Useful utilities for developers building Arabic and RTL Flutter/Dart applications.

arabic_dev_utils is a modular toolkit for MENA app development: normalize and classify Arabic text, search across it in a diacritic-insensitive way, convert between Arabic and Western digits, and infer RTL/LTR direction — without pulling in a framework or reimplementing locale-aware number formatting yourself.

Install

dependencies:
  arabic_dev_utils: ^0.1.0

Quick start

import 'package:arabic_dev_utils/arabic_dev_utils.dart';

// Text normalization
ArabicText.isArabic('مرحبا'); // true
ArabicText.normalize('أَحْمَد'); // 'احمد'
ArabicText.stripInvisibleChars('محمد‍علي'); // 'محمدعلي'

// Search
ArabicSearch.matches(query: 'احمد', text: 'أَحْمَد بن علي'); // true
ArabicSearch.rank('علي', ['محمد علي', 'علي حسن']); // ['علي حسن', 'محمد علي']

// Numbers
ArabicNumbers.toArabicDigits('100 ريال'); // '١٠٠ ريال'
ArabicNumbers.format(1234, locale: 'ar_EG'); // '١٬٢٣٤'

// RTL/direction
ArabicDirection.dominantDirection('مرحبا hello'); // ArabicTextDirection.mixed
ArabicDirectionality(
  basedOn: 'مرحبا',
  child: Text('مرحبا'), // rendered right-to-left automatically
);

Features

Text (ArabicText)

• isArabic, isArabicOnly, isMixed — script detection
• removeDiacritics, removeTatweel — strip tashkeel and kashida
• normalizeAlef, normalizeYeh, normalizeTehMarbuta — variant normalization
• normalize — applies all of the above in one call
• stripInvisibleChars — see Why invisible-character stripping matters
• wordCount, charCount — Arabic-aware counting

Search (ArabicSearch)

• matches — diacritic/tatweel/alef/yeh-insensitive containment check
• normalizeForSearch — the aggressive normalization used internally
• tokenize — normalized search tokens
• rank — sort candidates by match quality (exact → prefix → contains)

Numbers (ArabicNumbers)

• toArabicDigits, toEnglishDigits — convert between ASCII, Arabic-Indic (٠-٩), and Extended Arabic-Indic/Persian (۰-۹) digits
• containsArabicDigits — detect Arabic-Indic digits
• format — a thin, documented wrapper around package:intl's NumberFormat for Arabic-digit output; it exists for discoverability, not because intl is inadequate

RTL/direction (ArabicDirection)

• isRtl, isLtr, isMixedDirection — direction checks
• dominantDirection — returns an ArabicTextDirection (rtl, ltr, mixed, neutral)
• ArabicDirectionality — a Directionality-wrapping widget that infers direction from a string. package:intl's Bidi class already exposes direction detection, but this widget is easier to discover and use for the common "set this widget's direction from a string" case.

Why invisible-character stripping matters

Text copied from WhatsApp, Word, or many mobile keyboards can carry invisible characters — zero-width joiners, zero-width spaces, bidi control marks — that render identically to the naked eye but break equality checks, search, and diffs:

const fromWhatsApp = 'محمد‍علي'; // contains a hidden ZWJ
const typedByHand = 'محمدعلي';

fromWhatsApp == typedByHand; // false — a silent bug

ArabicText.stripInvisibleChars(fromWhatsApp) ==
    ArabicText.stripInvisibleChars(typedByHand); // true

Without stripping, this shows up as duplicate-looking entries in a database, search results that silently miss a match, or a diff tool reporting a change between two strings that look the same on screen.

Example app

See example/lib/main.dart for a full demo of all four modules, including a live search box over a list of Arabic names.

Contributing

Issues and pull requests are welcome. Please include tests for any new behavior and keep new public API fully documented with dartdoc comments.

License

MIT — see LICENSE.