.82%
⚡	Simulation	allocate_drop_arrow[0]	456.9 ns	402.7 ns	+13.45%

Tip

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

---

_{Comparing ji/probe-vtable-hook (264e0d4) with develop (b5f43ba)}

Footnotes

1. 218 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. ↩