PCREPATTERN(3)
💖 NAME 💖
okay so this is PCREPATTERN(3), the manual page for PCRE — Perl-compatible regular expressions — and bestie, we are about to get so into it. 💅
🌸 PCRE REGULAR EXPRESSION DETAILS 🌸
The syntax and semantics of the regular expressions that are supported by PCRE are described in detail below. There is a quick-reference syntax summary in the pcresyntax page. PCRE tries to match Perl syntax and semantics as closely as it can. PCRE also supports some alternative regular expression syntax (which does not conflict with the Perl syntax) in order to provide some compatibility with regular expressions in Python, .NET, and Oniguruma.
Perl's regular expressions are described in its own documentation, and regular expressions in general are covered in a number of books, some of which have copious examples. Jeffrey Friedl's "Mastering Regular Expressions", published by O'Reilly, covers regular expressions in great detail. This description of PCRE's regular expressions is intended as reference material.
This document discusses the patterns that are supported by PCRE when one its main matching functions, pcre_exec() (8-bit) or pcre[16|32]_exec() (16- or 32-bit), is used. PCRE also has alternative matching functions, pcre_dfa_exec() and pcre[16|32_dfa_exec(), which match using a different algorithm that is not Perl-compatible. Some of the features discussed below are not available when DFA matching is used. The advantages and disadvantages of the alternative functions, and how they differ from the normal functions, are discussed in the pcrematching page.
✨ SPECIAL START-OF-PATTERN ITEMS ✨
A number of options that can be passed to pcre_compile() can also be set by special items at the start of a pattern. These are not Perl-compatible, but are provided to make these options accessible to pattern writers who are not able to change the program that processes the pattern. Any number of these items may appear, but they must all be together right at the start of the pattern string, and the letters must be in upper case.
🦋 UTF support 🦋
The original operation of PCRE was on strings of one-byte characters. However, there is now also support for UTF-8 strings in the original library, an extra library that supports 16-bit and UTF-16 character strings, and a third library that supports 32-bit and UTF-32 character strings. To use these features, PCRE must be built to include appropriate support. When using UTF strings you must either call the compiling function with the PCRE_UTF8, PCRE_UTF16, or PCRE_UTF32 option, or the pattern must start with one of these special sequences:
(*UTF8)
(*UTF16)
(*UTF32)
(*UTF)(*UTF) is a generic sequence that can be used with any of the libraries. Starting a pattern with such a sequence is equivalent to setting the relevant option. How setting a UTF mode affects pattern matching is mentioned in several places below. There is also a summary of features in the pcreunicode page.
Some applications that allow their users to supply patterns may wish to restrict them to non-UTF data for security reasons. If the PCRE_NEVER_UTF option is set at compile time, (*UTF) etc. are not allowed, and their appearance causes an error.
🧚 Unicode property support 🧚
Another special sequence that may appear at the start of a pattern is (*UCP). This has the same effect as setting the PCRE_UCP option: it causes sequences such as \d and \w to use Unicode properties to determine character types, instead of recognizing only characters with codes less than 128 via a lookup table.
💅 Disabling auto-possessification 💅
If a pattern starts with (*NO_AUTO_POSSESS), it has the same effect as setting the PCRE_NO_AUTO_POSSESS option at compile time. This stops PCRE from making quantifiers possessive when what follows cannot match the repeated item. For example, by default a+b is treated as a++b. For more details, see the pcreapi documentation.
🎀 Disabling start-up optimizations 🎀
If a pattern starts with (*NO_START_OPT), it has the same effect as setting the PCRE_NO_START_OPTIMIZE option either at compile or matching time. This disables several optimizations for quickly reaching "no match" results. For more details, see the pcreapi documentation.
🌈 Newline conventions 🌈
PCRE supports five different conventions for indicating line breaks in strings: a single CR (carriage return) character, a single LF (linefeed) character, the two-character sequence CRLF, any of the three preceding, or any Unicode newline sequence. The pcreapi page has further discussion about newlines, and shows how to set the newline convention in the options arguments for the compiling and matching functions.
It is also possible to specify a newline convention by starting a pattern string with one of the following five sequences:
(*CR) carriage return
(*LF) linefeed
(*CRLF) carriage return, followed by linefeed
(*ANYCRLF) any of the three above
(*ANY) all Unicode newline sequencesThese override the default and the options given to the compiling function. For example, on a Unix system where LF is the default newline sequence, the pattern
(*CR)a.bchanges the convention to CR. That pattern matches "a\nb" because LF is no longer a newline. If more than one of these settings is present, the last one is used.
The newline convention affects where the circumflex and dollar assertions are true. It also affects the interpretation of the dot metacharacter when PCRE_DOTALL is not set, and the behaviour of \N. However, it does not affect what the \R escape sequence matches. By default, this is any Unicode newline sequence, for Perl compatibility. However, this can be changed; see the description of \R in the section entitled "Newline sequences" below. A change of \R setting can be combined with a change of newline convention.
⭐ Setting match and recursion limits ⭐
The caller of pcre_exec() can set a limit on the number of times the internal match() function is called and on the maximum depth of recursive calls. These facilities are provided to catch runaway matches that are provoked by patterns with huge matching trees (a typical example is a pattern with nested unlimited repeats) and to avoid running out of system stack by too much recursion. When one of these limits is reached, pcre_exec() gives an error return. The limits can also be set by items at the start of the pattern of the form
(*LIMIT_MATCH=d)
(*LIMIT_RECURSION=d)where d is any number of decimal digits. However, the value of the setting must be less than the value set (or defaulted) by the caller of pcre_exec() for it to have any effect. In other words, the pattern writer can lower the limits set by the programmer, but not raise them. If there is more than one setting of one of these limits, the lower value is used.
🔥 EBCDIC CHARACTER CODES 🔥
PCRE can be compiled to run in an environment that uses EBCDIC as its character code rather than ASCII or Unicode (typically a mainframe system). In the sections below, character code values are ASCII or Unicode; in an EBCDIC environment these characters may have different code values, and there are no code points greater than 255.
👑 CHARACTERS AND METACHARACTERS 👑
A regular expression is a pattern that is matched against a subject string from left to right. Most characters stand for themselves in a pattern, and match the corresponding characters in the subject. As a trivial example, the pattern
The quick brown foxmatches a portion of a subject string that is identical to itself. When caseless matching is specified (the PCRE_CASELESS option), letters are matched independently of case. In a UTF mode, PCRE always understands the concept of case for characters whose values are less than 128, so caseless matching is always possible. For characters with higher values, the concept of case is supported if PCRE is compiled with Unicode property support, but not otherwise. If you want to use caseless matching for characters 128 and above, you must ensure that PCRE is compiled with Unicode property support as well as with UTF support.
The power of regular expressions comes from the ability to include alternatives and repetitions in the pattern. These are encoded in the pattern by the use of metacharacters, which do not stand for themselves but instead are interpreted in some special way.
There are two different sets of metacharacters: those that are recognized anywhere in the pattern except within square brackets, and those that are recognized within square brackets. Outside square brackets, the metacharacters are as follows:
\ general escape character with several uses
^ assert start of string (or line, in multiline mode)
$ assert end of string (or line, in multiline mode)
. match any character except newline (by default)
[ start character class definition
| start of alternative branch
( start subpattern
) end subpattern
? extends the meaning of (
also 0 or 1 quantifier
also quantifier minimizer
* 0 or more quantifier
+ 1 or more quantifier
also "possessive quantifier"
{ start min/max quantifierPart of a pattern that is in square brackets is called a "character class". In a character class the only metacharacters are:
\ general escape character
^ negate the class, but only if the first character
- indicates character range
[ POSIX character class (only if followed by POSIX
syntax)
] terminates the character classThe following sections describe the use of each of the metacharacters.
💜 BACKSLASH 💜
The backslash character has several uses. Firstly, if it is followed by a character that is not a number or a letter, it takes away any special meaning that character may have. This use of backslash as an escape character applies both inside and outside character classes.
For example, if you want to match a * character, you write \* in the pattern. This escaping action applies whether or not the following character would otherwise be interpreted as a metacharacter, so it is always safe to precede a non-alphanumeric with backslash to specify that it stands for itself. In particular, if you want to match a backslash, you write \\.
In a UTF mode, only ASCII numbers and letters have any special meaning after a backslash. All other characters (in particular, those whose codepoints are greater than 127) are treated as literals.
okay so bestie, if your pattern gets compiled with the PCRE_EXTENDED option, most of the white space in it (not the white space inside a character class, tho — that stays) gets yeeted into the void. And anything from a # that's outside a character class all the way to the next newline, newline included? Also gone. But you can sneak a white space or a # back in as an actual pattern character with an escaping backslash. 💅
If you wanna strip the special meaning off a whole run of characters, just wrap them between \Q and \E. Here's the thing tho — this is different from Perl: in PCRE, $ and @ are handled as literals inside a \Q...\E sequence, whereas in Perl, $ and @ cause variable interpolation. Peep these examples:
Pattern PCRE matches Perl matches
\Qabc$xyz\E abc$xyz abc followed by the
contents of $xyz
\Qabc\$xyz\E abc\$xyz abc\$xyz
\Qabc\E\$\Qxyz\E abc$xyz abc$xyzThe \Q...\E sequence gets recognized both inside and outside character classes. An isolated \E that isn't preceded by \Q? Ignored, no cap. But if \Q isn't followed by \E later in the pattern, the literal interpretation just keeps going to the end of the pattern (that is, \E is assumed at the end). ⚠️ And if that isolated \Q is inside a character class, this causes an error, because the character class is not terminated.
✨ Non-printing characters ✨
A second use of backslash is giving you a way to encode non-printing characters in patterns in a way that's visible. There's genuinely no restriction on the appearance of non-printing characters, apart from the binary zero that terminates a pattern, but when a pattern is being prepped by text editing, it's often way easier to use one of these escape sequences than the binary character it represents. In an ASCII or Unicode environment, these escapes go like:
\a alarm, that is, the BEL character (hex 07)
\cx "control-x", where x is any ASCII character
\e escape (hex 1B)
\f form feed (hex 0C)
\n linefeed (hex 0A)
\r carriage return (hex 0D)
\t tab (hex 09)
\0dd character with octal code 0dd
\ddd character with octal code ddd, or back reference
\o{ddd..} character with octal code ddd..
\xhh character with hex code hh
\x{hhh..} character with hex code hhh.. (non-JavaScript mode)
\uhhhh character with hex code hhhh (JavaScript mode only)The precise effect of \cx on ASCII characters is literally this: if x is a lower case letter, it gets converted to upper case. Then bit 6 of the character (hex 40) gets inverted. So \cA to \cZ become hex 01 to hex 1A (A is 41, Z is 5A), but \c{ becomes hex 3B ({ is 7B), and \c; becomes hex 7B (; is 3B). If the data item (byte or 16-bit value) following \c has a value greater than 127, a compile-time error occurs. That locks out non-ASCII characters in all modes. 💀
When PCRE is compiled in EBCDIC mode, \a, \e, \f, \n, \r, and \t generate the appropriate EBCDIC code values. The \c escape is processed as specified for Perl in the perlebcdic document. The only characters allowed after \c are A-Z, a-z, or one of @, [, \, ], ^, _, or ?. Any other character provokes a compile-time error. The sequence \c@ encodes character code 0; after \c the letters (in either case) encode characters 1-26 (hex 01 to hex 1A); [, \, ], ^, and _ encode characters 27-31 (hex 1B to hex 1F), and \c? becomes either 255 (hex FF) or 95 (hex 5F).
So, apart from \c?, these escapes generate the same character code values as they do in an ASCII environment, though the meanings of the values mostly differ. For example, \cG always generates code value 7, which is BEL in ASCII but DEL in EBCDIC. wild.
The sequence \c? generates DEL (127, hex 7F) in an ASCII environment, but because 127 is not a control character in EBCDIC, Perl makes it generate the APC character. Unfortunately, there are several variants of EBCDIC. In most of them the APC character has the value 255 (hex FF), but in the one Perl calls POSIX-BC its value is 95 (hex 5F). If certain other characters have POSIX-BC values, PCRE makes \c? generate 95; otherwise it generates 255.
