Regular expressions - JavaScript | MDN

MDN Web Docs

Description

Regular expressions

are an important concept in formal language theory. They are a way to describe a possibly infinite set of character strings (called a language). A regular expression, at its core, needs the following features:

A set of characters that can be used in the language, called the alphabet.

Concatenation: ab means "the character a followed by the character b".

Union: a|b means "either a or b".

Kleene star: a* means "zero or more a characters".

Assuming a finite alphabet (such as the 26 letters of the English alphabet, or the entire Unicode character set), all regular languages can be generated by the features above. Of course, many patterns are very tedious to express this way (such as "10 digits" or "a character that's not a space"), so JavaScript regular expressions include many shorthands, introduced below.

Note: JavaScript regular expressions are in fact not regular, due to the existence of

backreferences

(regular expressions must have finite states). However, they are still a very useful feature.

Creating regular expressions

A regular expression is typically created as a literal by enclosing a pattern in forward slashes (/):

js

const regex1 = /ab+c/g; Regular expressions can also be created with the

RegExp()

constructor:

js

const regex2 = new RegExp("ab+c", "g"); They have no runtime differences, although they may have implications on performance, static analyzability, and authoring ergonomic issues with escaping characters. For more information, see the

RegExp

reference.

Regex flags

Flags are special parameters that can change the way a regular expression is interpreted or the way it interacts with the input text. Each flag corresponds to one accessor property on the RegExp object.

FlagDescriptionCorresponding propertydGenerate indices for substring matches.

hasIndices

gGlobal search.

global

iCase-insensitive search.

ignoreCase

mMakes ^ and $ match the start and end of each line instead of those of the entire string.

multiline

sAllows . to match newline characters.

dotAll

u"Unicode"; treat a pattern as a sequence of Unicode code points.

unicode

vAn upgrade to the u mode with more Unicode features.

unicodeSets

yPerform a "sticky" search that matches starting at the current position in the target string.

sticky

The i, m, and s flags can be enabled or disabled for specific parts of a regex using the

modifier

syntax.

The sections below list all available regex syntaxes, grouped by their syntactic nature.

Assertions

Assertions are constructs that test whether the string meets a certain condition at the specified position, but not consume characters. Assertions cannot be

quantified

.

Buffer boundary assertion: \A, \z, \Z

Asserts that the current position in the string is strictly at the start or end of the entire string (\Z also allows a trailing newline), regardless of the presence of the m flag.

Input boundary assertion: ^, $

Asserts that the current position is the start or end of input, or start or end of a line if the m flag is set.

Lookahead assertion: (?=...), (?!...)

Asserts that the current position is followed or not followed by a certain pattern.

Lookbehind assertion: (?<=...), (?<!...)

Asserts that the current position is preceded or not preceded by a certain pattern.

Word boundary assertion: \b, \B

Asserts that the current position is a word boundary.

Atoms

Atoms are the most basic units of a regular expression. Each atom consumes one or more characters in the string, and either fails the match or allows the pattern to continue matching with the next atom.

Backreference: \1, \2

Matches a previously matched subpattern captured with a capturing group.

Capturing group: (...)

Matches a subpattern and remembers information about the match.

Character class: [...], [^...]

Matches any character in or not in a set of characters. When the

v

flag is enabled, it can also be used to match finite-length strings.

Character class escape: \d, \D, \w, \W, \s, \S

Matches any character in or not in a predefined set of characters.

Character escape: \n, \u{...}

Matches a character that may not be able to be conveniently represented in its literal form.

Literal character: a, b

Matches a specific character.

Modifier: (?ims-ims:...)

Overrides flag settings in a specific part of a regular expression.

Named backreference: \k<name>

Matches a previously matched subpattern captured with a named capturing group.

Named capturing group: (?<name>...)

Matches a subpattern and remembers information about the match. The group can later be identified by a custom name instead of by its index in the pattern.

Non-capturing group: (?:...)

Matches a subpattern without remembering information about the match.

Unicode character class escape: \p{...}, \P{...}

Matches a set of characters specified by a Unicode property. When the

v

flag is enabled, it can also be used to match finite-length strings.

Wildcard: .

Matches any character except line terminators, unless the s flag is set.

