A lexical analyzer (tokenizer) generator in TypeScript.

TLEX provides a flexible tokenizer with regex-based pattern matching, support for both JavaScript and Flex-style regex syntax, and stateful lexing. It works in Node.js and browsers with full TypeScript support.

Full documentation and interactive examples: panyam.github.io/tlex

Try tokenizers in the browser: Playground

npm install tlex

import { Tokenizer } from "tlex";

const tokenizer = new Tokenizer();

// Add token rules
tokenizer.add(/[0-9]+/, { tag: "NUMBER" });
tokenizer.add(/[a-zA-Z_][a-zA-Z0-9_]*/, { tag: "IDENTIFIER" });
tokenizer.add(/\+|\-|\*|\//, { tag: "OPERATOR" });
tokenizer.add(/\s+/, { tag: "WS", skip: true });

// Tokenize input
const tokens = tokenizer.tokenize("x + 42 * y");
// Returns: [IDENTIFIER "x", OPERATOR "+", NUMBER "42", OPERATOR "*", IDENTIFIER "y"]

• Regex-based patterns - Use JavaScript regex or Flex-style patterns
• Rule priorities - Control which rule matches on conflicts
• Skip tokens - Automatically skip whitespace and comments
• Stateful lexing - Context-sensitive tokenization with lexer states
• Token callbacks - Custom handlers for token processing
• Lookahead support - TokenBuffer for parser lookahead
• TypeScript native - Full type definitions included

The documentation includes several runnable examples:

• JSON Tokenizer - Complete JSON lexer
• Calculator - Expression tokenizer
• C Lexer - C-style language with states

• API Reference - Tokenizer, Token, Rule classes
• JS Regex Syntax - JavaScript regex support
• Flex Syntax - Flex-style patterns
• Rule Configuration - Priority, skip, states

MIT