After \0 up to two further octal digits are read. If there are fewer than two digits, just those that are present are used. So the sequence \0\x\015 specifies two binary zeros followed by a CR character (code value 13). Make sure you supply two digits after the initial zero if the pattern character that follows is itself an octal digit. 🙈
The escape \o must be followed by a sequence of octal digits, enclosed in braces. An error occurs if this is not the case. This escape is a recent addition to Perl; it provides a way of specifying character code points as octal numbers greater than 0777, and it also lets octal numbers and back references be unambiguously specified.
For greater clarity and unambiguity, it's best to avoid following \ by a digit greater than zero. Instead, use \o{} or \x{} to specify character numbers, and \g{} to specify back references. The following paragraphs describe the old, ambiguous syntax. 👑
The handling of a backslash followed by a digit other than 0 is complicated, and Perl has changed in recent releases, causing PCRE also to change. Outside a character class, PCRE reads the digit and any following digits as a decimal number. If the number is less than 8, or if there have been at least that many previous capturing left parentheses in the expression, the entire sequence is taken as a back reference. A description of how this works is given later, following the discussion of parenthesized subpatterns.
Inside a character class, or if the decimal number following \ is greater than 7 and there have not been that many capturing subpatterns, PCRE handles \8 and \9 as the literal characters "8" and "9", and otherwise re-reads up to three octal digits following the backslash, using them to generate a data character. Any subsequent digits stand for themselves. For example:
\040 is another way of writing an ASCII space
\40 is the same, provided there are fewer than 40
previous capturing subpatterns
\7 is always a back reference
\11 might be a back reference, or another way of
writing a tab
\011 is always a tab
\0113 is a tab followed by the character "3"
\113 might be a back reference, otherwise the
character with octal code 113
\377 might be a back reference, otherwise
the value 255 (decimal)
\81 is either a back reference, or the two
characters "8" and "1"Note that octal values of 100 or greater that are specified using this syntax must not be introduced by a leading zero, because no more than three octal digits are ever read. fr fr.
By default, after \x that is not followed by {, from zero to two hexadecimal digits are read (letters can be in upper or lower case). Any number of hexadecimal digits may appear between \x{ and }. If a character other than a hexadecimal digit appears between \x{ and }, or if there is no terminating }, an error occurs.
If the PCRE_JAVASCRIPT_COMPAT option is set, the interpretation of \x is as just described only when it is followed by two hexadecimal digits. Otherwise, it matches a literal "x" character. In JavaScript mode, support for code points greater than 256 is provided by \u, which must be followed by four hexadecimal digits; otherwise it matches a literal "u" character.
Characters whose value is less than 256 can be defined by either of the two syntaxes for \x (or by \u in JavaScript mode). There is no difference in the way they are handled. For example, \xdc is exactly the same as \x{dc} (or \u00dc in JavaScript mode).
🦋 Constraints on character values 🦋
Characters that are specified using octal or hexadecimal numbers are limited to certain values, as follows:
8-bit non-UTF mode less than 0x100
8-bit UTF-8 mode less than 0x10ffff and a valid codepoint
16-bit non-UTF mode less than 0x10000
16-bit UTF-16 mode less than 0x10ffff and a valid codepoint
32-bit non-UTF mode less than 0x100000000
32-bit UTF-32 mode less than 0x10ffff and a valid codepointInvalid Unicode codepoints are the range 0xd800 to 0xdfff (the so-called "surrogate" codepoints), and 0xffef.
💜 Escape sequences in character classes 💜
All the sequences that define a single character value can be used both inside and outside character classes. In addition, inside a character class, \b is interpreted as the backspace character (hex 08).
\N is not allowed in a character class. \B, \R, and \X are not special inside a character class. Like other unrecognized escape sequences, they are treated as the literal characters "B", "R", and "X" by default, but cause an error if the PCRE_EXTRA option is set. Outside a character class, these sequences have different meanings.
🌈 Unsupported escape sequences 🌈
In Perl, the sequences \l, \L, \u, and \U are recognized by its string handler and used to modify the case of following characters. By default, PCRE does not support these escape sequences. However, if the PCRE_JAVASCRIPT_COMPAT option is set, \U matches a "U" character, and \u can be used to define a character by code point, as described in the previous section.
⭐ Absolute and relative back references ⭐
The sequence \g followed by an unsigned or a negative number, optionally enclosed in braces, is an absolute or relative back reference. A named back reference can be coded as \g{name}. Back references are discussed later, following the discussion of parenthesized subpatterns.
👑 Absolute and relative subroutine calls 👑
For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or a number enclosed either in angle brackets or single quotes, is an alternative syntax for referencing a subpattern as a "subroutine". Details are discussed later. Note that \g{...} (Perl syntax) and \g<...> (Oniguruma syntax) are not synonymous. The former is a back reference; the latter is a subroutine call.
🎀 Generic character types 🎀
Another use of backslash is for specifying generic character types:
\d any decimal digit
\D any character that is not a decimal digit
\h any horizontal white space character
\H any character that is not a horizontal white space character
\s any white space character
\S any character that is not a white space character
\v any vertical white space character
\V any character that is not a vertical white space character
\w any "word" character
\W any "non-word" characterThere's also the single sequence \N, which matches a non-newline character. This is the same as the "." metacharacter when PCRE_DOTALL is not set. Perl also uses \N to match characters by name; PCRE does not support this. 🫶
Okay so each pair of lower-case + upper-case escape sequences splits the ENTIRE character set into two disjoint sets 💅 Any given character matches one, and only one, of each pair. These sequences can show up both inside and outside character classes, and each one matches exactly one character of the appropriate type. But — and this is the tea ☕ — if your current matching point is sitting at the end of the subject string, every single one of them fails, because there's literally no character left to match. 💀
For compatibility with Perl, \s did not used to match the VT character (code 11), which made it different from the the POSIX "space" class. But then Perl added VT at release 5.18, and PCRE followed suit at release 8.34 ✨ The default \s characters are now HT (9), LF (10), VT (11), FF (12), CR (13), and space (32), which are defined as white space in the "C" locale. This list may vary if locale-specific matching is taking place — for example, in some locales the "non-breaking space" character (\xA0) is recognized as white space, and in others the VT character is not. Locales stay messy bestie 😭
A "word" character is an underscore or any character that is a letter or digit. By default, the definition of letters and digits is controlled by PCRE's low-valued character tables, and may vary if locale-specific matching is taking place (see "Locale support" in the pcreapi page). For example, in a French locale such as "fr_FR" in Unix-like systems, or "french" in Windows, some character codes greater than 127 are used for accented letters, and these are then matched by \w. Pro tip though: the use of locales with Unicode is discouraged 🙈
By default, characters whose code points are greater than 127 never match \d, \s, or \w, and always match \D, \S, and \W, although this may vary for characters in the range 128-255 when locale-specific matching is happening. These escape sequences retain their original meanings from before Unicode support was available, mainly for efficiency reasons. If PCRE is compiled with Unicode property support, and the PCRE_UCP option is set, the behaviour is changed so that Unicode properties are used to determine character types, as follows:
\d any character that matches \p{Nd} (decimal digit)
\s any character that matches \p{Z} or \h or \v
\w any character that matches \p{L} or \p{N}, plus underscoreThe upper case escapes match the inverse sets of characters 🫶 Note that \d matches only decimal digits, whereas \w matches any Unicode digit, as well as any Unicode letter, and underscore. Note also that PCRE_UCP affects \b, and \B because they are defined in terms of \w and \W. Just so you know: matching these sequences is noticeably slower when PCRE_UCP is set. Worth it? Your call 💅
The sequences \h, \H, \v, and \V are features that were added to Perl at release 5.10 ✨ In contrast to the other sequences, which match only ASCII characters by default, these always match certain high-valued code points, whether or not PCRE_UCP is set. The horizontal space characters are:
U+0009 Horizontal tab (HT)
U+0020 Space
U+00A0 Non-break space
U+1680 Ogham space mark
U+180E Mongolian vowel separator
U+2000 En quad
U+2001 Em quad
U+2002 En space
U+2003 Em space
U+2004 Three-per-em space
U+2005 Four-per-em space
U+2006 Six-per-em space
U+2007 Figure space
U+2008 Punctuation space
U+2009 Thin space
U+200A Hair space
U+202F Narrow no-break space
U+205F Medium mathematical space
U+3000 Ideographic spaceThe vertical space characters are:
U+000A Linefeed (LF)
U+000B Vertical tab (VT)
U+000C Form feed (FF)
U+000D Carriage return (CR)
U+0085 Next line (NEL)
U+2028 Line separator
U+2029 Paragraph separatorIn 8-bit, non-UTF-8 mode, only the characters with codepoints less than 256 are relevant 👑
🌀 Newline sequences
Outside a character class, by default, the escape sequence \R matches any Unicode newline sequence 🌈 In 8-bit non-UTF-8 mode \R is equivalent to the following:
(?>\r\n|\n|\x0b|\f|\r|\x85)This is an example of an "atomic group", details of which are given below 👀 This particular group matches either the two-character sequence CR followed by LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab, U+000B), FF (form feed, U+000C), CR (carriage return, U+000D), or NEL (next line, U+0085). The two-character sequence is treated as a single unit that cannot be split — no breaking up the besties 💖
In other modes, two additional characters whose codepoints are greater than 255 are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029). Unicode character property support is not needed for these characters to be recognized, periodt ✨
It is possible to restrict \R to match only CR, LF, or CRLF (instead of the complete set of Unicode line endings) by setting the option PCRE_BSR_ANYCRLF either at compile time or when the pattern is matched. (BSR is an abbreviation for "backslash R".) This can be made the default when PCRE is built; if this is the case, the other behaviour can be requested via the PCRE_BSR_UNICODE option. It is also possible to specify these settings by starting a pattern string with one of the following sequences:
(*BSR_ANYCRLF) CR, LF, or CRLF only
(*BSR_UNICODE) any Unicode newline sequenceThese override the default and the options given to the compiling function, but they can themselves be overridden by options given to a matching function 👑 Note that these special settings, which are not Perl-compatible, are recognized only at the very start of a pattern, and that they must be in upper case. If more than one of them is present, the last one is used. They can be combined with a change of newline convention; for example, a pattern can start with:
(*ANY)(*BSR_ANYCRLF)They can also be combined with the (*UTF8), (*UTF16), (*UTF32), (*UTF) or (*UCP) special sequences. Inside a character class, \R is treated as an unrecognized escape sequence, and so matches the letter "R" by default, but causes an error if PCRE_EXTRA is set 🙈
🦋 Unicode character properties
When PCRE is built with Unicode character property support, three additional escape sequences that match characters with specific properties are available ✨ When in 8-bit non-UTF-8 mode, these sequences are of course limited to testing characters whose codepoints are less than 256, but they do work in this mode. The extra escape sequences are:
\p{xx} a character with the xx property
\P{xx} a character without the xx property
\X a Unicode extended grapheme clusterThe property names represented by xx above are limited to the Unicode script names, the general category properties, "Any", which matches any character (including newline), and some special PCRE properties (described in the next section) 💅 Other Perl properties such as "InMusicalSymbols" are not currently supported by PCRE. Note that \P{Any} does not match any characters, so always causes a match failure 💀
Sets of Unicode characters are defined as belonging to certain scripts. A character from one of these sets can be matched using a script name. For example:
\p{Greek}
\P{Han}Those that are not part of an identified script are lumped together as "Common". The current list of scripts is:
Arabic, Armenian, Avestan, Balinese, Bamum, Bassa_Vah, Batak, Bengali, Bopomofo, Brahmi, Braille, Buginese, Buhid, Canadian_Aboriginal, Carian, Caucasian_Albanian, Chakma, Cham, Cherokee, Common, Coptic, Cuneiform, Cypriot, Cyrillic, Deseret, Devanagari, Duployan, Egyptian_Hieroglyphs, Elbasan, Ethiopic, Georgian, Glagolitic, Gothic, Grantha, Greek, Gujarati, Gurmukhi, Han, Hangul, Hanunoo, Hebrew, Hiragana, Imperial_Aramaic, Inherited, Inscriptional_Pahlavi, Inscriptional_Parthian, Javanese, Kaithi, Kannada, Katakana, Kayah_Li, Kharoshthi, Khmer, Khojki, Khudawadi, Lao, Latin, Lepcha, Limbu, Linear_A, Linear_B, Lisu, Lycian, Lydian, Mahajani, Malayalam, Mandaic, Manichaean, Meetei_Mayek, Mende_Kikakui, Meroitic_Cursive, Meroitic_Hieroglyphs, Miao, Modi, Mongolian, Mro, Myanmar, Nabataean, New_Tai_Lue, Nko, Ogham, Ol_Chiki, Old_Italic, Old_North_Arabian, Old_Permic, Old_Persian, Old_South_Arabian, Old_Turkic, Oriya, Osmanya, Pahawh_Hmong, Palmyrene, Pau_Cin_Hau, Phags_Pa, Phoenician, Psalter_Pahlavi, Rejang, Runic, Samaritan, Saurashtra, Sharada, Shavian, Siddham, Sinhala, Sora_Sompeng, Sundanese, Syloti_Nagri, Syriac, Tagalog, Tagbanwa, Tai_Le, Tai_Tham, Tai_Viet, Takri, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh, Tirhuta, Ugaritic, Vai, Warang_Citi, Yi.
