[javascript] What characters are valid for JavaScript variable names?

Which characters can be used for naming a JavaScript variable?

I want to create a small "extension library" for my non-JavaScript users here at work (who all seem to be squeamish when it comes to the language). I love how jQuery and Prototype have both use the $ dollar sign, and since I use jQuery, I'm looking for another nice single-character symbol to use.

I realize that I could just test out a number of characters, but I'm hoping to narrow down my list of characters to start with (in consideration of future integration with another popular library, perhaps).

in case regular expressions is not a must, wouldn't it be better to just ask the browser to decide using eval ?

function isValidVarName( name ) {
    try {
        // Update, previoulsy it was
        // eval('(function() { var ' + name + '; })()');
        Function('var ' + name);
    } catch( e ) {
        return false;
    }
    return true;
}

isValidVarName('my_var');     // true
isValidVarName('1');          // false

From the ECMAScript specification in section 7.6 Identifier Names and Identifiers, a valid identifier is defined as:

Identifier :: 
    IdentifierName but not ReservedWord

IdentifierName :: 
    IdentifierStart 
    IdentifierName IdentifierPart 

IdentifierStart :: 
    UnicodeLetter 
    $ 
    _ 
    \ UnicodeEscapeSequence 

IdentifierPart :: 
    IdentifierStart 
    UnicodeCombiningMark 
    UnicodeDigit 
    UnicodeConnectorPunctuation 
    \ UnicodeEscapeSequence 

UnicodeLetter 
    any character in the Unicode categories “Uppercase letter (Lu)”, “Lowercase letter (Ll)”, “Titlecase letter (Lt)”, 
    “Modifier letter (Lm)”, “Other letter (Lo)”, or “Letter number (Nl)”. 

UnicodeCombiningMark 
    any character in the Unicode categories “Non-spacing mark (Mn)” or “Combining spacing mark (Mc)” 

UnicodeDigit 
    any character in the Unicode category “Decimal number (Nd)” 

UnicodeConnectorPunctuation 
    any character in the Unicode category “Connector punctuation (Pc)” 

UnicodeEscapeSequence 
    see 7.8.4. 

HexDigit :: one of 
    0 1 2 3 4 5 6 7 8 9 a b c d e f A B C D E F

which creates a lot of opportunities for naming variables and also in golfing. Let's try some examples.

A valid identifier could start with either a UnicodeLetter, $, _, or \ UnicodeEscapeSequence. A unicode letter is any character from these categories (see all categories):

• Uppercase letter (Lu)
• Lowercase letter (Ll)
• Titlecase letter (Lt)
• Modifier letter (Lm)
• Other letter (Lo)
• Letter number (Nl)

This alone accounts for some crazy possibilities - working examples. If it doesn't work in all browsers, then call it a bug, cause it should.

var ? = "something";
var HELLO = "hello";
var ???? = "less than? wtf";
var ????????????? = "javascript"; // ok that's JavaScript in hindi
var KingGeorge? = "Roman numerals, awesome!";

Javascript variables can have letters, digits, dollar signs ($) and underscores (_). They can't start with digits.

Usually libraries use $ and _ as shortcuts for functions that you'll be using everywhere. Although the names $ or _ aren't meaningful, they're useful for their shortness and since you'll be using the function everywhere you're expected to know what they mean.

If your library doesn't consist on getting a single function being used everywhere, I'd recommend that you use more meaningful names as those will help you and others understand what your code is doing without necessarily compromising the source code niceness.

You could for instance take a look at the awesome DateJS library and at the syntatic sugar it allows without the need of any symbol or short-named variables.

You should first get your code to be practical, and only after try making it pretty.

The accepted answer would rule out a lot of valid identifiers, as far as I can see. Here is a regular expression that I put together which should follow the spec (see chapter 7.6 on identifiers). Created it using RegexBuddy and you can find an export of the explanation at http://samples.geekality.net/js-identifiers.

^[$_\p{L}][$_\p{L}\p{Mn}\p{Mc}\p{Nd}\p{Pc}\u200C\u200D]*+$

In addition, the name cannot be one of the following reserved words.

break, do, instanceof, typeof, case, else, new, var, catch, finally, return, void, continue, for, switch, while, debugger, function, this, with, default, if, throw, delete, in, try, class, enum, extends, super, const, export, import, implements, let, private, public, yield, interface, package, protected, static, null, true, false

Here is one quick suggestion for creating variable names. If you want the variable not to conflict when being used in FireFox, do not use the variable name "_content" as this variable name is already being used by the browser. I found this out the hard way and had to change all of the places I used the variable "_content" in a large JavaScript application.

