test(proptest): expand property suite to 8 tests across binary/XML/JSON

[dchud]

Summary

• Adds 4 structural-invariant properties over the emitted ISO 2709 bytes (leader length matches byte count, directory entries tile the data area, indicator bytes are digit-or-space, subfield codes are [a-z0-9]) plus MARCXML and MARCJSON round-trip properties that exercise format-specific escaping (< > & " ' for XML; \t \n \r \\ " for JSON).
• Broadens arb_record to cover zero-control-field and zero-data-field records; caps runtime with ProptestConfig { cases: 64 } (suite runs in ~3s locally, well under the 10s budget).
• Removes the stray commented #proptest = "1.0" from Cargo.toml [dependencies] (the real entry has always been in [dev-dependencies]).
• Documents regression-seed policy inline and adds tests/*.proptest-regressions.pending / tests/proptest-regressions/*.pending to .gitignore so only accepted seeds are tracked.

The MARCXML round-trip property surfaced a real bug: quick-xml + serde(rename = "$value") strips whitespace-only text content, so a <controlfield tag=\"001\"> </controlfield> fails to deserialize. The XML strategy now excludes whitespace-only values and the fix is tracked in a separate bead (bd-zm6m).

Test plan

• cargo test --test properties --package mrrc → 8 passed in 2.6s
• .cargo/check.sh (full local CI) → all checks passed
• CI green on GitHub across all platforms

Bead: bd-ouji
Closes #91

🤖 Generated with Claude Code

[codspeed-hq]

Merging this PR will degrade performance by 35.39%

⚠️ 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

⚡ 1 improved benchmark
❌ 55 regressed benchmarks
✅ 4 untouched benchmarks
⏩ 16 skipped benchmarks¹

⚠️ Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	WallTime	test_threaded_reading_4x_10k	692.2 ms	895.7 ms	-22.71%
❌	WallTime	test_threaded_reading_4x_1k	66 ms	85.7 ms	-22.95%
❌	WallTime	test_threaded_reading_1k	20.8 ms	25.6 ms	-19.02%
❌	WallTime	test_sequential_2x_reading_10k	133.5 ms	157.3 ms	-15.12%
❌	WallTime	test_sequential_2x_reading_1k	12.9 ms	15.6 ms	-16.79%
❌	WallTime	test_sequential_10k	66.2 ms	78.6 ms	-15.88%
❌	WallTime	test_threaded_reading_2x_10k	204.7 ms	262.7 ms	-22.09%
❌	WallTime	test_sequential_4x_reading_1k	26.1 ms	30.7 ms	-14.89%
❌	WallTime	test_sequential_with_title_extraction_10k	69.1 ms	81.3 ms	-15.02%
❌	WallTime	test_process_4_files_sequential	67.4 ms	78.7 ms	-14.31%
❌	WallTime	test_iterator_vs_while_loop_1k	6.3 ms	7.4 ms	-14.22%
❌	WallTime	test_process_4_files_parallel_4_threads	85.9 ms	97.6 ms	-12%
❌	WallTime	test_read_10k_records	64.4 ms	76.3 ms	-15.63%
❌	WallTime	test_read_1k_records	6.4 ms	7.5 ms	-15.1%
❌	WallTime	test_threading_speedup_4x_10k	668.6 ms	863.7 ms	-22.6%
❌	WallTime	test_roundtrip_1k_records	9.5 ms	11.5 ms	-17.27%
❌	WallTime	test_threaded_with_title_extraction_2x_10k	221.2 ms	272.1 ms	-18.73%
❌	WallTime	test_read_and_extract_titles_10k	67 ms	79.7 ms	-15.97%
❌	WallTime	test_parallel_read_with_extract_4x_1k	70.7 ms	94.5 ms	-25.19%
❌	WallTime	test_threading_speedup_2x_10k	202.4 ms	263.5 ms	-23.19%
...	...	...	...	...	...

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

---

_{Comparing bd-ouji-proptest-expansion (bff1c57) with main (1ebc492)²}

Footnotes

1. 16 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. ↩

2. No successful run was found on main (bff1c57) during the generation of this report, so 1ebc492 was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