Each character has exactly one Unicode general category property, specified by a two-letter abbreviation. For compatibility with Perl, negation can be specified by including a circumflex between the opening brace and the property name. For example, \p{^Lu} is the same as \P{Lu} 🫶
If only one letter is specified with \p or \P, it includes all the general category properties that start with that letter. In this case, in the absence of negation, the curly brackets in the escape sequence are optional; these two examples have the same effect:
\p{L}
\pLThe following general category property codes are supported:
C Other
Cc Control
Cf Format
Cn Unassigned
Co Private use
Cs Surrogate
L Letter
Ll Lower case letter
Lm Modifier letter
Lo Other letter
Lt Title case letter
Lu Upper case letter
M Mark
Mc Spacing mark
Me Enclosing mark
Mn Non-spacing mark
N Number
Nd Decimal number
Nl Letter number
No Other number
P Punctuation
Pc Connector punctuation
Pd Dash punctuation
Pe Close punctuation
Pf Final punctuation
Pi Initial punctuation
Po Other punctuation
Ps Open punctuation
S Symbol
Sc Currency symbol
Sk Modifier symbol
Sm Mathematical symbol
So Other symbol
Z Separator
Zl Line separator
Zp Paragraph separator
Zs Space separatorThe special property L& is also supported: it matches a character that has the Lu, Ll, or Lt property, in other words, a letter that is not classified as a modifier or "other" ✨
The Cs (Surrogate) property applies only to characters in the range U+D800 to U+DFFF ⚠️ Such characters are not valid in Unicode strings and so cannot be tested by PCRE, unless UTF validity checking has been turned off (see the discussion of PCRE_NO_UTF8_CHECK, PCRE_NO_UTF16_CHECK and PCRE_NO_UTF32_CHECK in the pcreapi page). Perl does not support the Cs property 🙈
Oh and also — the long fancy property names that Perl lets you write (like \p{Letter}) are not a thing in PCRE, bestie. Sorry. And you can't slap an Is prefix on any of these properties either. It's just not that kind of party. 💅
No character that actually lives in the Unicode table has the Cn (unassigned) property. Instead, PCRE just assumes that property for any code point that isn't in the Unicode table. So Cn is basically the "we don't know her" bucket. 💀
Also, turning on caseless matching does not touch these escape sequences. Like, \p{Lu} is still gonna match upper case letters and only upper case letters, no matter what flags you set. That's genuinely different from how current versions of Perl behave, so don't come for me when your Perl-brain gets confused. 🙈
And real talk: matching characters by Unicode property is not fast, because PCRE has to do a whole multistage table lookup just to find out what a character's property even is. That's literally why the classic escape sequences like \d and \w do not use Unicode properties in PCRE by default. If you want them to, you can set the PCRE_UCP option, or start your pattern with (*UCP). Your call, bestie. 🫶
✨ Extended grapheme clusters ✨
The \X escape matches any number of Unicode characters that together form an "extended grapheme cluster", and it treats that whole sequence as an atomic group (more on atomic groups below). 👑
Up to and including release 8.31, PCRE matched an earlier, simpler definition that was equivalent to:
(?>\PM\pM*)So like — it matched a character without the "mark" property, followed by zero or more characters with the "mark" property. And characters with the "mark" property are typically non-spacing accents that affect the character before them. Cute, but limited. 🎀
Then Unicode went and extended that simple definition to include more complicated kinds of composite character, by giving every character a grapheme breaking property and creating rules that use these properties to define where extended grapheme clusters start and end. In releases of PCRE later than 8.31, \X matches one of these clusters. Glow up fr fr. 🌈
\X always matches at least one character. Then it decides whether to add more characters according to these rules for ending a cluster:
- End at the end of the subject string.
- Do not end between
CRandLF; otherwise end after any control character. - Do not break Hangul (a Korean script) syllable sequences. Hangul characters come in five types:
L,V,T,LV, andLVT. AnLcharacter may be followed by anL,V,LV, orLVTcharacter; anLVorVcharacter may be followed by aVorTcharacter; anLVTorTcharacter may be followed only by aTcharacter. - Do not end before extending characters or spacing marks. Characters with the "mark" property always have the "extend" grapheme breaking property.
- Do not end after prepend characters.
- Otherwise, end the cluster.
🌸 PCRE's additional properties 🌸
On top of all the standard Unicode properties described above, PCRE supports four more that make it possible to convert traditional escape sequences like \w and \s into Unicode-property-based ones. PCRE uses these non-standard, non-Perl properties internally whenever PCRE_UCP is set. But bestie, you can also use them explicitly. They are:
Xan Any alphanumeric character
Xps Any POSIX space character
Xsp Any Perl space character
Xwd Any Perl "word" characterXan matches characters that have either the L (letter) or the N (number) property. Xps matches the characters tab, linefeed, vertical tab, form feed, or carriage return, plus any other character that has the Z (separator) property. Xsp is the same as Xps; it used to exclude vertical tab for Perl compatibility, but Perl changed, and so PCRE followed along at release 8.34. Growth! 🌱 And Xwd matches the same characters as Xan, plus underscore.
There's another non-standard property too: Xuc, which matches any character that can be represented by a Universal Character Name in C++ and other programming languages. Those are the characters $, @, ` (grave accent), and all characters with Unicode code points greater than or equal to U+00A0, except for the surrogates U+D800 to U+DFFF. Heads up though: most base (ASCII) characters are excluded. (Universal Character Names are of the form \uHHHH or \UHHHHHHHH where H is a hexadecimal digit. And note that the Xuc property does not match these sequences — it matches the characters that they represent. Important distinction, don't get it twisted. 💜)
💖 Resetting the match start 💖
The escape sequence \K causes any previously matched characters to not be included in the final matched sequence. Iconic, honestly. For example, the pattern:
foo\Kbarmatches "foobar", but reports that it has matched "bar". Sneaky! This feature is similar to a lookbehind assertion (described below). However — and this is the key part — in this case, the part of the subject before the real match does not have to be of fixed length, the way it does with lookbehind assertions. Flexible queen. 👑
Using \K also doesn't interfere with the setting of captured substrings. For example, when the pattern
(foo)\Kbarmatches "foobar", the first substring is still set to "foo". Captures stay intact, we love to see it. 🎀
Perl documents that using \K within assertions is "not well defined". In PCRE, \K is acted upon when it occurs inside positive assertions, but is ignored in negative assertions. And watch out: when a pattern such as (?=ab\K) matches, the reported start of the match can be greater than the end of the match. Chaos. 💀
🦋 Simple assertions 🦋
The final use of backslash is for certain simple assertions. An assertion specifies a condition that has to be met at a particular point in a match, without consuming any characters from the subject string. It's just vibing there, checking. ✨ The use of subpatterns for more complicated assertions is described below. The backslashed assertions are:
\b matches at a word boundary
\B matches when not at a word boundary
\A matches at the start of the subject
\Z matches at the end of the subject
also matches before a newline at the end of the subject
\z matches only at the end of the subject
\G matches at the first matching position in the subjectInside a character class, \b has a different meaning; it matches the backspace character. And if any of the other assertions here show up in a character class, by default it matches the corresponding literal character (so like, \B matches the letter B). However, if the PCRE_EXTRA option is set, you get an "invalid escape sequence" error instead. Surprise! 🫶
A word boundary is a position in the subject string where the current character and the previous character do not both match \w or \W (i.e. one matches \w and the other matches \W), or the start or end of the string if the first or last character matches \w, respectively. In a UTF mode, the meanings of \w and \W can be changed by setting the PCRE_UCP option. When you do that, it also affects \b and \B. Neither PCRE nor Perl has a separate "start of word" or "end of word" metasequence. But whatever follows \b normally determines which one it is. For example, the fragment \ba matches "a" at the start of a word. Slay. 🌟
The \A, \Z, and \z assertions differ from the traditional circumflex and dollar (described in the next section) in that they only ever match at the very start and end of the subject string, whatever options are set. So they're fully independent of multiline mode — no notes. 🧚 These three assertions are not affected by the PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the circumflex and dollar metacharacters. However, if the startoffset argument of pcre_exec() is non-zero — meaning matching starts at a point other than the beginning of the subject — \A can never match. Not once. 💀 The difference between \Z and \z is that \Z matches before a newline at the end of the string as well as at the very end, whereas \z matches only at the end.
The \G assertion is true only when the current matching position is at the start point of the match, as specified by the startoffset argument of pcre_exec(). It differs from \A when the value of startoffset is non-zero. By calling pcre_exec() multiple times with appropriate arguments, you can mimic Perl's /g option, and it's in this kind of implementation where \G can be useful. Main character energy. 👑
Note, however, that PCRE's interpretation of \G, as the start of the current match, is subtly different from Perl's, which defines it as the end of the previous match. In Perl, these can be different when the previously matched string was empty. And because PCRE does just one match at a time, it cannot reproduce this behaviour. Sorry bestie, that's just the way it is. 🙈
If all the alternatives of a pattern begin with \G, the expression is anchored to the starting match position, and the "anchored" flag is set in the compiled regular expression. Locked in. 🔒
CIRCUMFLEX AND DOLLAR 💅
So the circumflex (^) and dollar ($) metacharacters are zero-width assertions. That means they just test whether a particular condition is true, no cap — they do not consume any characters from the subject string. They just stand there and judge. Periodt.
Outside a character class, in the default matching mode, ^ is an assertion that's true only if the current matching point is at the start of the subject string. If the startoffset argument of pcre_exec() is non-zero, ^ can never match if the PCRE_MULTILINE option is unset. Inside a character class, ^ has an entirely different meaning — see below, bestie. 🙈
^ doesn't have to be the first character of the pattern if a bunch of alternatives are involved, but it really should be the first thing in each alternative where it appears if the pattern is ever gonna match that branch. If all possible alternatives start with ^, meaning the pattern is constrained to match only at the start of the subject, it's called an "anchored" pattern. (There are also other constructs that can cause a pattern to be anchored, iykyk.)
The dollar character ($) is an assertion that's true only if the current matching point is at the end of the subject string, or immediately before a newline at the end of the string (by default). But heads up: it does not actually match that newline. $ doesn't have to be the last character of the pattern if a number of alternatives are involved, but it should be the last item in any branch where it appears. And inside a character class? $ has no special meaning at all. She's just a regular character there. 💅
You can change what $ means so it matches only at the very end of the string by setting the PCRE_DOLLAR_ENDONLY option at compile time. This does not affect the \Z assertion, just so you know. ✨
The meanings of ^ and $ get changed if the PCRE_MULTILINE option is set. When that's the case, ^ matches immediately after internal newlines as well as at the start of the subject string. It does not match after a newline that ends the string. $ matches before any newlines in the string, as well as at the very end, when PCRE_MULTILINE is set. When newline is specified as the two-character sequence CRLF, isolated CR and LF characters do not indicate newlines. No cap.
