v0.2.2

Changed

• Rewrote sanitizer to use rdBase.BlockLogs() + Chem.DetectChemistryProblems() instead of unsafe sys.stderr redirection
• _add_atoms returns dict[str, int] for O(1) bond lookup (was O(n) list scan)
• _str_to_float returns float | None to distinguish missing CIF values from real zeros
• Reject degenerate conformers (>1 atom at origin) instead of checking single-atom origin
• sanitize() always works on a copy; the original input molecule is never modified

Added

• ty type checker in dev dependencies and CI lint job
• Release version validation (tag vs pyproject.toml) in release workflow
• Ruff rules: SIM, PERF, RUF, FURB
• Dependency upper bounds (gemmi<1, rdkit<2026, rich<15, typer<1)
• 38 new tests (73 → 95 total), overall coverage 82% → 91%
• API reference and advanced usage examples in README

Fixed

• Version mismatch between __init__.py and pyproject.toml (now uses importlib.metadata)
• sanitize() no longer mutates the input molecule on failure/exception paths
• Bond error handling: KeyError for missing atoms, RuntimeError for duplicate bonds

v0.2.1

Fixed

• Updated README with optional CLI installation instructions
• Updated CHANGELOG for v0.2.0 changes

v0.2.0

chore: bump version to 0.2.0

v0.1.0

Added

• Initial release
• CIF file parsing using gemmi
• Conversion to RDKit molecule objects
• Support for Ideal and Model 3D conformers
• Automatic metal bond to dative bond conversion
• Stereochemistry assignment from 3D coordinates
• CLI tool with convert and info commands
• Rich terminal output support
• Options for sanitization, conformer handling, and hydrogen removal