happy-codec

Zero-dependency codec library for UTF-8, Base64, Hex and ByteString encoding/decoding.

---

API Documentation

---

Features

• Zero dependencies - No external runtime dependencies
• Universal runtime - Works in any JavaScript environment (Browser, Node.js, Deno, Bun, Web Workers, 小程序/小游戏 (如微信小游戏), etc.) regardless of DOM/BOM support
• UTF-8 - Native TextEncoder/TextDecoder when available, pure JS fallback otherwise
• Base64 - Native Uint8Array.prototype.toBase64/Uint8Array.fromBase64 (ES2026) when available, pure JS fallback otherwise
• Hex - Native Uint8Array.prototype.toHex/Uint8Array.fromHex (ES2026) when available, pure JS fallback otherwise
• ByteString - Binary string conversions

Installation

# npm
npm install happy-codec

# yarn
yarn add happy-codec

# pnpm
pnpm add happy-codec

# JSR (Deno)
deno add @happy-js/happy-codec

# JSR (Bun)
bunx jsr add @happy-js/happy-codec

Usage

import {
    decodeBase64,
    decodeByteString,
    decodeHex,
    decodeUtf8,
    encodeBase64,
    encodeByteString,
    encodeHex,
    encodeUtf8,
} from 'happy-codec';

// UTF-8
const utf8Bytes = encodeUtf8('你好'); // Uint8Array [228, 189, 160, ...]
const text = decodeUtf8(utf8Bytes); // '你好'
// Invalid bytes are replaced with U+FFFD by default
const lossy = decodeUtf8(new Uint8Array([0xff, 0xfe])); // '��'
// Use { fatal: true } to throw on invalid bytes
decodeUtf8(new Uint8Array([0xff]), { fatal: true }); // throws TypeError
// BOM is stripped by default, use { ignoreBOM: true } to keep it
const withBOM = new Uint8Array([0xef, 0xbb, 0xbf, 0x48, 0x69]); // BOM + 'Hi'
decodeUtf8(withBOM); // 'Hi'
decodeUtf8(withBOM, { ignoreBOM: true }); // '\uFEFFHi'

// Base64
const base64 = encodeBase64('Hello, World!'); // 'SGVsbG8sIFdvcmxkIQ=='
const bytes = decodeBase64(base64); // Uint8Array
// Base64url alphabet
const url = encodeBase64('Hello', { alphabet: 'base64url' }); // 'SGVsbG8'
decodeBase64(url, { alphabet: 'base64url' }); // Uint8Array
// Omit padding
encodeBase64(new Uint8Array([65]), { omitPadding: true }); // 'QQ' (no trailing '==')
// Strict decoding: require correct padding
decodeBase64('QQ==', { lastChunkHandling: 'strict' }); // Uint8Array [65]
// decodeBase64('QQ', { lastChunkHandling: 'strict' }); // throws SyntaxError

// Hex
const hex = encodeHex(new Uint8Array([255, 0, 128])); // 'ff0080'
const data = decodeHex('ff0080'); // Uint8Array [255, 0, 128]

// ByteString
const byteStr = encodeByteString(new Uint8Array([72, 101, 108, 108, 111])); // 'Hello'
const byteArr = decodeByteString('Hello'); // Uint8Array [72, 101, 108, 108, 111]

Performance

All encoding/decoding functions use native APIs when available, with pure JS fallback for environments that don't support them yet.

• Base64: Native Uint8Array.prototype.toBase64 / Uint8Array.fromBase64 (ES2026) when available, pure JS fallback otherwise.
• Hex: Native Uint8Array.prototype.toHex / Uint8Array.fromHex (ES2026) when available, pure JS fallback otherwise.
• UTF-8: Native TextEncoder / TextDecoder when available, pure JS fallback otherwise.

Run pnpm run bench to compare fallback vs native performance on your own environment.

License

MIT