unret

unret takes a regexp syntax tree generated by ret and converts it back to a string.

Installation

$ npm install unret

Usage

var ret   = require("ret");
var unret = require("unret");
 
// Parse an expression with ret
var syntaxTree = ret(/regexp/.source);
 
// Manipulate the syntax tree
var newTree = doSomething(syntaxTree);
 
// Convert the syntax tree to a string
var newExpression = unret(newTree);
 
// Print the new expression
console.log(newExpression);
 
// Create a new RegExp object
var myExp = RegExp(newExpression);
 
// etc, etc

Testing

Tests use Chai and Mocha.

$ npm test

Limitations

Certain aspects of ret's syntax tree mean that parsing an expression and then immediately converting it to a string will not return the original expression.

• Quantifier symbols - ?, +, and * - are expanded to {min,max} quantifiers
• Character class escapes - \w, \d, etc - are expanded to the equivalent character class, e.g. [_a-zA-Z0-9]
• Escaped characters - \n, \u0065, etc - are expanded to the equivalent character.

In these cases, unret does not attempt to guess the original notation or take an opinion on which is preferable. So long as ret is generating a syntax tree equivalent to the original expression, the string returned by unret should represent an equivalent expression.

In all cases, once a ret tree has been converted to a string by unret, parsing it with ret and again converting to a string (however many times) should yield the first string returned, i.e. a fixed point is reached.