Madhu ML TT — Core Malayalam Text Engine

Madhu ML TT is a high-performance, developer-focused Malayalam text engine designed for bidirectional conversion between modern Unicode Malayalam and legacy ML-TT (Malayalam TrueType) font encodings.

Capabilities & Features

• Bidirectional Conversion: Accurate Unicode -> ML-TT and ML-TT -> Unicode transformation.
• Sequence-Aware Matra Reordering: Handles pre-base matras (െ, േ, ൈ, ്ര) and split matras (ൊ, ോ, ൌ).
• Atomic Chillu Standardization: Full support for both atomic Unicode chillus (U+0D7K–U+0D7F) and legacy virama-based sequences (ന+്+ZWJ).
• Mixed Content Preservation: Retains English words, numbers, emails, URLs, punctuation (.,!?;:()[]{}'"-/_), and whitespace.
• Data-Driven Architecture: Decoupled engine algorithms and font mapping data (ML-TTKarthika C-DIT / GIST standard).
• Framework Independent: Pure ES Module with zero runtime dependencies. Runs in Node.js, browsers, or web workers.

Installation & Basic Usage

import { unicodeToMLTT, mlttToUnicode } from './src/index.js';

// 1. Unicode to ML-TT Conversion
const mltt = unicodeToMLTT('കേരളം');
console.log(mltt); // Output: "tIcfw"

// 2. ML-TT to Unicode Conversion
const unicode = mlttToUnicode('tIcfw');
console.log(unicode); // Output: "കേരളം"

// 3. Mixed Text Conversion with English Preservation
const mixedMltt = unicodeToMLTT('Hello കേരളം 2026!');
console.log(mixedMltt); // Output: "Hello tIcfw 2026!"

const restoredUnicode = mlttToUnicode(mixedMltt, { preserveEnglish: true });
console.log(restoredUnicode); // Output: "Hello കേരളം 2026!"

Verification & Test Suite

Run the master test runner containing 19 test suites:

npm test

Included Test Categories

tests/
├── conformance/      # Fixed reference fixture tests for both directions
├── unicode-to-mltt/  # Vowels, consonants, matras, conjuncts, chillus, mixed
├── mltt-to-unicode/  # Basic, matras, conjuncts, chillus, mixed
├── edge-cases/       # Malformed input, unknown chars, normalization matrix
├── regression/       # Bug fix regression protection
├── roundtrip/        # 100% roundtrip fidelity suite (50 assertions)
├── performance/      # Throughput benchmarks (100 B to 1 MB)
└── font-verification.py # fontTools glyph correspondence verification

Current Scope & Future Roadmap

• Phase 1 (Complete): Technical specification, font audit, reordering engine, test suite, and core release readiness.
• Phase 2 (Planned): Developer SDK (@madhu-mltt/core), React components, and Vite plugin integration.

License

MIT License.