For example, the pattern /^abc$/ matches the subject string "def\nabc" (where \n represents a newline) in multiline mode, but not otherwise. Consequently, patterns that are anchored in single line mode because all branches start with ^ are not anchored in multiline mode, and a match for ^ is possible when the startoffset argument of pcre_exec() is non-zero. The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set. She said what she said. 💅
Note that the sequences \A, \Z, and \z can be used to match the start and end of the subject in both modes, and if all branches of a pattern start with \A it is always anchored, whether or not PCRE_MULTILINE is set. Periodt. 🫶
FULL STOP (PERIOD, DOT) AND \N 🦋
Outside a character class, a dot (.) in the pattern matches any one character in the subject string except — by default — a character that signifies the end of a line. So she's giving "anyone but the line-ender" energy. 💅
When a line ending is defined as a single character, dot never matches that character. When the two-character sequence CRLF is used, dot does not match CR if it is immediately followed by LF, but otherwise it matches all characters (including isolated CRs and LFs). When any Unicode line endings are being recognized, dot does not match CR or LF or any of the other line ending characters. She's picky, okay? 🙈
The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL option is set, a dot matches any one character, without exception. If the two-character sequence CRLF is present in the subject string, it takes two dots to match it. Math is mathing. ✨
The handling of dot is entirely independent of the handling of ^ and $ — the only relationship being that they both involve newlines. Dot has no special meaning in a character class. None. Zero. 💀
The escape sequence \N behaves like a dot, except that it is not affected by the PCRE_DOTALL option. In other words, it matches any character except one that signifies the end of a line. Perl also uses \N to match characters by name; PCRE does not support this. Sorry bestie, not today. 😭
MATCHING A SINGLE DATA UNIT 🎀
Outside a character class, the escape sequence \C matches any one data unit, whether or not a UTF mode is set. In the 8-bit library, one data unit is one byte; in the 16-bit library it is a 16-bit unit; in the 32-bit library it is a 32-bit unit. Unlike a dot, \C always matches line-ending characters. She does not care. 😤
The feature is provided in Perl in order to match individual bytes in UTF-8 mode, but it is unclear how it can usefully be used. Because \C breaks up characters into individual data units, matching one unit with \C in a UTF mode means that the rest of the string may start with a malformed UTF character. This has undefined results, because PCRE assumes that it is dealing with valid UTF strings (and by default it checks this at the start of processing unless the PCRE_NO_UTF8_CHECK, PCRE_NO_UTF16_CHECK or PCRE_NO_UTF32_CHECK option is used). So like, danger. 🚨
PCRE does not allow \C to appear in lookbehind assertions (described below) in a UTF mode, because this would make it impossible to calculate the length of the lookbehind. And lookbehind needs to know its length, bestie. Periodt. 💅
In general, the \C escape sequence is best avoided. However, one way of using it that avoids the problem of malformed UTF characters is to use a lookahead to check the length of the next character, as in this pattern, which could be used with a UTF-8 string (ignore white space and line breaks):
(?| (?=[\x00-\x7f])(\C) |
(?=[\x80-\x{7ff}])(\C)(\C) |
(?=[\x{800}-\x{ffff}])(\C)(\C)(\C) |
(?=[\x{10000}-\x{1fffff}])(\C)(\C)(\C)(\C))A group that starts with (?| resets the capturing parentheses numbers in each alternative (see "Duplicate Subpattern Numbers" below). The assertions at the start of each branch check the next UTF-8 character for values whose encoding uses 1, 2, 3, or 4 bytes, respectively. The character's individual bytes are then captured by the appropriate number of groups. It's giving ✨organized chaos✨.
💖 SQUARE BRACKETS AND CHARACTER CLASSES 💖
Okay bestie, this is where it gets fun. An opening square bracket [ starts a character class, and a closing square bracket ] ends it. Periodt. A closing square bracket all on its own is not special by default. BUT — plot twist — if the PCRE_JAVASCRIPT_COMPAT option is set, a lone closing square bracket throws a compile-time error. So if you actually need a ] as a member of the class, either make it the first data character in the class (right after an initial circumflex, if there is one) or escape it with a backslash. Slay. ✨
A character class matches a single character in the subject. In a UTF mode, that character might be more than one data unit long, no big deal. The matched character has to be in the set of characters the class defines — unless the very first character in the class definition is a circumflex, in which case the subject character must not be in that set. If you genuinely need a circumflex as a member of the class, just don't put it first, or escape it with a backslash. Easy. 🎀
For example, the character class [aeiou] matches any lower case vowel, while [^aeiou] matches any character that is not a lower case vowel. Note that a circumflex is literally just a convenient notation for specifying the characters in the class by listing all the ones that aren't. A class starting with a circumflex is not an assertion — it still consumes a character from the subject string, so it fails if the current pointer is sitting at the end of the string. it's giving "not everything is what it looks like." 😭
In UTF-8 (UTF-16, UTF-32) mode, characters with values greater than 255 (0xffff) can go in a class as a literal string of data units, or by using the \x{ escaping mechanism. iconic.
When caseless matching is set, any letters in a class represent both their upper case and lower case versions. So for example, a caseless [aeiou] matches "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a caseful version would. In a UTF mode, PCRE always understands the concept of case for characters whose values are less than 128, so caseless matching is always possible there. For characters with higher values, case is supported only if PCRE was compiled with Unicode property support — otherwise, nope. So if you want caseless matching in a UTF mode for characters 128 and above, you must make sure PCRE is compiled with Unicode property support as well as with UTF support. double-check that build, bestie 🫶
Characters that might indicate line breaks are never treated in any special way when matching character classes — doesn't matter what line-ending sequence is in use, doesn't matter what PCRE_DOTALL and PCRE_MULTILINE are set to. A class like [^a] always matches one of those characters. 💅
The minus (hyphen) character can specify a range of characters in a class. For example, [d-m] matches any letter between d and m, inclusive. If you need a literal minus in a class, escape it with a backslash or put it somewhere it can't be read as a range — typically as the first or last character in the class, or immediately after a range. For example, [b-d-z] matches letters in the range b to d, a hyphen character, or z. 🌈
It is not possible to have the literal character "]" as the end character of a range. A pattern like [W-]46] is interpreted as a class of two characters ("W" and "-") followed by a literal string "46]", so it would match "W46]" or "-46]". However, if the "]" is escaped with a backslash it is interpreted as the end of the range, so [W-\]46] is interpreted as a class containing a range followed by two other characters. The octal or hexadecimal representation of "]" can also be used to end a range. the way that just works 💀
An error is generated if a POSIX character class (peep below) or an escape sequence other than one that defines a single character shows up at a point where a range ending character is expected. For example, [z-\xff] is valid, but [A-\d] and [A-[:digit:]] are not. 🙈
Ranges operate in the collating sequence of character values. They can also be used for characters specified numerically, for example [\000-\037]. Ranges can include any characters that are valid for the current mode.
If a range that includes letters is used when caseless matching is set, it matches the letters in either case. For example, [W-c] is equivalent to [][\^_wxyzabc], matched caselessly, and in a non-UTF mode, if character tables for a French locale are in use, [\xc8-\xcb]` matches accented E characters in both cases. In UTF modes, PCRE supports the concept of case for characters with values greater than 128 only when it is compiled with Unicode property support. lowkey a build-flag situation, always 😌
The character escape sequences \d, \D, \h, \H, \p, \P, \s, \S, \v, \V, \w, and \W may appear in a character class, and they add the characters they match to the class. For example, [\dABCDEF] matches any hexadecimal digit. In UTF modes, the PCRE_UCP option affects the meanings of \d, \s, \w and their upper case partners, exactly like it does when they appear outside a character class, as described in the section entitled "Generic character types" above. The escape sequence \b means something different inside a character class — it matches the backspace character. The sequences \B, \N, \R, and \X are not special inside a character class. Like any other unrecognized escape sequences, they're treated as the literal characters "B", "N", "R", and "X" by default, but they cause an error if the PCRE_EXTRA option is set. ⚠️
A circumflex can conveniently be used with the upper case character types to specify a more restricted set of characters than the matching lower case type. For example, the class [^\W_] matches any letter or digit, but not underscore, whereas [\w] includes underscore. A positive character class should be read as "something OR something OR ..." and a negative class as "NOT something AND NOT something AND NOT ...". 🧚
The only metacharacters that are recognized in character classes are:
- backslash
- hyphen (only where it can be interpreted as specifying a range)
- circumflex (only at the start)
- opening square bracket (only when it can be interpreted as introducing a POSIX class name, or for a special compatibility feature — see the next two sections)
- the terminating closing square bracket
However, escaping other non-alphanumeric characters does no harm. harmless queen behavior 💜
🌸 POSIX CHARACTER CLASSES 🌸
Perl supports the POSIX notation for character classes, which uses names enclosed by [: and :] within the enclosing square brackets. PCRE supports this notation too. For example,
[01[:alpha:]%]matches "0", "1", any alphabetic character, or "%". The supported class names are:
alnum— letters and digitsalpha— lettersascii— character codes 0 - 127blank— space or tab onlycntrl— control charactersdigit— decimal digits (same as\d)graph— printing characters, excluding spacelower— lower case lettersprint— printing characters, including spacepunct— printing characters, excluding letters and digits and spacespace— white space (the same as\sfrom PCRE 8.34)upper— upper case lettersword—"word"characters (same as\w)xdigit— hexadecimal digits
The default "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13), and space (32). If locale-specific matching is happening, the list of space characters may be different — there might be fewer or more of them. Fun fact for the lore: "Space" used to be different from \s, which didn't include VT, for Perl compatibility. But Perl changed at release 5.18, and PCRE followed at release 8.34. So "Space" and \s now match the same set of characters. character development!! 🦋
The name "word" is a Perl extension, and "blank" is a GNU extension from Perl 5.8. Another Perl extension is negation, indicated by a ^ character after the colon. For example,
[12[:^digit:]]matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not supported, and an error is given if they are encountered. 🚫
By default, characters with values greater than 128 do not match any of the POSIX character classes. However, if the PCRE_UCP option is passed to pcre_compile(), some of the classes change so that Unicode character properties are used. This is achieved by replacing certain POSIX classes by other sequences, as follows:
[:alnum:]becomes\p{Xan}[:alpha:]becomes\p{L}[:blank:]becomes\h[:digit:]becomes\p{Nd}[:lower:]becomes\p{Ll}[:space:]becomes\p{Xps}[:upper:]becomes\p{Lu}[:word:]becomes\p{Xwd}
Negated versions, such as [:^alpha:] use \P instead of \p. Three other POSIX classes are handled specially in UCP mode:
[:graph:]— This matches characters that have glyphs that mark the page when printed. In Unicode property terms, it matches all characters with the L, M, N, P, S, or Cf properties, except for:U+061C— Arabic Letter MarkU+180E— Mongolian Vowel SeparatorU+2066 - U+2069— Various"isolate"s
[:print:]— This matches the same characters as[:graph:]plus space characters that are not controls, that is, characters with theZsproperty.[:punct:]— This matches all characters that have the Unicode P (punctuation) property, plus those characters whose code points are less than 128 that have the S (Symbol) property.
The other POSIX classes are unchanged, and match only characters with code points less than 128.
👑 COMPATIBILITY FEATURE FOR WORD BOUNDARIES 👑
So, back in the POSIX.2 compliant library that was included in 4.4BSD Unix, there was this ugly syntax — [[:<:]] and [[:>:]] — used for matching "start of word" and "end of word". PCRE treats these items as follows:
[[:<:]] is converted to \b(?=\w)
[[:>:]] is converted to \b(?<=\w)Only these exact character sequences are recognized. A sequence like [a[:<:]b] provokes an error for an unrecognized POSIX class name. This support is not compatible with Perl. It's provided to help migrations from other environments, and it is best not used in any new patterns. Note that \b matches at the start and the end of a word (see "Simple assertions" above), and in a Perl-style pattern the preceding or following character normally shows which one you want, without needing the assertions used above in order to give exactly the POSIX behaviour. so yeah, legacy vibes only 💅✨
🎀 VERTICAL BAR 🎀
Ok so the vertical bar? That's the "or" girlie of the pattern world. Alternatives, bestie. You use | to separate them.
Like, this pattern:
gilbert|sullivanmatches either "gilbert" or "sullivan". Either one, pick your fighter. And you can have as many alternatives as you want — go off, honestly. An empty alternative is also totally permitted (and yeah, that matches the empty string, she's valid).
Here's the tea on how it decides: the matching process tries each alternative in turn, from left to right, and the first one that succeeds is the one that gets used. No take-backs.