Other features

These features do not specify any pattern themselves, but are used to compose patterns.

Disjunction: |

Matches any of a set of alternatives separated by the | character.

Quantifier: *, +, ?, {n}, {n,}, {n,m}

Matches an atom a certain number of times.

Escape sequences

Escape sequences in regexes refer to any kind of syntax formed by \ followed by one or more characters. They may serve very different purposes depending on what follow \. Below is a list of all valid "escape sequences":

Escape sequenceFollowed byMeaning\ANone

Buffer boundary assertion

\BNone

Non-word-boundary assertion

\DNone

Character class escape

representing non-digit characters\P{, a Unicode property and/or value, then }

Unicode character class escape

representing characters without the specified Unicode property\SNone

Character class escape

representing non-white-space characters\WNone

Character class escape

representing non-word characters\ZNone

Buffer boundary assertion

\bNone

Word boundary assertion

; inside

character classes

, represents U+0008 (BACKSPACE)\cA letter from A to Z or a to zA

character escape

representing the control character with value equal to the letter's character value modulo 32\dNone

Character class escape

representing digit characters (0 to 9)\fNone

Character escape

representing U+000C (FORM FEED)\k<, an identifier, then >A

named backreference

\nNone

Character escape

representing U+000A (LINE FEED)\p{, a Unicode property and/or value, then }

Unicode character class escape

representing characters with the specified Unicode property\q{, a string, then a }Only valid inside

v-mode character classes

; represents the string to be matched literally\rNone

Character escape

representing U+000D (CARRIAGE RETURN)\sNone

Character class escape

representing whitespace characters\tNone

Character escape

representing U+0009 (CHARACTER TABULATION)\u4 hexadecimal digits; or {, 1 to 6 hexadecimal digits, then }

Character escape

representing the character with the given code point\vNone

Character escape

representing U+000B (LINE TABULATION)\wNone

Character class escape

representing word characters (A to Z, a to z, 0 to 9, _)\x2 hexadecimal digits

Character escape

representing the character with the given value\zNone

Buffer boundary assertion

\0None

Character escape

representing U+0000 (NULL)\ followed by 0 and another digit becomes a

legacy octal escape sequence

, which is forbidden in

Unicode-aware mode

. \ followed by any other digit sequence becomes a

backreference

.

In addition, \ can be followed by some non-letter-or-digit characters, in which case the escape sequence is always a

character escape

representing the escaped character itself:

\$, \(, \), \*, \+, \., \/, \?, \[, \\, \], \^, \{, \|, \}: valid everywhere

\-: only valid inside

character classes

\!, \#, \%, \&, \,, \:, \;, \<, \=, \>, \@, \`, \~: only valid inside

v-mode character classes

The other

ASCII

characters, namely space character, ", ', _, and any letter character not mentioned above, are not valid escape sequences. In

Unicode-unaware mode

, escape sequences that are not one of the above become identity escapes: they represent the character that follows the backslash. For example, \a represents the character a. This behavior limits the ability to introduce new escape sequences without causing backward compatibility issues, and is therefore forbidden in Unicode-aware mode.

Specifications

Specification

ECMAScript® 2027 Language Specification# prod-PatternCharacter

ECMAScript® 2027 Language Specification# prod-CharacterClass

ECMAScript® 2027 Language Specification# prod-RegularExpressionModifiers

ECMAScript® 2027 Language Specification# prod-Disjunction

ECMAScript® 2027 Language Specification# prod-CharacterClassEscape

ECMAScript® 2027 Language Specification# prod-Atom

ECMAScript® 2027 Language Specification# prod-RegExpUnicodeEscapeSequence

ECMAScript® 2027 Language Specification# prod-CharacterEscape

ECMAScript® 2027 Language Specification# prod-DecimalEscape

ECMAScript® 2027 Language Specification# prod-AtomEscape

ECMAScript® 2027 Language Specification# prod-Assertion

ECMAScript® 2027 Language Specification# prod-Quantifier

ECMAScript® 2027 Language Specification# sec-patterns-static-semantics-early-errors

Regular Expression Buffer Boundaries for ECMAScript# sec-patterns

Browser compatibility

See also

Regular expressions

guide

RegExp