fix(mpa): skip blank/commented lines and warn on rarefaction-excluded samples

[ilypopv]

📝 Summary

Hardens the Kraken/Bracken report parser to gracefully skip blank and #-commented lines, fixing a latent crash where such lines fell through to a ValueError on tuple unpacking. Additionally, calc_beta_div now emits a UserWarning naming every sample excluded for falling below the rarefaction depth instead of silently dropping them.

🛠 Type of Change

• Bug fix
• New feature

🔍 Key Changes

• transform2mpa._parse_line: strip the line and return [] immediately for blank or #-commented input, preventing downstream unpacking errors.
• transform2mpa.kreport_to_mpa: replace the is None check with not report_vals so the empty-list return from _parse_line is correctly treated as a skip.
• krakenparser.py: add import warnings and override warnings.formatwarning to suppress source file/line noise, showing only the message text; broaden the main_callback error handler to also catch ValueError.
• stats/diversity.calc_beta_div: track excluded samples with their total read counts and emit a UserWarning (plus a log warning) listing them when any are dropped below rarefaction_depth; added a Warns: docstring section.
• Tests: added unit tests for blank/commented line handling in _parse_line, an integration test verifying noisy reports produce identical output to clean ones, and a test asserting the exclusion warning names the affected sample.

🧪 How Has This Been Tested?

• Unit tests (test_units.py) verify _parse_line returns [] for blank ("\n", " \n") and commented ("# some comment\n") lines.
• Integration test (test_integration.py) confirms a report with leading blank/comment lines yields byte-identical output to the clean report.
• Integration test asserts calc_beta_div raises UserWarning matching the excluded sample id (S3) when a sample is below the rarefaction depth.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!