0.1.6

Added

• New peek() function to extract BAM file metadata (contigs and modifications) without full processing
• New seq_table() function to extract read sequences and base qualities for a genomic region as a Polars DataFrame, with insertions shown as lowercase, deletions as periods, and modifications as 'Z'
• New tag parameter for polars_bam_mods, read_info, window_reads, and seq_table functions to filter by specific modification type (e.g., single-letter code "m" or ChEBI code "76792")
• New win_op parameter for window_reads() function supporting "density" (default) and "grad_density" modes to measure modification density or gradient within each window
• PyPI publishing via GitHub Actions using OIDC trusted publishing
• Comprehensive filtering parameter tests for polars_bam_mods, read_info, window_reads, and seq_table functions
• Gradient window tests with new example BAM files (example_10, example_11)
• New two_mods_bam test fixture for testing multiple modification types
• New pynanalogue-specific test data files for seq_table testing

Changed

• Removed flaky 10% sampling test case in seq_table testing due to stochastic fluctuations with small sample sizes
• Updated nanalogue dependency from 0.1.4 to 0.1.6
• Consolidated build-wheels.yml into publish_to_pypi.yml workflow
• Re-enabled musl benchmark job in CI
• Renamed test_filtering.py to test_polars_bam_mods_filtering.py for clarity

Fixed

• Fixed mismatch generation bug when simulating BAM files (via nanalogue 0.1.6)
• Fixed full_region misinterpretation when an open-ended interval or entire contig was provided in the region parameter (via nanalogue 0.1.6)