Extended JavaScript regular expressions
Pull request Compare This branch is 136 commits behind slevithan:master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


XRegExp 3.0.0-pre

XRegExp provides augmented and extensible JavaScript regular expressions. You get new syntax, flags, and methods beyond what browsers support natively. XRegExp is also a regex utility belt with tools to make your client-side grepping simpler and more powerful, while freeing you from worrying about pesky cross-browser inconsistencies and the dubious lastIndex property.

XRegExp supports all native ES5 regular expression syntax. It works with Internet Explorer 5.5+, Firefox 1.5+, Chrome, Safari 3+, and Opera 11+. You can also use it on the server with Node.js, or as a RequireJS module. The base library is about 3.8 KB, minified and gzipped.

See what's new in version 3.0.0-pre.


XRegExp regexes compile to native RegExp objects, and therefore perform just as fast as native regular expressions. There is a small extra cost when compiling a pattern for the first time.

Usage examples

// Using named capture and flag x (free-spacing and line comments)
var date = XRegExp('(?<year>  [0-9]{4} ) -?  # year  \n\
                    (?<month> [0-9]{2} ) -?  # month \n\
                    (?<day>   [0-9]{2} )     # day   ', 'x');

// XRegExp.exec gives you named backreferences on the match result
var match = XRegExp.exec('2012-02-22', date);
match.year; // -> '2012'

// It also includes optional pos and sticky arguments
var pos = 3, result = [];
while (match = XRegExp.exec('<1><2><3><4>5<6>', /<(\d+)>/, pos, 'sticky')) {
    pos = match.index + match[0].length;
} // result -> ['2', '3', '4']

// XRegExp.replace allows named backreferences in replacements
XRegExp.replace('2012-02-22', date, '${month}/${day}/${year}'); // -> '02/22/2012'
XRegExp.replace('2012-02-22', date, function(match) {
    return match.month + '/' + match.day + '/' + match.year;
}); // -> '02/22/2012'

// In fact, XRegExps compile to RegExps and work perfectly with native methods
date.test('2012-02-22'); // -> true

// The *only* caveat is that named captures must be referenced using numbered backreferences
'2012-02-22'.replace(date, '$2/$3/$1'); // -> '02/22/2012'

// If you want, you can extend native methods so you don't have to worry about this.
// Doing so also fixes numerous browser bugs in the native methods
'2012-02-22'.replace(date, '${month}/${day}/${year}'); // -> '02/22/2012'
'2012-02-22'.replace(date, function(match) {
    return match.month + '/' + match.day + '/' + match.year;
}); // -> '02/22/2012'
date.exec('2012-02-22').year; // -> '2012'

// Extract every other digit from a string using XRegExp.forEach
XRegExp.forEach('1a2345', /\d/, function(match, i) {
    if (i % 2) this.push(+match[0]);
}, []); // -> [2, 4]

// Get numbers within <b> tags using XRegExp.matchChain
XRegExp.matchChain('1 <b>2</b> 3 <b>4 a 56</b>', [
]); // -> ['2', '4', '56']

// You can also pass forward and return specific backreferences
var html = '<a href="http://xregexp.com/">XRegExp</a>' +
           '<a href="http://www.google.com/">Google</a>';
XRegExp.matchChain(html, [
    {regex: /<a href="([^"]+)">/i, backref: 1},
    {regex: XRegExp('(?i)^https?://(?<domain>[^/?#]+)'), backref: 'domain'}
]); // -> ['xregexp.com', 'www.google.com']

// Merge strings and regexes into a single pattern, safely rewriting backreferences
XRegExp.union(['a+b*c', /(dogs)\1/, /(cats)\1/], 'i');
// -> /a\+b\*c|(dogs)\1|(cats)\2/i

These examples should give you the flavor of what's possible, but XRegExp has more syntax, flags, methods, options, and browser fixes that aren't shown here. You can even augment XRegExp's regular expression syntax with addons (see below) or write your own. See xregexp.com for more details.


In browsers, you can either load addons individually, or bundle all addons together with XRegExp by loading xregexp-all.js. XRegExp's npm package uses xregexp-all.js, which means that the addons are always available when XRegExp is installed on the server using npm.


In browsers, first include the Unicode Base script:

<script src="src/xregexp.js"></script>
<script src="src/addons/unicode/unicode-base.js"></script>

Then you can do this:

var unicodeWord = XRegExp('^\\p{L}+$');
unicodeWord.test('Русский'); // -> true
unicodeWord.test('日本語'); // -> true
unicodeWord.test('العربية'); // -> true

The base script adds \p{L} and its full name \p{Letter}, but other Unicode categories, scripts, blocks, and properties require addon packages. Try these next examples after additionally including unicode-scripts.js:

XRegExp('^\\p{Hiragana}+$').test('ひらがな'); // -> true
XRegExp('^[\\p{Latin}\\p{Common}]+$').test('Über Café.'); // -> true

By default, \p{…} and \P{…} support the Basic Multilingual Plane (i.e., code points up to U+FFFF). You can opt in to full 21-bit Unicode support (with code points up to U+10FFFF) on a per-regex basis by using flag A. In XRegExp, this is called astral mode. You can implicitly apply astral mode for all new regexes by running XRegExp.install('astral'). When in astral mode, \p{…} and \P{…} always match a full code point rather than a code unit, using surrogate pairs for code points above U+FFFF.

