Add to_unicode, permissive, and recovery_mode flags to MARCReader

[dchud]

Summary

Fixes #78 — adds pymarc-compatible to_unicode and permissive kwargs to MARCReader, and exposes mrrc's existing RecoveryMode as a recovery_mode kwarg.

• to_unicode: accepted for pymarc compat; mrrc always converts MARC-8 → UTF-8, so False emits a warning but has no effect
• permissive=True: yields None for records that fail to parse, matching pymarc's behavior exactly
• recovery_mode: exposes mrrc's Rust-native RecoveryMode ("strict", "lenient", "permissive") for salvaging partial data from damaged records
• Combining permissive=True with non-strict recovery_mode raises ValueError (conflicting strategies)

No Rust core changes — permissive is handled in the Python wrapper's __next__, recovery_mode is passed through PyO3 to the existing MarcReader::with_recovery_mode().

Test plan

• 15 new Python tests in test_marcreader_flags.py covering all kwargs, edge cases, and conflict validation
• All 775 Rust tests pass
• All 611 Python tests pass
• Full .cargo/check.sh passes

Bead: bd-y331

🤖 Generated with Claude Code

[codspeed-hq]

Merging this PR will improve performance by 28.68%

⚠️ 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.

⚡ 8 improved benchmarks
✅ 52 untouched benchmarks
⏩ 16 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	WallTime	test_process_4_files_sequential	85.7 ms	75.7 ms	+13.26%
⚡	WallTime	test_pipeline_parallel_2x_10k_threaded	50.7 ms	41.8 ms	+21.23%
⚡	WallTime	test_pipeline_parallel_4x_10k_threaded	104.4 ms	88.1 ms	+18.54%
⚡	WallTime	test_process_4_files_parallel_4_threads	108.7 ms	93.7 ms	+16%
⚡	WallTime	test_pipeline_sequential_extraction_4x_10k	105.5 ms	94 ms	+12.17%
⚡	WallTime	test_pipeline_sequential_4x_10k	84.5 ms	76.6 ms	+10.28%
⚡	WallTime	test_file_parallel_4x_10k_with_extraction	1,096.1 ms	851.8 ms	+28.68%
⚡	WallTime	test_pipeline_sequential_1x_10k	20.4 ms	18.4 ms	+11.12%

---

_{Comparing bd-y331-marcreader-flags (9ec3c31) with main (dd4c671)}

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. ↩