Changelog

First release.

Added#

• slug() with script: 'keep', which leaves the Arabic letters in place — the option no other slug library offers, and the only way to produce a URL an Arabic reader can read.
• Zero-width joiners, non-joiners and kashidas treated as what they are: marks inside a word, never separators. Two of the most downloaded slug libraries turn می‌روم into two words.
• Vowel marks used when an author has written them, including a shadda doubling the consonant rather than the vowel — which Unicode's ordering makes easy to get backwards — and the silent alef after a tanween.
• Language-aware readings for Arabic, Persian and Urdu: positional و and ی, final heh as eh, and the letters پ چ ژ گ ٹ ڈ ڑ ں that an Arabic-only table drops entirely.
• Letter unification that follows the content's language: Persian and Urdu keep ی and ک, Arabic keeps ي and ك, and both spellings still collapse to one slug.
• A closing ta marbuta as a and an internal one as t.
• The definite article as al-, with الله as the single exception.
• Arabic-Indic and Persian digits converted to 0–9.
• Direction marks stripped, so a right-to-left override can never make a URL read as something other than what it is.
• uniqueSlug() for collisions, which romanised Arabic produces far more often than Latin text does.
• prefersScript(), normalise(), transliterate(), westernDigits(), stripInvisible() and hasTashkeel() exported on their own — the same normalisation makes Arabic search work.
• separator, lower, maxLength that cuts on a word boundary, stopWords that never empty a title, and keep for extra characters.
• 2.0 kB gzip, zero dependencies.