chore: forbid the locking Id constructors with a clippy lint

[miniex]

Summary

Follow-up to #8614. A clippy disallowed-methods lint now forbids Id::new and Id::new_static workspace-wide, since both intern into a globally locked table (ThreadedRodeo) on every call. Static identifiers must use a CachedId static instead; the sites that intern a genuinely dynamic string (proto/flatbuffer deserialization, arrow extension lookups, Python read contexts, tests) keep Id::new behind an #[expect(clippy::disallowed_methods, reason = "...")]. CachedId itself now interns through Id::new_static, so it is the single sanctioned path for static ids.

This is the enforcement half of #8380; #8614 did the migration. It is stacked on #8614, so it stays a draft until that merges; the shared migration commit drops out once this is rebased onto a merged #8614.

Refs: #8380

Testing

cargo clippy --all-targets is clean across the edited crates (vortex-array, vortex-layout, vortex-file, vortex-ipc, vortex-zstd, vortex-python, vortex-duckdb, vortex-session), with disallowed_methods and unfulfilled_lint_expectations both at zero; cargo +nightly fmt --all --check clean. vortex-cuda could not be built locally (no CUDA toolchain).

---

I'm Korean, so sorry if any wording reads a little awkward.

[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.

⚡ 5 improved benchmarks
❌ 2 regressed benchmarks
✅ 1588 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_varbinview_into_canonical[(1000, 10)]	168.9 µs	205.7 µs	-17.89%
❌	Simulation	slice_empty_vortex	339.4 ns	397.8 ns	-14.66%
⚡	Simulation	chunked_bool_canonical_into[(1000, 10)]	26.9 µs	15.7 µs	+71.06%
⚡	Simulation	chunked_varbinview_canonical_into[(100, 100)]	259.5 µs	224.3 µs	+15.69%
⚡	Simulation	chunked_varbinview_into_canonical[(100, 100)]	306.8 µs	271.9 µs	+12.82%
⚡	Simulation	bitwise_not_vortex_buffer_mut[128]	273.6 ns	244.4 ns	+11.93%
⚡	Simulation	eq_i64_constant	318.9 µs	288.2 µs	+10.68%

Tip

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

---

_{Comparing miniex:perf/disallow-id-new-lint (f58aa0b) with develop (733ab9e)}

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. ↩