- HOTFIX: Prevent error message about missing file in CLI use
0.2.8
- FIX: Fixed inheritence when it chains, e.g. Algerian Arabic inheriting from Tunisian Arabic which inherits from Standard Arabic - FIX: Fixed inheritence missing `marks`, `design_notes` and `note` - TWEAK: Make sure `marks` are saved in ordered form, so saving does not arbitrarily alter the order - TESTS: Added tests for orthography inheritance - DATA: Constrained speaker counts to integers only
0.2.7
- DATA: Fixed various speaker counts containing malformed data - DATA: More design notes for Latin-script languages - DATA: Khmer added as draft, Armenian, Buginese, Georgian, Burmese, Lao and Thai refined - TWEAK: Implemented validation for speaker count data
0.2.6
- DATA: Various status updates, notes and reviewed orthographies - DATA: Introduced `marks` attribute containing all combinging marks needed for an orthography - FEATURE: Automatically extract and save `marks` from `base` data, plus retain any explicitly added `marks` in the data - TWEAK: For default `hyperglot-save` calls automatically run validation to flag any remaining issues - TWEAK: Flag legacy marks being used in charset data
0.2.5
- DATA: Introduced `design_note` parameter - DATA: Various language data updates and smaller fixes - DATA: Several orthography fixes, thanks Denis Moyogo Jacquerye - TWEAK: Changed orthography status names to `todo, draft, preliminary, verified` - TWEAK: Improved `Language.get_orthography` to return better default picks and allow getting orthographies of specific script or status