And if the alternatives are chilling inside a subpattern (we get to what that is in a sec, hold on), then "succeeds" means something bigger: it means matching the rest of the main pattern as well as the alternative in the subpattern. So it's not just "did this branch work", it's "did this branch work and does everything after it still line up". 💅
✨ INTERNAL OPTION SETTING ✨
So you know how options get set by the application? Plot twist: the pattern itself can change them from the inside. The settings of PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and PCRE_EXTENDED (all the Perl-compatible ones) can be flipped by a lil sequence of Perl option letters wrapped in (? and ).
The letters, bestie:
i for PCRE_CASELESS
m for PCRE_MULTILINE
s for PCRE_DOTALL
x for PCRE_EXTENDEDSo like, (?im) sets caseless + multiline matching. Iconic.
You can also unset them by putting a hyphen before the letter. And yeah, you can do both at once — a combined setting-and-unsetting like (?im-sx), which sets PCRE_CASELESS and PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is completely allowed. 🌈
Heads up: if a letter shows up both before and after the hyphen, the option gets unset. The hyphen wins. Periodt.
Now for the PCRE-specific ones — PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA — you change those the exact same way, just using the characters J, U and X respectively. Slay.
Here's where it gets interesting. When one of these option changes happens at top level (meaning: not inside subpattern parentheses), the change applies to the remainder of the pattern that follows. But an option change inside a subpattern (subpattern description incoming below, promise) only affects that part of the subpattern that follows it. So:
(a(?i)b)cmatches abc and aBc and no other strings (assuming PCRE_CASELESS is not already in use). That's the whole point — you can make options have different settings in different parts of the pattern. Customize your lil heart out. 💖
And any changes made in one alternative do carry on into subsequent branches within the same subpattern. Example:
(a(?i)b|c)matches "ab", "aB", "c", and "C" — even though when matching "C", the first branch gets abandoned before the option setting even happens. Why?? Because the effects of option settings happen at compile time. Otherwise there'd be some truly unhinged behaviour and we simply cannot.
Note: there are other PCRE-specific options the application can set when the compiling or matching functions are called. In some cases the pattern can contain special leading sequences such as(*CRLF)to override what the application has set or what has been defaulted. Details live in the section entitled "Newline sequences" above. There are also the(*UTF8),(*UTF16),(*UTF32), and(*UCP)leading sequences that can be used to set UTF and Unicode property modes — they're equivalent to setting thePCRE_UTF8,PCRE_UTF16,PCRE_UTF32and thePCRE_UCPoptions, respectively. The(*UTF)sequence is a generic version that can be used with any of the libraries. However — the application can set thePCRE_NEVER_UTFoption, which locks out the use of the(*UTF)sequences. No entry, sorry babe.
🦋 SUBPATTERNS 🦋
Subpatterns are delimited by parentheses (round brackets), and yes, they can be nested. Little subpattern matryoshka dolls. 🪆
Turning part of a pattern into a subpattern does two things:
- It localizes a set of alternatives. For example, the pattern
cat(aract|erpillar|)matches "cataract", "caterpillar", or "cat". Without the parentheses, it would match "cataract", "erpillar" or an empty string — which is not the vibe. See how much the brackets are carrying?
- It sets up the subpattern as a capturing subpattern. Meaning: when the whole pattern matches, that portion of the subject string that matched the subpattern gets passed back to the caller via the
ovectorargument of the matching function. (Important caveat: this only applies to the traditional matching functions. The DFA matching functions do not support capturing. F in the chat for DFA, she's doing her own thing. 💀)
Opening parentheses are counted from left to right, starting from 1, to get the numbers for the capturing subpatterns. So like, if the string "the red king" is matched against:
the ((red|white) (king|queen))the captured substrings are "red king", "red", and "king", and they're numbered 1, 2, and 3, respectively. Numbered in the order they open, not the order they close. Noted. ✨
Now — the fact that plain parentheses pull double duty (grouping and capturing) is not always helpful, ngl. Sometimes you just want grouping with zero capturing energy. So: if an opening parenthesis is followed by a question mark and a colon, the subpattern does no capturing at all, and it's not counted when computing the number of any subsequent capturing subpatterns. Bestie said "I'm just here to organize, not to be perceived."
For example, matching "the white queen" against:
the ((?:red|white) (king|queen))gives you captured substrings "white queen" and "queen", numbered 1 and 2. Clean. And the maximum number of capturing subpatterns is 65535 — that's the ceiling, don't test it.
As a convenient lil shorthand: if you need option settings at the start of a non-capturing subpattern, the option letters can go between the ? and the :. So these two patterns:
(?i:saturday|sunday)
(?:(?i)saturday|sunday)match exactly the same set of strings. Why? Because alternative branches are tried left to right, and options are not reset until the end of the subpattern is reached, so an option setting in one branch does affect subsequent branches. Which is why both of those patterns match "SUNDAY" as well as "Saturday". The i is doing overtime and we love her for it. 💅
👑 DUPLICATE SUBPATTERN NUMBERS 👑
Okay so Perl 5.10 introduced this feature where every alternative in a subpattern uses the same numbers for its capturing parentheses. A subpattern like this starts with (?| and is itself a non-capturing subpattern. Very "share the number, share the love."
Check it:
(?|(Sat)ur|(Sun))dayBecause both alternatives are inside a (?| group, both sets of capturing parentheses are numbered one. So when the pattern matches, you can just look at captured substring number one — whichever alternative matched. Chef's kiss for when you want to capture part, but not all, of one of a bunch of alternatives. 🍓
Inside a (?| group, parentheses are numbered as usual, but the number gets reset at the start of each branch. And the numbers of any capturing parentheses that follow the subpattern start after the highest number used in any branch.
Here's the example straight outta the Perl documentation — the numbers underneath show in which buffer the captured content will be stored:
# before ---------------branch-reset----------- after
/ ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
# 1 2 2 3 2 3 4Look at that numbering map. She's giving diagram, she's giving clarity. 🌈
A back reference to a numbered subpattern uses the most recent value that is set for that number by any subpattern. So:
/(?|(abc)|(def))\1/matches "abcabc" or "defdef". Whoever set that number last, that's who you're referencing. Gossip chain but make it regex.
In contrast, a subroutine call to a numbered subpattern always refers to the first one in the pattern with the given number. So:
/(?|(abc)|(def))(?1)/matches "abcabc" or "defabc". Notice the second one — def then abc?? Yeah. Because (?1) reaches back to the first subpattern of that number, no matter which branch ran. Loyal to the original. 🫶
And if a condition test for a subpattern's having matched refers to a non-unique number, the test is true if any of the subpatterns of that number have matched. Any one of them counts, she's not picky.
An alternative approach to using this "branch reset" feature is to use duplicate named subpatterns, as described in the next section. Just saying. Options, bestie. 💖
✨ NAMED SUBPATTERNS ✨
okay bestie, so like — identifying capturing parentheses by number is simple, but it can be very hard to keep track of the numbers in complicated regular expressions. and worse? if an expression gets modified, the numbers may change. rude. 😭
to help with this difficulty, PCRE supports the naming of subpatterns. this feature was not added to Perl until release 5.10. Python had the feature earlier, and PCRE introduced it at release 4.0, using the Python syntax. PCRE now supports both the Perl and the Python syntax. Perl allows identically numbered subpatterns to have different names, but PCRE does not. periodt.
in PCRE, a subpattern can be named in one of three ways: (?<name>...) or (?'name'...) as in Perl, or (?P<name>...) as in Python. references to capturing parentheses from other parts of the pattern, such as back references, recursion, and conditions, can be made by name as well as by number. iconic behavior, honestly. 💅
names consist of up to 32 alphanumeric characters and underscores, but must start with a non-digit. named capturing parentheses are still allocated numbers as well as names, exactly as if the names were not present. the PCRE API provides function calls for extracting the name-to-number translation table from a compiled pattern. there is also a convenience function for extracting a captured substring by name. slay.
by default, a name must be unique within a pattern, but it is possible to relax this constraint by setting the PCRE_DUPNAMES option at compile time. (duplicate names are also always permitted for subpatterns with the same number, set up as described in the previous section.) duplicate names can be useful for patterns where only one instance of the named parentheses can match. suppose you want to match the name of a weekday, either as a 3-letter abbreviation or as the full name, and in both cases you want to extract the abbreviation. this pattern (ignoring the line breaks) does the job:
(?<DN>Mon|Fri|Sun)(?:day)?|
(?<DN>Tue)(?:sday)?|
(?<DN>Wed)(?:nesday)?|
(?<DN>Thu)(?:rsday)?|
(?<DN>Sat)(?:urday)?there are five capturing substrings, but only one is ever set after a match. (an alternative way of solving this problem is to use a "branch reset" subpattern, as described in the previous section.) 🫶
the convenience function for extracting the data by name returns the substring for the first (and in this example, the only) subpattern of that name that matched. this saves searching to find which numbered subpattern it was. lowkey a lifesaver.
if you make a back reference to a non-unique named subpattern from elsewhere in the pattern, the subpatterns to which the name refers are checked in the order in which they appear in the overall pattern. the first one that is set is used for the reference. for example, this pattern matches both "foofoo" and "barbar" but not "foobar" or "barfoo":
(?:(?<n>foo)|(?<n>bar))\k<n>if you make a subroutine call to a non-unique named subpattern, the one that corresponds to the first occurrence of the name is used. in the absence of duplicate numbers (see the previous section) this is the one with the lowest number.
if you use a named reference in a condition test (see the section about conditions below), either to check whether a subpattern has matched, or to check for recursion, all subpatterns with the same name are tested. if the condition is true for any one of them, the overall condition is true. this is the same behaviour as testing by number. for further details of the interfaces for handling named subpatterns, see the pcreapi documentation. iykyk.
⚠️ warning: you cannot use different names to distinguish between two subpatterns with the same number because PCRE uses only the numbers when matching. for this reason, an error is given at compile time if different names are given to subpatterns with the same number. however, you can always give the same name to subpatterns with the same number, even when PCRE_DUPNAMES is not set.
✨ REPETITION ✨
repetition is specified by quantifiers, which can follow any of the following items:
- a literal data character
- the dot metacharacter
- the
\Cescape sequence - the
\Xescape sequence - the
\Rescape sequence - an escape such as
\dor\pLthat matches a single character - a character class
- a back reference (see next section)
- a parenthesized subpattern (including assertions)
- a subroutine call to a subpattern (recursive or otherwise)
the general repetition quantifier specifies a minimum and maximum number of permitted matches, by giving the two numbers in curly brackets (braces), separated by a comma. the numbers must be less than 65536, and the first must be less than or equal to the second. for example:
z{2,4}matches "zz", "zzz", or "zzzz". a closing brace on its own is not a special character. if the second number is omitted, but the comma is present, there is no upper limit; if the second number and the comma are both omitted, the quantifier specifies an exact number of required matches. thus
[aeiou]{3,}matches at least 3 successive vowels, but may match many more, while
\d{8}matches exactly 8 digits. an opening curly bracket that appears in a position where a quantifier is not allowed, or one that does not match the syntax of a quantifier, is taken as a literal character. for example, {,6} is not a quantifier, but a literal string of four characters. 💀
in UTF modes, quantifiers apply to characters rather than to individual data units. thus, for example, \x{100}{2} matches two characters, each of which is represented by a two-byte sequence in a UTF-8 string. similarly, \X{3} matches three Unicode extended grapheme clusters, each of which may be several data units long (and they may be of different lengths).
the quantifier {0} is permitted, causing the expression to behave as if the previous item and the quantifier were not present. this may be useful for subpatterns that are referenced as subroutines from elsewhere in the pattern (but see also the section entitled "Defining subpatterns for use by reference only" below). items other than subpatterns that have a {0} quantifier are omitted from the compiled pattern.
