Perl and Emacs regular expressions compared

This page is intended as a memory-jogger about the different syntax of Perl and Emacs regular expressions. The Emacs regular expressions described here are the command-line regular expressions used in things like replace-regexp rather than the elisp regular expressions.

Perl regex	Emacs regex	Notes
(, )	$, $	Emacs uses backslashes in the style of "grep" for capturing parentheses, whereas Perl uses the "egrep" style of capturing parentheses without backslashes.
{, }	\{, \}	Emacs uses backslashes in the style of "grep" for curly brackets used for the number of matches, whereas Perl uses the "egrep" style without backslashes.
\s	\s-	\s matches whitespace in Perl. See What characters match a regular expression? for exactly what characters it matches. In Emacs, \s- matches the whitespace character class. The \s designates a character class and the minus sign makes it the "whitespace" class.
$1, $2, $3, ...	\1, \2, \3, ...	Emacs puts the results of capturing parentheses into \1, \2, etc. Perl uses a dollar sign, as in $1, $2, etc.
[, ]	[, ]	Emacs and Perl use the same characters, [ and ], to create character classes.
\d	[0-9]	Emacs does not have a way to match digits except by using posix character classes, as in `[[:digits:]]`.
\b	\b	Perl and Emacs use the same \b to match word boundaries.
\w	\w	Perl and Emacs use the same \w character to match word characters.

Regex Tool is a tool for showing what matches a regex inside Emacs. It works with both Emacs and Perl regular expressions. There is also something called "regexp-builder" in Emacs, but this actually uses a different form of regular expressions (the ones for Emacs Lisp) and so it is not very useful in practice.

Web links

Question about "comparison table for emacs regexp and perl compatible regular expression (PCRE)?" - stackoverflow.com

A little information on differences at a community website.
RegularExpression - Emacs Wiki

A very in-depth guide to Emacs regular expressions, which gets to the point much more rapidly than the meandering Emacs documentation.
Emacs Lisp for Perl Programmers

A table comparing Emacs Lisp and Perl has a little information about regular expressions.

Perl regex	Emacs regex	Notes
(, )	\(, \)	Emacs uses backslashes in the style of "grep" for capturing parentheses, whereas Perl uses the "egrep" style of capturing parentheses without backslashes.
{, }	\{, \}	Emacs uses backslashes in the style of "grep" for curly brackets used for the number of matches, whereas Perl uses the "egrep" style without backslashes.
\s	\s-	\s matches whitespace in Perl. See What characters match a regular expression? for exactly what characters it matches. In Emacs, \s- matches the whitespace character class. The \s designates a character class and the minus sign makes it the "whitespace" class.
$1, $2, $3, ...	\1, \2, \3, ...	Emacs puts the results of capturing parentheses into \1, \2, etc. Perl uses a dollar sign, as in $1, $2, etc.
[, ]	[, ]	Emacs and Perl use the same characters, [ and ], to create character classes.
\d	[0-9]	Emacs does not have a way to match digits except by using posix character classes, as in `[[:digits:]]`.
\b	\b	Perl and Emacs use the same \b to match word boundaries.
\w	\w	Perl and Emacs use the same \w character to match word characters.