Add optional Rust accelerator for bidict

[jab]

Summary

• add an optional Rust/PyO3 native helper for bulk bidict build and update paths
• extend the native path to cover mapping-specialized bulk operations, forceupdate routing, and a no-dup fast path in the Rust ingestion loop
• add Rust tooling and hooks via nix, plus native-aware test and benchmark CI coverage
• improve the benchmark PR comment so it reports notable per-benchmark deltas from CI

Benchmark notes

• local and Linux A/B measurements show clear wins on large bulk initialization and update workloads
• benchmark CI is now wired to post notable improvements and regressions back to the PR comment

Follow-up before ready for review

• update CHANGELOG.rst with release-note coverage for the optional native helper and the related CI/tooling additions

[github-actions]

Benchmark report

⚠️ Informational benchmark regression detected.

• Baseline asset: GHA-linux-cachegrind-x86_64-CPython-3.14-baseline.json
• Threshold: min:5%
• Result: one or more benchmarks slowed down by at least min:5%; workflow kept green because PR benchmarking is informational
• Run: benchmark workflow run

This benchmark check is informational and does not block merging.

Deltas below compare benchmark min time against GHA-linux-cachegrind-x86_64-CPython-3.14-baseline.json for benchmarks with baseline >= 1 ms and absolute delta >= 0.1 ms.

Notable improvements

Benchmark	Baseline	PR	Delta
test_bi_init_from_dict[10000-int]	661.963 ms	55.507 ms	-91.61%
test_bi_init_from_dict[10000-str]	693.033 ms	61.633 ms	-91.11%
test_bi_init_fail_late_dupval[10000]	663.970 ms	59.886 ms	-90.98%
test_bi_init_from_dict[1000-int]	67.781 ms	6.549 ms	-90.34%
test_bi_init_from_dict[1000-str]	69.929 ms	7.132 ms	-89.80%
test_bi_init_fail_late_dupval[1000]	67.399 ms	7.151 ms	-89.39%
test_bi_update_partial_overlap[10000]	467.907 ms	70.205 ms	-85.00%
test_bi_update_partial_overlap[1000]	47.581 ms	7.999 ms	-83.19%

Notable regressions

Benchmark	Baseline	PR	Delta
test_bi_update_fail_early_dupval[10000]	1.016 ms	8.765 ms	+762.35%
test_bi_setitem_replace_existing_key[10000]	2.019 ms	2.292 ms	+13.53%
test_bi_forceput_existing_value[10000]	1.909 ms	2.096 ms	+9.81%
test_bi_forceput_existing_value[100]	1.719 ms	1.849 ms	+7.58%
test_bi_setitem_new_item[10000]	2.002 ms	2.110 ms	+5.39%
test_orderedbi_items_equals_with_equal_dict_items[10000]	10.646 ms	10.956 ms	+2.91%
test_bi_equals_with_equal_dict[10000]	10.648 ms	10.949 ms	+2.83%