for convenience, the three most common quantifiers have single-character abbreviations:
* is equivalent to {0,}
+ is equivalent to {1,}
? is equivalent to {0,1}it is possible to construct infinite loops by following a subpattern that can match no characters with a quantifier that has no upper limit, for example:
(a?)*earlier versions of Perl and PCRE used to give an error at compile time for such patterns. however, because there are cases where this can be useful, such patterns are now accepted, but if any repetition of the subpattern does in fact match no characters, the loop is forcibly broken. not me learning this the hard way. 🙈
by default, the quantifiers are "greedy", that is, they match as much as possible (up to the maximum number of permitted times), without causing the rest of the pattern to fail. the classic example of where this gives problems is in trying to match comments in C programs. these appear between /* and */ and within the comment, individual * and / characters may appear. an attempt to match C comments by applying the pattern
/\*.*\*/to the string
/* first comment */ not comment /* second comment */fails, because it matches the entire string owing to the greediness of the .* item. it's giving chaos. 😭
however, if a quantifier is followed by a question mark, it ceases to be greedy, and instead matches the minimum number of times possible, so the pattern
/\*.*?\*/does the right thing with the C comments. the meaning of the various quantifiers is not otherwise changed, just the preferred number of matches. do not confuse this use of question mark with its use as a quantifier in its own right. because it has two uses, it can sometimes appear doubled, as in
\d??\dwhich matches one digit by preference, but can match two if that is the only way the rest of the pattern matches. main character energy, honestly.
if the PCRE_UNGREEDY option is set (an option that is not available in Perl), the quantifiers are not greedy by default, but individual ones can be made greedy by following them with a question mark. in other words, it inverts the default behaviour. ✨
when a parenthesized subpattern is quantified with a minimum repeat count that is greater than 1 or with a limited maximum, more memory is required for the compiled pattern, in proportion to the size of the minimum or maximum. just so you know.
if a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is implicitly anchored, because whatever follows will be tried against every character position in the subject string, so there is no point in retrying the overall match at any position after the first. PCRE normally treats such a pattern as though it were preceded by \A. no cap.
in cases where it is known that the subject string contains no newlines, it is worth setting PCRE_DOTALL in order to obtain this optimization, or alternatively using ^ to indicate anchoring explicitly.
however, there are some cases where the optimization cannot be used. when .* is inside capturing parentheses that are the subject of a back reference elsewhere in the pattern, a match at the start may fail where a later one succeeds. consider, for example:
(.*)abc\1if the subject is "xyz123abc123" the match point is the fourth character. for this reason, such a pattern is not implicitly anchored.
another case where implicit anchoring is not applied is when the leading .* is inside an atomic group. once again, a match at the start may fail where a later one succeeds. consider this pattern:
(?>. *?a)bit matches "ab" in the subject "aab". the use of the backtracking control verbs (*PRUNE) and (*SKIP) also disable this optimization. 💅
when a capturing subpattern is repeated, the value captured is the substring that matched the final iteration. for example, after
(tweedle[dume]{3}\s*)+has matched "tweedledum tweedledee" the value of the captured substring is "tweedledee". however, if there are nested capturing subpatterns, the corresponding captured values may have been set in previous iterations. for example, after
/(a|(b))+/matches "aba" the value of the second captured substring is "b". and that's the vibe. 🦋
💖 ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS 💖
Okay so bestie, picture this: both maximizing ("greedy") and minimizing ("ungreedy" or "lazy") repetition have this whole thing where, if the stuff after them fails, the repeated item gets re-evaluated to see if a different number of repeats would let the rest of the pattern match. And sometimes? You just don't want that. Either you're changing the nature of the match, or you wanna make it fail earlier than it otherwise would, because you, the pattern author, already know there's no point in carrying on. ✨
Take the pattern \d+foo when it's applied to the subject line
123456barAfter matching all 6 digits and then failing to match "foo", the normal matcher behavior is to try again with only 5 digits matching the \d+ item, then with 4, and so on, until it ultimately just fails. "Atomic grouping" (a term borrowed from Jeffrey Friedl's book) is the tool that lets you say: once this subpattern has matched, it is not to be re-evaluated this way. Periodt. 💅
So if we use atomic grouping on that previous example, the matcher gives up immediately the first time it fails to match "foo". The notation is a special kind of parenthesis that starts with (?> — like this:
(?>\d+)fooThis kind of parenthesis "locks up" the part of the pattern it contains once it's matched, and a failure further into the pattern is prevented from backtracking into it. Backtracking past it to previous items, though? Totally works as normal. 🫶
Another way to say it: a subpattern of this type matches the string of characters that an identical standalone pattern would match, if it were anchored at the current point in the subject string.
Atomic grouping subpatterns are not capturing subpatterns. Simple cases like the example above can be thought of as a maximizing repeat that must swallow everything it can. So while both \d+ and \d+? are prepared to adjust the number of digits they match to make the rest of the pattern match, (?>\d+) can only match an entire sequence of digits. No negotiating. 🎀
Atomic groups in general can of course contain arbitrarily complicated subpatterns, and yes, they can be nested. But when the subpattern for an atomic group is just a single repeated item, like in the example above, there's a simpler notation — a "possessive quantifier". It's just an extra + character following a quantifier. Using it, the previous example can be rewritten as
\d++fooNote that a possessive quantifier can be used with an entire group, for example:
(abc|xyz){2,3}+Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY option is ignored. They're a convenient notation for the simpler forms of atomic group. However — and this is important — there is no difference in meaning between a possessive quantifier and the equivalent atomic group, though there may be a performance difference; possessive quantifiers should be slightly faster. 🏃♀️💨
The possessive quantifier syntax is an extension to the Perl 5.8 syntax. Jeffrey Friedl originated the idea (and the name) in the first edition of his book. Mike McCloskey liked it, so he implemented it when he built Sun's Java package, and PCRE copied it from there. It ultimately found its way into Perl at release 5.10. Iconic lineage honestly. 👑
PCRE has an optimization that automatically "possessifies" certain simple pattern constructs. For example, the sequence A+B is treated as A++B because there's no point in backtracking into a sequence of A's when B must follow.
Now here's the part where things can get painfully slow if you're not careful. When a pattern contains an unlimited repeat inside a subpattern that can itself be repeated an unlimited number of times, using an atomic group is the only way to avoid some failing matches taking a very long time indeed.
The pattern
(\D+|<\d+>)*[!?]matches an unlimited number of substrings that either consist of non-digits, or digits enclosed in <>, followed by either ! or ?. When it matches, it runs quickly. But if it's applied to
aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaait takes a long time before reporting failure. Why? Because the string can be divided between the internal \D+ repeat and the external * repeat in a huge number of ways, and all of them have to be tried. 😭 (The example uses [!?] rather than a single character at the end, because both PCRE and Perl have an optimization that allows for fast failure when a single character is used. They remember the last single character that's required for a match, and fail early if it isn't present in the string.) If the pattern is changed so that it uses an atomic group, like this:
((?>\D+)|<\d+>)*[!?]then sequences of non-digits cannot be broken, and failure happens quickly. Slay. ✨
🔙 BACK REFERENCES 🔙
Outside a character class, a backslash followed by a digit greater than 0 (and possibly further digits) is a back reference to a capturing subpattern earlier (that is, to its left) in the pattern — provided there have been that many previous capturing left parentheses.
However! If the decimal number following the backslash is less than 10, it is always taken as a back reference, and it causes an error only if there aren't that many capturing left parentheses in the entire pattern. In other words, the parentheses being referenced don't need to be to the left of the reference for numbers less than 10. A "forward back reference" of this type can make sense when a repetition is involved and the subpattern to the right has participated in an earlier iteration. 🌀
It is not possible to have a numerical "forward back reference" to a subpattern whose number is 10 or more using this syntax, because a sequence such as \50 is interpreted as a character defined in octal. (See the subsection entitled "Non-printing characters" above for further details of the handling of digits following a backslash.) There's no such problem when named parentheses are used — a back reference to any subpattern is possible using named parentheses (see below). 🎀
Another way of avoiding the ambiguity inherent in the use of digits following a backslash is to use the \g escape sequence. This escape must be followed by an unsigned number or a negative number, optionally enclosed in braces. These examples are all identical:
(ring), \1
(ring), \g1
(ring), \g{1}An unsigned number specifies an absolute reference without the ambiguity that's present in the older syntax. It's also useful when literal digits follow the reference. A negative number is a relative reference. Consider this example:
(abc(def)ghi)\g{-1}The sequence \g{-1} is a reference to the most recently started capturing subpattern before \g, that is, it's equivalent to \2 in this example. Similarly, \g{-2} would be equivalent to \1. Relative references are super helpful in long patterns, and also in patterns that are created by joining together fragments that contain references within themselves. 🧩
A back reference matches whatever actually matched the capturing subpattern in the current subject string, rather than anything matching the subpattern itself (see "Subpatterns as subroutines" below for a way of doing that). So the pattern
(sens|respons)e and \1ibilitymatches "sense and sensibility" and "response and responsibility", but not "sense and responsibility". And if caseful matching is in force at the time of the back reference, the case of letters matters. For example,
((?i)rah)\s+\1matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original capturing subpattern is matched caselessly. Rude but true. 🙈
There are several different ways of writing back references to named subpatterns. The .NET syntax \k{name} and the Perl syntax \k<name> or \k'name' are supported, as is the Python syntax (?P=name). Perl 5.10's unified back reference syntax, in which \g can be used for both numeric and named references, is also supported. We could rewrite the above example in any of the following ways:
(?<p1>(?i)rah)\s+\k<p1>
(?'p1'(?i)rah)\s+\k{p1}
(?P<p1>(?i)rah)\s+(?P=p1)
(?<p1>(?i)rah)\s+\g{p1}A subpattern that is referenced by name may appear in the pattern before or after the reference. 💫
There may be more than one back reference to the same subpattern. If a subpattern has not actually been used in a particular match, any back references to it always fail by default. For example, the pattern
(a|(bc))\2always fails if it starts to match "a" rather than "bc". However, if the PCRE_JAVASCRIPT_COMPAT option is set at compile time, a back reference to an unset value matches an empty string. 🫠
Because there may be many capturing parentheses in a pattern, all digits following a backslash are taken as part of a potential back reference number. If the pattern continues with a digit character, some delimiter must be used to terminate the back reference. If the PCRE_EXTENDED option is set, this can be white space. Otherwise, the \g{ syntax or an empty comment (see "Comments" below) can be used.
🔁 Recursive back references 🔁
A back reference that occurs inside the parentheses to which it refers fails when the subpattern is first used, so, for example, (a\1) never matches. However, such references can be useful inside repeated subpatterns. For example, the pattern
(a|b\1)+matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of the subpattern, the back reference matches the character string corresponding to the previous iteration. For this to work, the pattern must be such that the first iteration does not need to match the back reference. You can do that using alternation, as in the example above, or by a quantifier with a minimum of zero. 🦋
Heads up bestie: back references of this type cause the group that they reference to be treated as an atomic group. Once the whole group has been matched, a subsequent matching failure cannot cause backtracking into the middle of the group. 💀
🎀 ASSERTIONS 🎀
okay bestie, an assertion is a lil test on the characters after or before the current matching point that literally does not consume any characters. it's just vibing there, checking the vibe, then stepping aside. the simple ones — \b, \B, \A, \G, \Z, \z, ^ and $ — we already covered up above 💖
the more complicated assertions are coded as subpatterns. there's two flavors:
- ones that look ahead of the current position in the subject string
- ones that look behind it
an assertion subpattern gets matched the normal way, except for one thing: it does not change the current matching position. no cap.
assertion subpatterns are not capturing subpatterns. but if an assertion has capturing subpatterns inside it, those still count for the purposes of numbering the capturing subpatterns in the whole pattern. however, substring capturing only actually happens for positive assertions. (Perl sometimes, but not always, does do capturing in negative assertions — so don't @ us.) 🙈
WARNING: if a positive assertion containing one or more capturing subpatterns succeeds, but a failure to match later in the pattern causes backtracking over this assertion, the captures within the assertion are reset only if no higher numbered captures are already set. this is, unfortunately, a fundamental limitation of the current implementation, and since PCRE1 is now in maintenance-only status, it is unlikely ever to change. 💀 iconic behavior, but not in a good way.
for compatibility with Perl, assertion subpatterns may be repeated; even though it makes zero sense to assert the same thing several times, the side effect of capturing parentheses may occasionally be useful. in practice, there's only three cases:
- if the quantifier is
{0}, the assertion is never obeyed during matching. however, it may contain internal capturing parenthesized groups that are called from elsewhere via the subroutine mechanism. sneaky. - if the quantifier is
{0,n}where n is greater than zero, it's treated as if it were{0,1}. at run time, the rest of the pattern match is tried with and without the assertion, the order depending on the greediness of the quantifier. - if the minimum repetition is greater than zero, the quantifier is ignored. the assertion is obeyed just once when encountered during matching. periodt.
