JavaScript/Regular expressions

Overview

JavaScript implements regular expressions (regex for short) when searching for matches within a string. As with other scripting languages, this allows searching beyond a simple letter-by-letter match, and can even be used to parse strings in a certain format.

Unlike strings, regular expressions are delimited by the slash (/) character, and may have some options appended.

Regular expressions most commonly appear in conjunction with the string.match() and string.replace() methods.

At a glance, by example:

strArray = "Hello world!".match(/world/); // Singleton array; note the slashes
strArray = "Hello!".match(/l/g); // Matched strings are returned in a string array
"abc".match(/a(b)c/)[1] === "b" // Matched subgroup is the 2nd item (index 1)
str1 = "Hey there".replace(/Hey/g, "Hello");
str2 = "N/A".replace(/\//g, ","); // Slash is escaped with \
str3 = "Hello".replace(/l/g, "m").replace(/H/g, "L").replace(/o/g, "a"); // Pile
if (str3.match(/emma/)) { console.log("Yes"); }
if (str3.match("emma")) { console.log("Yes"); } // Quotes work as well
"abbc".replace(/(.)\1/g, "$1") === "abc" // Backreference

Compatibility

JavaScript's set of regular expressions follows the extended set. While copying a Regex pattern from JavaScript to another location may work as expected, some older programs may not function as expected.

In the search term, \1 is used to back reference a matched group, as in other implementations.
In the replacement string, $1 is substituted with a matched group in the search, instead of \1.
- Example: "abbc".replace(/(.)\1/g, "$1") => "abc"
| is magic, \| is literal
( is magic, \( is literal
The syntaxes (?=...), (?!...), (?<=...), and (?<!...) are not available.

Examples

Matching
- string = "Hello world!".match(/world/);
- stringArray = "Hello world!".match(/l/g); // Matched strings are returned in a string array
- "abc".match(/a(b)c/)[1] => "b" // Matched subgroup is the second member (having the index "1") of the resulting array
Replacement
- string = string.replace(/expression without quotation marks/g, "replacement");
- string = string.replace(/escape the slash in this\/way/g, "replacement");
- string = string.replace( ... ).replace ( ... ). replace( ... );
Test
- if (string.match(/regexp without quotation marks/)) {

Modifiers

Single-letter modifiers:

g	Global. The list of matches is returned in an array.
i	Case-insensitive search
m	Multiline. If the operand string has multiple lines, ^ and $ match the beginning and end of each line within the string, instead of matching the beginning and end of the whole string only: "a\nb\nc".replace(/^b$/g,"d") === "a\nb\nc" "a\nb\nc".replace(/^b$/gm,"d") === "a\nd\nc"

Operators

Operator	Effect
\b	Matches boundary of a word.
\w	Matches an alphanumeric character, including "_".
\W	Negation of \w.
\s	Matches a whitespace character (space, tab, newline, formfeed)
\S	Negation of \s.
\d	Matches a digit.
\D	Negation of \d.

Function call

For complex operations, a function can process the matched substrings. In the following code, we are capitalizing all the words. It can't be done by a simple replacement, as each letter to capitalize is a different character:

var capitalize = function(matchobj) {
  var group1 = matchobj.replace(/^(\W)[a-zA-Z]+$/g, "$1");
  var group2 = matchobj.replace(/^\W([a-zA-Z])[a-zA-Z]+$/g, "$1");
  var group3 = matchobj.replace(/^\W[a-zA-Z]([a-zA-Z]+)$/g, "$1");
  return group1 + group2.toUpperCase() + group3;
};

var classicText = "To be or not to be?";

var changedClassicText = classicText.replace(/\W[a-zA-Z]+/g, capitalize);

console.log(changedClassicText==="To Be Or Not To Be?");

The function is called for each substring. Here is the signature of the function:

function (''<matchedSubstring>[, <capture1>, ...<captureN>, <indexInText>, <entireText>]'') {
 ...
 return ''<stringThatWillReplaceInText>'';
}

The first parameter is the substring that matches the pattern.
The next parameters are the captures in the substrings. There are as many parameters as there are captures.
The next parameter is the index of the beginning of the substring starting from the beginning of the text.
The last parameter is a remainder of the entire text.
The return value will be put in the text instead of the matching substring.