v0.1.0 is live: native-speaker validation for the next lexicon batch

[Samielakkad]

darija-tools now has a first tagged release: v0.1.0.

It provides dependency-free Moroccan Darija text normalization, Arabizi-to-Arabic transliteration, optional digit and whitespace normalization, loanword preservation, and a small CLI. The package is tested on Python 3.9 through 3.13.

The next priority is native-speaker validation rather than adding a large unreviewed word list. Issue #5 contains the first proposed batch and its current fallback output.

Useful contributions include:

• confirming or correcting one proposed Arabic spelling;
• adding a common regional Arabizi variant that maps to the same form;
• supplying a minimal sentence that demonstrates the expected spelling;
• implementing a validated subset with focused tests.

Please mention your region when a spelling is regional, and claim a subset before opening a PR so work is not duplicated. The transliterator is intentionally transparent and conservative: a small reviewed mapping is more useful than a large speculative one.