👀 Lookahead assertions
lookahead assertions start with (?= for positive assertions and (?! for negative assertions. for example,
\w+(?=;)matches a word followed by a semicolon, but does not include the semicolon in the match ✨ and
foo(?!bar)matches any occurrence of "foo" that is not followed by "bar". but bestie, watch out — the apparently similar pattern
(?!foo)bardoes not find an occurrence of "bar" that is preceded by something other than "foo". it finds any occurrence of "bar" whatsoever, because the assertion (?!foo) is always true when the next three characters are "bar". 😭 if you want the other effect, you need a lookbehind assertion. it's giving "not me thinking these were the same thing".
if you want to force a matching failure at some point in a pattern, the most convenient way to do it is with (?!) because an empty string always matches, so an assertion that requires there not to be an empty string must always fail. slay. the backtracking control verb (*FAIL) or (*F) is a synonym for (?!).
🔙 Lookbehind assertions
lookbehind assertions start with (?<= for positive assertions and (?<! for negative assertions. for example,
(?<!foo)bardoes find an occurrence of "bar" that is not preceded by "foo" 👑
the contents of a lookbehind assertion are restricted such that all the strings it matches must have a fixed length. however, if there are several top-level alternatives, they do not all have to have the same fixed length. so this:
(?<=bullock|donkey)is permitted, but this:
(?<!dogs?|cats?)causes an error at compile time. 💀 branches that match different length strings are permitted only at the top level of a lookbehind assertion. this is an extension compared with Perl, which requires all branches to match the same length of string. so
(?<=ab(c|de))is not permitted, because its single top-level branch can match two different lengths — but it is acceptable to PCRE if rewritten to use two top-level branches:
(?<=abc|abde)in some cases, the escape sequence \K (see above) can be used instead of a lookbehind assertion to get round the fixed-length restriction. workaround queen 🎀
the implementation of lookbehind assertions is, for each alternative, to temporarily move the current position back by the fixed length and then try to match. if there are insufficient characters before the current position, the assertion fails. simple as.
in a UTF mode, PCRE does not allow the \C escape (which matches a single data unit even in a UTF mode) to appear in lookbehind assertions, because it makes it impossible to calculate the length of the lookbehind. the \X and \R escapes, which can match different numbers of data units, are also not permitted. sorry babe, the math has to math. 🧮
"subroutine" calls (see below) such as (?2) or (?&X) are permitted in lookbehinds, as long as the subpattern matches a fixed-length string. recursion, however, is not supported. nope nope nope.
possessive quantifiers can be used in conjunction with lookbehind assertions to specify efficient matching of fixed-length strings at the end of subject strings. consider a simple pattern such as
abcd$when applied to a long string that does not match. because matching proceeds from left to right, PCRE will look for each "a" in the subject and then see if what follows matches the rest of the pattern. if the pattern is specified as
^.*abcd$the initial .* matches the entire string at first, but when this fails (because there is no following "a"), it backtracks to match all but the last character, then all but the last two characters, and so on. once again the search for "a" covers the entire string, from right to left, so we are no better off. 😭 however, if the pattern is written as
^.*+(?<=abcd)there can be no backtracking for the .*+ item; it can match only the entire string. the subsequent lookbehind assertion does a single test on the last four characters. if it fails, the match fails immediately. for long strings, this approach makes a significant difference to the processing time. performance queen, we stan 🌈
🌸 Using multiple assertions
several assertions (of any sort) may occur in succession. for example,
(?<=\d{3})(?<!999)foomatches "foo" preceded by three digits that are not "999". notice that each of the assertions is applied independently at the same point in the subject string. first there's a check that the previous three characters are all digits, and then there's a check that the same three characters are not "999". this pattern does not match "foo" preceded by six characters, the first of which are digits and the last three of which are not "999". for example, it doesn't match "123abcfoo". a pattern to do that is
(?<=\d{3}...)(?<!999)foothis time the first assertion looks at the preceding six characters, checking that the first three are digits, and then the second assertion checks that the preceding three characters are not "999". teamwork makes the dream work 💖
assertions can be nested in any combination. for example,
(?<=(?<!foo)bar)bazmatches an occurrence of "baz" that is preceded by "bar" which in turn is not preceded by "foo", while
(?<=\d{3}(?!999)...)foois another pattern that matches "foo" preceded by three digits and any three characters that are not "999". nesting like russian dolls but make it regex 🪆✨
🦋 CONDITIONAL SUBPATTERNS 🦋
okay bestie, this is where patterns get choices. you can make the matching process obey a subpattern conditionally, or pick between two alternative subpatterns, depending on the result of an assertion — or on whether some specific capturing subpattern has already matched. 💅
the two possible forms of conditional subpattern are:
(?(condition)yes-pattern)
(?(condition)yes-pattern|no-pattern)if the condition is satisfied, the yes-pattern is used; otherwise the no-pattern (if present) is used. and if there are more than two alternatives in the subpattern, a compile-time error occurs. periodt. 😤
each of the two alternatives may itself contain nested subpatterns of any form, including conditional subpatterns — the restriction to two alternatives applies only at the level of the condition. here's an example fragment where the alternatives are complex:
(?(1) (A|B|C) | (D | (?(2)E|F) | E) )there are four kinds of condition: references to subpatterns, references to recursion, a pseudo-condition called DEFINE, and assertions. 🎀
✨ Checking for a used subpattern by number ✨
if the text between the parentheses consists of a sequence of digits, the condition is true if a capturing subpattern of that number has previously matched. and if there's more than one capturing subpattern with the same number (see the earlier section about duplicate subpattern numbers 👀), the condition is true if any of them have matched.
you can also precede the digits with a plus or minus sign — then the subpattern number is relative rather than absolute. the most recently opened parentheses can be referenced by (?(-1), the next most recent by (?(-2), and so on. and inside loops it can also make sense to refer to subsequent groups: the next parentheses to be opened can be referenced as (?(+1), and so on. (the value zero in any of these forms is not used; it provokes a compile-time error. 🚫)
consider the following pattern, which contains non-significant white space to make it more readable (assume the PCRE_EXTENDED option) and to divide it into three parts for ease of discussion:
( \( )? [^()]+ (?(1) \) )the first part matches an optional opening parenthesis, and if that character is present, sets it as the first captured substring. the second part matches one or more characters that are not parentheses. the third part is a conditional subpattern that tests whether or not the first set of parentheses matched. if they did — that is, if the subject started with an opening parenthesis — the condition is true, and so the yes-pattern is executed and a closing parenthesis is required. otherwise, since no-pattern is not present, the subpattern matches nothing. in other words, this pattern matches a sequence of non-parentheses, optionally enclosed in parentheses. 🫶
if you were embedding this pattern in a larger one, you could use a relative reference:
...other stuff... ( \( )? [^()]+ (?(-1) \) ) ...this makes the fragment independent of the parentheses in the larger pattern. lowkey genius tbh. 💖
🌸 Checking for a used subpattern by name 🌸
Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a used subpattern by name. for compatibility with earlier versions of PCRE, which had this facility before Perl did, the syntax (?(name)...) is also recognized.
rewriting the above example to use a named subpattern gives this:
(?<OPEN> \( )? [^()]+ (?(<OPEN>) \) )and if the name used in a condition of this kind is a duplicate, the test is applied to all subpatterns of the same name, and is true if any one of them has matched. 👑
🧚 Checking for pattern recursion 🧚
if the condition is the string (R), and there is no subpattern with the name R, the condition is true if a recursive call to the whole pattern or any subpattern has been made. and if digits or a name preceded by ampersand follow the letter R, for example:
(?(R3)...) or (?(R&name)...)then the condition is true if the most recent recursion is into a subpattern whose number or name is given. note: this condition does not check the entire recursion stack, bestie. 🙈 and if the name used in a condition of this kind is a duplicate, the test is applied to all subpatterns of the same name, and is true if any one of them is the most recent recursion.
at "top level", all these recursion test conditions are false. the syntax for recursive patterns is described below. (iykyk, we'll get there 💅)
💜 Defining subpatterns for use by reference only 💜
if the condition is the string (DEFINE), and there is no subpattern with the name DEFINE, the condition is always false. in this case, there may be only one alternative in the subpattern. it is always skipped if control reaches this point in the pattern. the idea of DEFINE is that it can be used to define subroutines that can be referenced from elsewhere. (the use of subroutines is described below.)
for example, a pattern to match an IPv4 address such as "192.168.23.245" could be written like this (ignore white space and line breaks):
(?(DEFINE) (?<byte> 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
\b (?&byte) (\.(?&byte)){3} \bthe first part of the pattern is a DEFINE group inside which another group named "byte" is defined. this matches an individual component of an IPv4 address (a number less than 256). when matching takes place, this part of the pattern is skipped because DEFINE acts like a false condition. the rest of the pattern uses references to the named group to match the four dot-separated components of an IPv4 address, insisting on a word boundary at each end. slay. 🌈
⭐ Assertion conditions ⭐
if the condition is not in any of the above formats, it must be an assertion. this may be a positive or negative lookahead or lookbehind assertion. consider this pattern, again containing non-significant white space, and with the two alternatives on the second line:
(?(?=[^a-z]*[a-z])
\d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )the condition is a positive lookahead assertion that matches an optional sequence of non-letters followed by a letter. in other words, it tests for the presence of at least one letter in the subject. if a letter is found, the subject is matched against the first alternative; otherwise it is matched against the second. this pattern matches strings in one of the two forms dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits. 🎀
🎀 COMMENTS 🎀
there are two ways of including comments in patterns that are processed by PCRE. in both cases, the start of the comment must not be in a character class, nor in the middle of any other sequence of related characters such as (?: or a subpattern name or number. 🚫 and the characters that make up a comment play no part in the pattern matching.
- the sequence
(?#marks the start of a comment that continues up to the next closing parenthesis. nested parentheses are not permitted. - if the
PCRE_EXTENDEDoption is set, an unescaped#character also introduces a comment, which in this case continues to immediately after the next newline character or character sequence in the pattern.
which characters are interpreted as newlines is controlled by the options passed to a compiling function or by a special sequence at the start of the pattern, as described in the section entitled "Newline conventions" above. 👀
note that the end of this type of comment is a literal newline sequence in the pattern — escape sequences that happen to represent a newline do not count. 💀
for example, consider this pattern when PCRE_EXTENDED is set, and the default newline convention is in force:
abc #comment \n still commenton encountering the # character, pcre_compile() skips along, looking for a newline in the pattern. the sequence \n is still literal at this stage, so it does not terminate the comment. only an actual character with the code value 0x0a (the default newline) does so. and that's just how it is, bestie. ✨
✨ RECURSIVE PATTERNS ✨
okay bestie, imagine you gotta match a string that's wrapped in parentheses, and the parentheses can nest infinitely deep. like, no ceiling. without recursion, the absolute best you can do is a pattern that matches up to some fixed depth of nesting. it is literally not possible to handle an arbitrary nesting depth. that's the whole vibe. 💀
for a while now, Perl has had a thing that lets regular expressions recurse (amongst other things). it does this by interpolating Perl code in the expression at run time, and that code can refer to the expression itself. a Perl pattern using code interpolation to solve the parentheses problem looks like this:
$re = qr{\( (?: (?>[^()]+) | (?p{$re}) )* \)}x;the (?p{...}) item interpolates Perl code at run time, and in this case it refers recursively to the pattern it appears in. iconic behaviour, honestly.
