Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

v0.1.0 - 2026-08-02

First tagged release of garabic — a set of functions for Arabic text processing in Go.

Normalize — normalize Arabic text for searching and indexing.
RemoveHarakat — strip diacritics (tashkeel) from Arabic text.
Shape — reconstruct Arabic and Persian glyphs (including پ چ ژ گ) so text renders connected and in the right order in images and other non-RTL-aware surfaces.
SpellNumber — spell out numbers in Arabic words.
Tashkeel — add matching diacritics to Arabic text (initial support).
Justify — justify an Arabic sentence to a given width using kashida (tatweel) stretching, with space-padding fallback.
ToArabicDigits, ToPersianDigits, ToEnglishDigits — convert between English, Arabic and Persian digits.
IsArabic, IsArabicLetter — Arabic-script detection helpers.
Test suites for Arabic and Persian shaping, digits, justification and tashkeel (96% coverage).
GitHub Actions workflows for build checks and race-enabled tests.

Shape no longer misplaces diacritics on words combining tashkeel with a lam-alef ligature (لا / لأ / لإ / لآ), and handles tatweel in input correctly.
Tashkeel no longer panics when a sentence ends with a preposition (حرف جر).
Justify no longer panics on empty input, single words, or tashkeel with lam-alef; width math is now rune-based and exact.
ToPersianDigits now emits proper Extended Arabic-Indic codepoints (۰–۹, U+06F0–U+06F9) for all ten digits instead of a mix of Arabic and Persian forms; ToEnglishDigits accepts both.