nmarusov / antlr4ts

Optimized TypeScript target for ANTLR 4

antlr4ts - TypeScript/JavaScript target for ANTLR 4

Overview

• Releases: See the GitHub Releases page for release notes and links to the distribution
• Feedback: Use GitHub Issues for issues (bugs, enhancements, features, and questions)

Requirements

This project has separate requirements for developers and end users.

💡 The requirements listed on this page only cover user scenarios - that is, scenarios where developers wish to use ANTLR 4 for parsing tasks inside of a TypeScript application. If you are interested in contributing to ANTLR 4 itself, see CONTRIBUTING.md for contributor documentation.

End user requirements

Parsers generated by the ANTLR 4 TypeScript target have a runtime dependency on the antlr4ts package. The package is tested and known to work with Node.js 6.7.

Development requirements

The tool used to generate TypeScript code from an ANTLR 4 grammar is written in Java. To fully utilize the ANTLR 4 TypeScript target (including the ability to regenerate code from a grammar file after changes are made), a Java Runtime Environment (JRE) needs to be installed on the developer machine. The generated code itself uses several features new to TypeScript 2.0.

• Java Runtime Environment 1.6+ (1.8+ recommended)
• TypeScript 2.0+

Getting started

1. Install antlr4ts as a runtime dependency using your preferred package manager.

npm install antlr4ts --save

yarn add antlr4ts

2. Install antlr4ts-cli as a development dependency using your preferred package manager.

npm install antlr4ts-cli --save-dev

yarn add -D antlr4ts-cli

3. Add a grammar to your project, e.g. path/to/MyGrammar.g4

4. Add a script to package.json for compiling your grammar to TypeScript

   "scripts": {
     // ...
     "antlr4ts": "antlr4ts -visitor path/to/MyGrammar.g4"
   }
   

5. Use your grammar in TypeScript

   import { ANTLRInputStream, CommonTokenStream } from 'antlr4ts';
   
   // Create the lexer and parser
   let inputStream = new ANTLRInputStream("text");
   let lexer = new MyGrammarLexer(inputStream);
   let tokenStream = new CommonTokenStream(lexer);
   let parser = new MyGrammarParser(tokenStream);
   
   // Parse the input, where `compilationUnit` is whatever entry point you defined
   let tree = parser.compilationUnit();

   The two main ways to inspect the tree are by using a listener or a visitor, you can read about the differences between the two here.

   Listener Approach

   // ...
   import { MyGrammarParserListener } from './MyGrammarParserListener'
   import { FunctionDeclarationContext } from './MyGrammarParser'
   import { ParseTreeWalker } from 'antlr4ts/tree/ParseTreeWalker'
   
   
   class EnterFunctionListener implements MyGrammarParserListener {
     // Assuming a parser rule with name: `functionDeclaration`
     enterFunctionDeclaration(context: FunctionDeclarationContext) {
       console.log(`Function start line number ${context._start.line}`)
       // ...
     }
   
     // other enterX functions...
   }
   
   // Create the listener
   const listener: MyGrammarParserListener = new EnterFunctionListener();
   // Use the entry point for listeners
   ParseTreeWalker.DEFAULT.walk(listener, tree)

   Visitor Approach

   Note you must pass the -visitor flag to antlr4ts to get the generated visitor file.

   // ...
   import { MyGrammarParserVisitor } from './MyGrammarParserVisitor'
   import { AbstractParseTreeVisitor } from 'antlr4ts/tree/AbstractParseTreeVisitor'
   
   // Extend the AbstractParseTreeVisitor to get default visitor behaviour
   class CountFunctionsVisitor extends AbstractParseTreeVisitor<number> implements MyGrammarParserVisitor<number> {
   
     defaultResult() {
       return 0
     }
   
     aggregateResult(aggregate: number, nextResult: number) {
       return aggregate + nextResult
     }
   
     visitFunctionDeclaration(context: FunctionDeclarationContext): number {
       return 1 + super.visitChildren(context)
     }
   }
   
   // Create the visitor
   const countFunctionsVisitor = new CountFunctionsVisitor()
   // Use the visitor entry point
   countFunctionsVisitor.visit(tree)