Newest stable release for `tivars_lib_py`. Install from this tag or [from PyPI](https://pypi.org/project/tivars/0.9.1/).
**Full Changelog**: https://github.com/TI-Toolkit/tivars_lib_py/commits/v0.9.1
New features
- Smart tokenization mode — token munching better reflects program context
- New `tokenizer.normalize` function to normalize certain Unicode chars for tokenization
- New `Self` type annotations for 3.11+ to help static type analyzers
- All `typing` type aliases (to-be deprecated in a future version of Python) have been removed
- More warnings about invalid files and entry names
Bug fixes
- Fixed `ListEntry.load_list` and `TIMatrix.load_matrix` not working
- Entry name tokenization is now more consistent
Breaking changes
- Smart tokenization mode is now the default; use `mode=max` to enable previous behavior
- Some functions in `tokenizer` now use keyword-only arguments
Contributions
- adriweb - Set the spec for smart tokenization, reflecting much of the TI Connect CE program editor's behavior