// Using flag A. The test string uses a surrogate pair to represent U+1F4A9
XRegExp('^\\pS$', 'A').test('\uD83D\uDCA9'); // -> true

// Implicit flag A
XRegExp('^\\pS$').test('\uD83D\uDCA9'); // -> true

Opting in to astral mode disables the use of \p{…} and \P{…} within character classes. In astral mode, use e.g. (?:\pL|\pM|[0-9_])+ instead of [\pL\pM0-9_]+.

XRegExp uses Unicode 6.2.0. Support for astral code points is new in XRegExp 3.0.0.


In browsers, first include the script:

<script src="src/xregexp.js"></script>
<script src="src/addons/build.js"></script>

You can then build regular expressions using named subpatterns, for readability and pattern reuse:

var time = XRegExp.build('(?x)^ {{hours}} ({{minutes}}) $', {
    hours: XRegExp.build('{{h12}} : | {{h24}}', {
        h12: /1[0-2]|0?[1-9]/,
        h24: /2[0-3]|[01][0-9]/
    minutes: /^[0-5][0-9]$/

time.test('10:59'); // -> true
XRegExp.exec('10:59', time).minutes; // -> '59'

Named subpatterns can be provided as strings or regex objects. A leading ^ and trailing unescaped $ are stripped from subpatterns if both are present, which allows embedding independently-useful anchored patterns. {{…}} tokens can be quantified as a single unit. Backreferences in the outer pattern and provided subpatterns are automatically renumbered to work correctly within the larger combined pattern. The syntax ({{name}}) works as shorthand for named capture via (?<name>{{name}}). Named subpatterns cannot be embedded within character classes.

See also: Creating Grammatical Regexes Using XRegExp.build.


In browsers, first include the script:

<script src="src/xregexp.js"></script>
<script src="src/addons/matchrecursive.js"></script>

You can then match recursive constructs using XRegExp pattern strings as left and right delimiters:

var str = '(t((e))s)t()(ing)';
XRegExp.matchRecursive(str, '\\(', '\\)', 'g');
// -> ['t((e))s', '', 'ing']

// Extended information mode with valueNames
str = 'Here is <div> <div>an</div></div> example';
XRegExp.matchRecursive(str, '<div\\s*>', '</div>', 'gi', {
    valueNames: ['between', 'left', 'match', 'right']
/* -> [
{name: 'between', value: 'Here is ',       start: 0,  end: 8},
{name: 'left',    value: '<div>',          start: 8,  end: 13},
{name: 'match',   value: ' <div>an</div>', start: 13, end: 27},
{name: 'right',   value: '</div>',         start: 27, end: 33},
{name: 'between', value: ' example',       start: 33, end: 41}
] */

// Omitting unneeded parts with null valueNames, and using escapeChar
str = '...{1}\\{{function(x,y){return y+x;}}';
XRegExp.matchRecursive(str, '{', '}', 'g', {
    valueNames: ['literal', null, 'value', null],
    escapeChar: '\\'
/* -> [
{name: 'literal', value: '...', start: 0, end: 3},
{name: 'value',   value: '1',   start: 4, end: 5},
{name: 'literal', value: '\\{', start: 6, end: 8},
{name: 'value',   value: 'function(x,y){return y+x;}', start: 9, end: 35}
] */

// Sticky mode via flag y
str = '<1><<<2>>><3>4<5>';
XRegExp.matchRecursive(str, '<', '>', 'gy');
// -> ['1', '<<2>>', '3']

XRegExp.matchRecursive throws an error if it sees an unbalanced delimiter in the target string.

Installation and usage

In browsers:

<script src="build/xregexp-min.js"></script>

Or, to bundle XRegExp with all of its addons:

<script src="build/xregexp-all-min.js"></script>

Using npm:

npm install xregexp

In Node.js:

var XRegExp = require('xregexp').XRegExp;

In an AMD loader like RequireJS:

require({paths: {xregexp: 'build/xregexp-all-min'}}, ['xregexp'], function(XRegExp) {



XRegExp and addons copyright 2007-2012 by Steven Levithan.

Tools: Unicode range generators by Mathias Bynens, and adapted from his unicode-data project. Source file concatenator by Bjarke Walling.

Tests: Uses Jasmine for unit tests, and Benchmark.js for performance tests.

Prior art: XRegExp.build inspired by Lea Verou's RegExp.create. XRegExp.union inspired by Ruby. XRegExp's syntax extensions and flags come from Perl, .NET, etc.

All code, including addons, tools, and tests, are released under the terms of the MIT License.

Fork me to show support, fix, and extend.