regexp-to-ast

Reads a JavaScript Regular Expression literal(text) and outputs an Abstract Syntax Tree.

Installation

• npm

    npm install regexp-to-ast

• Browser

    <script src="https://unpkg.com/regexp-to-ast/lib/parser.js"></script>

API

The API is defined as a TypeScript definition file.

Usage

• Parsing to an AST:

  const RegExpParser = require("regexp-to-ast").RegExpParser
  const regexpParser = new RegExpParser.parser()
  
  // from a regexp text
  const astOutput = regexpParser.pattern("/a|b|c/g")
  
  // text from regexp instance.
  const input2 = /a|b/.toString()
  // The same parser instance can be reused
  const anotherAstOutput = regexpParser.pattern(input2)

• Visiting the AST:

  // parse to an AST as before.
  const { RegExpParser, BaseRegExpVisitor } = require("regexp-to-ast")
  const regexpParser = new RegExpParser.parser()
  const regExpAst = regexpParser.pattern("/a|b|c/g")
  
  // Override the visitor methods to add your logic.
  class MyRegExpVisitor extends BaseRegExpVisitor {
      visitPattern(node) {}
  
      visitFlags(node) {}
  
      visitDisjunction(node) {}
  
      visitAlternative(node) {}
  
      // Assertion
      visitStartAnchor(node) {}
  
      visitEndAnchor(node) {}
  
      visitWordBoundary(node) {}
  
      visitNonWordBoundary(node) {}
  
      visitLookahead(node) {}
  
      visitNegativeLookahead(node) {}
  
      // atoms
      visitCharacter(node) {}
  
      visitSet(node) {}
  
      visitGroup(node) {}
  
      visitGroupBackReference(node) {}
  
      visitQuantifier(node) {}
  }
  
  const myVisitor = new MyRegExpVisitor()
  myVisitor.visit(regExpAst)
  // extract visit results from the visitor state.

Compatibility

This library is written in ES5 style and is compatiable with all major browsers and modern node.js versions.

TODO / Limitations

• Use polyFill for string.prototype.at to support unicode characters outside BMP.
• Descriptive error messages.
• Position information in error messages.
• Support unicode flag escapes.
• Ensure edge cases described in "The madness of parsing real world JavaScript regexps" are supported.
• Support deprecated octal escapes