Wrote a glitch workspace that iterates over all the codepoints and emit the character if eval('var ' + String.fromCodePoint(#) + ' = 1') works.

It just keeps going, and going, and going....

Before JavaScript 1.5: ^[a-zA-Z_$][0-9a-zA-Z_$]*$

In English: It must start with a dollar sign, underscore or one of letters in the 26-character alphabet, upper or lower case. Subsequent characters (if any) can be one of any of those or a decimal digit.

JavaScript 1.5 and later * : ^[\p{L}\p{Nl}$_][\p{L}\p{Nl}$\p{Mn}\p{Mc}\p{Nd}\p{Pc}]*$

This is more difficult to express in English, but it is conceptually similar to the older syntax with the addition that the letters and digits can be from any language. After the first character, there are also allowed additional underscore-like characters (collectively called “connectors”) and additional character combining marks (“modifiers”). (Other currency symbols are not included in this extended set.)

JavaScript 1.5 and later also allows Unicode escape sequences, provided that the result is a character that would be allowed in the above regular expression.

Identifiers also must not be a current reserved word or one that is considered for future use.

There is no practical limit to the length of an identifier. (Browsers vary, but you’ll safely have 1000 characters and probably several more orders of magnitude than that.)

Links to the character categories:

• Letters: Lu, Ll, Lt, Lm, Lo, Nl
  (combined in the regex above as “L”)
• Combining marks (“modifiers”): Mn, Mc
• Digits: Nd
• Connectors: Pc

*n.b. This Perl regex is intended to describe the syntax only — it won’t work in JavaScript, which doesn’t (yet) include support for Unicode Properties. (There are some third-party packages that claim to add such support.)

I've taken Anas Nakawa's idea and improved it. First of all, there is no reason to actually run the function being declared. We want to know whether it parses correctly, not whether the code works. Second, a literal object is a better context for our purpose than var XXX as it's harder to break out of.

    function isValidVarName( name ) {
    try {
        return name.indexOf('}') === -1 && eval('(function() { a = {' + name + ':1}; a.' + name + '; var ' + name + '; }); true');
    } catch( e ) {
        return false;
    }
    return true;
}

// so we can see the test code
var _eval = eval;
window.eval = function(s) {
    console.log(s);
    return _eval(s);
}

console.log(isValidVarName('name'));
console.log(isValidVarName('$name'));
console.log(isValidVarName('not a name'));
console.log(isValidVarName('a:2,b'));
console.log(isValidVarName('"a string"'));

console.log(isValidVarName('xss = alert("I\'m in your vars executin mah scrip\'s");;;;;'));
console.log(isValidVarName('_;;;'));
console.log(isValidVarName('_=location="#!?"'));

console.log(isValidVarName('?'));
console.log(isValidVarName('HELLO'));
console.log(isValidVarName('????'));
console.log(isValidVarName('?????????????'));
console.log(isValidVarName('KingGeorge?'));
console.log(isValidVarName('}; }); alert("I\'m in your vars executin\' mah scripts"); true; // yeah, super valid'));
console.log(isValidVarName('if'));

Basically, in regular expression form: [a-zA-Z_$][0-9a-zA-Z_$]*. In other words, the first character can be a letter or _ or $, and the other characters can be letters or _ or $ or numbers.

Note: While other answers have pointed out that you can use Unicode characters in JavaScript identifiers, the actual question was "What characters should I use for the name of an extension library like jQuery?" This is an answer to that question. You can use Unicode characters in identifiers, but don't do it. Encodings get screwed up all the time. Keep your public identifiers in the 32-126 ASCII range where it's safe.

Actually, ECMAScript says on page 15: That an identifier may start with a $, an underscore or a UnicodeLetter, and then it goes on (just below that) to specify that a UnicodeLetter can be any character from the unicode catagories, Lo, Ll, Lu, Lt, Lm and Nl. And when you look up those catagories you will see that this opens up a lot more possibilities than just latin letters. Just search for "unicode catagories" in google and you can find them.

Javascript Variables

You can start a variable with any letter, $, or _ character. As long as it doesn't start with a number, you can include numbers as well.

Start: [a-z], $, _

Contain: [a-z], [0-9], $, _

jQuery

You can use _ for your library so that it will stand side-by-side with jQuery. However, there is a configuration you can set so that jQuery will not use $. It will instead use jQuery. To do this, simply set:

jQuery.noConflict();

This page explains how to do this.