Parker

Parker is a small JavaScript parser and tokenizer that turns JavaScript/TypeScript-like source code into an abstract syntax tree (AST). It is designed for tooling, code analysis, and experimentation rather than being a full drop-in replacement for production parsers. Work in progress.

Features

• Lexer: Converts source text into a stream of tokens.
• Parser: Produces a rich AST using node classes from src/node.js.
• Options support: Optional location metadata on nodes (line / column).
• Playground: Simple CLI playground under playground/ for quick experiments.
• Go prototype: Early Go-based lexer implementation under go/src/.

Getting started

Prerequisites

• Node.js (recommended: current LTS)
• npm (or another Node package manager)

Installation

From the parker directory:

npm install

Parker is published as a Node module inside this repo (see package.json). You can either:

• Use it directly from this repo (require('./src')), or
• Publish it / link it locally and depend on it from another project.

Usage

Basic parsing

const Parker = require('parker'); // or: require('./src')

const source = `
  export class Renderer {
  }
`;

const ast = Parker.parse(source, { locations: true });
console.log(JSON.stringify(ast, null, 2));

Tokenizing

const Parker = require('parker'); // or: require('./src')

const source = `
  export class Renderer {
  }
`;

const tokens = Parker.tokenizeAll(source, { locations: true });
console.log(tokens);

Options

Parker accepts an options object in both parse and tokenizeAll:

• locations: boolean
  When true, the parser and lexer attach location information to tokens and AST nodes.

Example:

const ast = Parker.parse(source, { locations: true });

API

Parker.parse(source, options?)

• source: string – Source code to parse.
• options: object (optional)
  • locations?: boolean

Returns: An AST NodeProgram instance representing the full program.

Parker.tokenizeAll(source, options?)

• source: string – Source code to tokenize.
• options: object (optional)
  • locations?: boolean

Returns: An array of token objects from the underlying lexer.

Playground

This repository includes a simple playground script to inspect the AST for a sample file.

From the parker directory:

npm run play

This will:

• Read playground/test.js
• Parse it with Parker.parse
• Print the resulting AST using Node’s util.inspect.

You can change the sample code by editing playground/test.js.

Development

• Main entry: src/index.js (exports the Parker class from src/parker.js).
• Lexer: src/lexer.js
• Parser: src/parser.js
• Nodes: src/node.js
• Options: src/options.js

For local hacking:

# Install dependencies
npm install

# Run playground
npm run play

License

MIT