obviously, PCRE cannot support the interpolation of Perl code. instead, it supports special syntax for recursion of the entire pattern, and also for individual subpattern recursion. after its introduction in PCRE and Python, this kind of recursion was subsequently introduced into Perl at release 5.10. so like, PCRE was early. trendsetter energy. 👑
so here's the deal: a special item that consists of (? followed by a number greater than zero and a closing parenthesis is a recursive subroutine call of the subpattern of the given number — provided that it occurs inside that subpattern. (if it doesn't, it's a non-recursive subroutine call, which is described in the next section, so hold that thought.) and the special item (?R) or (?0) is a recursive call of the entire regular expression. 🌀
this PCRE pattern solves the nested parentheses problem (assume the PCRE_EXTENDED option is set so that white space is ignored):
\( ( [^()]++ | (?R) )* \)so first it matches an opening parenthesis. then it matches any number of substrings which can either be a sequence of non-parentheses, or a recursive match of the pattern itself (that is, a correctly parenthesized substring). finally there's a closing parenthesis. and note the use of a possessive quantifier to avoid backtracking into sequences of non-parentheses. that part is not optional, bestie, that's the whole trick. 💅
if this were part of a larger pattern, you would not want to recurse the entire pattern, so instead you could use this:
( \( ( [^()]++ | (?1) )* \) )we put the pattern into parentheses, and caused the recursion to refer to them instead of the whole pattern. chef's kiss. 🎀
in a larger pattern, keeping track of parenthesis numbers can be tricky. this is made easier by the use of relative references. instead of (?1) in the pattern above you can write (?-2) to refer to the second most recently opened parentheses preceding the recursion. in other words, a negative number counts capturing parentheses leftwards from the point at which it is encountered. counting backwards, but make it regex. 🧚
it is also possible to refer to subsequently opened parentheses, by writing references such as (?+2). however, these cannot be recursive, because the reference is not inside the parentheses that are referenced. they are always non-recursive subroutine calls, as described in the next section. no recursion for you, forward references. 🙈
an alternative approach is to use named parentheses instead. the Perl syntax for this is (?&name); PCRE's earlier syntax (?P>name) is also supported. we could rewrite the above example as follows:
(?<pn> \( ( [^()]++ | (?&pn) )* \) )if there is more than one subpattern with the same name, the earliest one is used. first come, first served, periodt.
now, this particular example pattern that we've been looking at contains nested unlimited repeats, and so the use of a possessive quantifier for matching strings of non-parentheses is important when applying the pattern to strings that do not match. for example, when this pattern is applied to
(aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()it yields "no match" quickly. however, if a possessive quantifier is not used, the match runs for a very long time indeed, because there are so many different ways the + and * repeats can carve up the subject, and all of them have to be tested before failure can be reported. like, go get a snack, go touch grass, it'll still be running. 😭
at the end of a match, the values of capturing parentheses are those from the outermost level. if you want to obtain intermediate values, a callout function can be used (see below and the pcrecallout documentation). if the pattern above is matched against
(ab(cd)ef)the value for the inner capturing parentheses (numbered 2) is "ef", which is the last value taken on at the top level. if a capturing subpattern is not matched at the top level, its final captured value is unset, even if it was (temporarily) set at a deeper level during the matching process. the deepest drama doesn't get the final say. 💅
if there are more than 15 capturing parentheses in a pattern, PCRE has to obtain extra memory to store data during a recursion, which it does by using pcre_malloc, freeing it via pcre_free afterwards. if no memory can be obtained, the match fails with the PCRE_ERROR_NOMEMORY error. 🫶
do not confuse the(?R)item with the condition(R), which tests for recursion. they are two different girls, they are not the same girl.
consider this pattern, which matches text in angle brackets, allowing for arbitrary nesting. only digits are allowed in nested brackets (that is, when recursing), whereas any characters are permitted at the outer level:
< (?: (?(R) \d++ | [^<>]*+) | (?R)) * >in this pattern, (?(R) is the start of a conditional subpattern, with two different alternatives for the recursive and non-recursive cases. the (?R) item is the actual recursive call. conditional, then recursive. two separate jobs, bestie, don't merge them. 🌈
🦋 Differences in recursion processing between PCRE and Perl 🦋
recursion processing in PCRE differs from Perl in two important ways. let's get into it.
way number one: in PCRE (like Python, but unlike Perl), a recursive subpattern call is always treated as an atomic group. that is, once it has matched some of the subject string, it is never re-entered, even if it contains untried alternatives and there is a subsequent matching failure. it said what it said and it's not coming back. 💀
this can be illustrated by the following pattern, which purports to match a palindromic string that contains an odd number of characters (for example, "a", "aba", "abcba", "abcdcba"):
^(.|(.)(?1)\2)$the idea is that it either matches a single character, or two identical characters surrounding a sub-palindrome. in Perl, this pattern works; in PCRE it does not if the pattern is longer than three characters. consider the subject string "abcba":
at the top level, the first character is matched, but as it is not at the end of the string, the first alternative fails; the second alternative is taken and the recursion kicks in. the recursive call to subpattern 1 successfully matches the next character ("b"). (note that the beginning and end of line tests are not part of the recursion.) 🌀
back at the top level, the next character ("c") is compared with what subpattern 2 matched, which was "a". this fails. because the recursion is treated as an atomic group, there are now no backtracking points, and so the entire match fails. (Perl is able, at this point, to re-enter the recursion and try the second alternative.) however, if the pattern is written with the alternatives in the other order, things are different:
^((.)(?1)\2|.)$this time, the recursing alternative is tried first, and continues to recurse until it runs out of characters, at which point the recursion fails. but this time we do have another alternative to try at the higher level. that is the big difference: in the previous case the remaining alternative is at a deeper recursion level, which PCRE cannot use. 🙈
to change the pattern so that it matches all palindromic strings, not just those with an odd number of characters, it is tempting to change the pattern to this:
^((.)(?1)\2|.?)$again, this works in Perl, but not in PCRE, and for the same reason. when a deeper recursion has matched a single character, it cannot be entered again in order to match an empty string. the solution is to separate the two cases, and write out the odd and even cases as alternatives at the higher level:
^(?:((.)(?1)\2|)|((.)(?3)\4|.))if you want to match typical palindromic phrases, the pattern has to ignore all non-word characters, which can be done like this:
^\W*+(?:((.)\W*+(?1)\W*+\2|)|((.)\W*+(?3)\W*+\4|\W*+.\W*+))\W*+$if run with the PCRE_CASELESS option, this pattern matches phrases such as "A man, a plan, a canal: Panama!" and it works well in both PCRE and Perl. note the use of the possessive quantifier *+ to avoid backtracking into sequences of non-word characters. without this, PCRE takes a great deal longer (ten times or more) to match typical phrases, and Perl takes so long that you think it has gone into a loop. literally just sitting there. staring at you. 😭
WARNING: the palindrome-matching patterns above work only if the subject string does not start with a palindrome that is shorter than the entire string. for example, although"abcba"is correctly matched, if the subject is"ababa", PCRE finds the palindrome"aba"at the start, then fails at top level because the end of the string does not follow. once again, it cannot jump back into the recursion to try other alternatives, so the entire match fails.
way number two: the second way in which PCRE and Perl differ in their recursion processing is in the handling of captured values. in Perl, when a subpattern is called recursively or as a subpattern (see the next section), it has no access to any values that were captured outside the recursion, whereas in PCRE these values can be referenced. consider this pattern:
^(.)(\1|a(?2))in PCRE, this pattern matches "bab". the first capturing parentheses match "b", then in the second group, when the back reference \1 fails to match "b", the second alternative matches "a" and then recurses. in the recursion, \1 does now match "b" and so the whole match succeeds. in Perl, the pattern fails to match because inside the recursive call \1 cannot access the externally set value. different house rules, different results. 💜
💖 SUBPATTERNS AS SUBROUTINES 💖
okay bestie so you know how we just did recursive calls? plot twist: if you use that same recursive-call syntax (whether you're calling by number or by name) outside the parentheses it's pointing at, it doesn't recurse at all — it acts like a subroutine in a normal programming language. wild, right? 🦋
the called subpattern can be defined before or after the reference, so no "you have to declare it first" nonsense. and a numbered reference can be absolute or relative, like this:
(...(absolute)...)...(?2)...
(...(relative)...)...(?-1)...
(...(?+1)...(relative)...remember earlier when we clocked that (sens|respons)e and \1ibility matches "sense and sensibility" and "response and responsibility", but not "sense and responsibility"? well. if you use this instead:
(sens|respons)e and (?1)ibilityit does match "sense and responsibility" too, plus the other two strings. slay. there's another example chilling in the DEFINE discussion above if you wanna peek. 👑
real talk though: all subroutine calls, recursive or not, are always treated as atomic groups. so once a subroutine has matched some of the subject string, it is never re-entered — even if it still has untried alternatives sitting there and there's a matching failure later. and any capturing parentheses that get set during the subroutine call revert to their previous values afterwards. no take-backs, no redos. 💀
also, processing options like case-independence are fixed when a subpattern is defined. so if that subpattern gets used as a subroutine, those options can't be changed for different calls. like, consider:
(abc)(?i:(?-1))it matches "abcabc". it does not match "abcABC", because the change of processing option does not affect the called subpattern. the vibes are locked in at definition time, periodt. 🔒
🌸 ONIGURUMA SUBROUTINE SYNTAX 🌸
for compatibility with Oniguruma, there's a non-Perl syntax: \g followed by a name or a number enclosed either in angle brackets or single quotes. it's an alternative syntax for referencing a subpattern as a subroutine, possibly recursively. here are two of the examples from up above, rewritten with this syntax:
(?<pn> \( ( (?>[^()]+) | \g<pn> )* \) )
(sens|respons)e and \g'1'ibilityand PCRE supports an extension to Oniguruma: if a number is preceded by a plus or a minus sign, it's taken as a relative reference. for example:
(abc)(?i:\g<-1>)do NOT mix these up, bestie:\g{...}(Perl syntax) and\g<...>(Oniguruma syntax) are not synonymous. the former is a back reference; the latter is a subroutine call. completely different girlies. 🙈
✨ CALLOUTS ✨
so Perl has this feature where using the sequence (?{...}) causes arbitrary Perl code to be obeyed in the middle of matching a regular expression. among other things, this makes it possible to extract different substrings that match the same pair of parentheses when there's a repetition. iconic behavior honestly.
PCRE provides a similar feature, but obviously it cannot obey arbitrary Perl code (it's not gonna run your whole script mid-match, babe). this feature is called a "callout". the caller of PCRE provides an external function by putting its entry point in the global variable pcre_callout (8-bit library) or pcre[16|32]_callout (16-bit or 32-bit library). by default, this variable contains NULL, which disables all calling out. 🚫
within a regular expression, (?C) indicates the points at which the external function is to be called. if you want to identify different callout points, you can put a number less than 256 after the letter C. the default value is zero. for example, this pattern has two callout points:
(?C1)abc(?C2)defif the PCRE_AUTO_CALLOUT flag is passed to a compiling function, callouts are automatically installed before each item in the pattern. they are all numbered 255. and if there's a conditional group in the pattern whose condition is an assertion, an additional callout is inserted just before the condition. an explicit callout may also be set at this position, like this:
(?(?C9)(?=a)abc|def)note that this applies only to assertion conditions, not to other types of condition. read that again. 📝
during matching, when PCRE reaches a callout point, the external function is called. it's provided with the number of the callout, the position in the pattern, and, optionally, one item of data originally supplied by the caller of the matching function. the callout function may cause matching to proceed, to backtrack, or to fail altogether. main character energy, it does whatever it wants. 💅
by default, PCRE implements a number of optimizations at compile time and matching time, and one side-effect is that sometimes callouts get skipped. if you need all possible callouts to happen, you need to set options that disable the relevant optimizations. more details, plus a complete description of the interface to the callout function, are given in the pcrecallout documentation. 📚
🔥 BACKTRACKING CONTROL 🔥
okay so. Perl 5.10 introduced a bunch of "Special Backtracking Control Verbs", which the Perl documentation still describes as "experimental and subject to change or removal in a future version of Perl". it then goes on to say: "Their usage in production code should be noted to avoid problems during upgrades."
heads up bestie: the same remarks apply to the PCRE features described in this section. this is experimental territory. tread carefully. ⚠️
the new verbs make use of what was previously invalid syntax: an opening parenthesis followed by an asterisk. they're generally of the form (*VERB) or (*VERB:NAME). some may take either form, possibly behaving differently depending on whether or not a name is present.
a name is any sequence of characters that does not include a closing parenthesis. the maximum length of a name is 255 in the 8-bit library and 65535 in the 16-bit and 32-bit libraries. if the name is empty — that is, if the closing parenthesis immediately follows the colon — the effect is as if the colon were not there. any number of these verbs may occur in a pattern. go wild. 🌈
since these verbs are specifically related to backtracking, most of them can be used only when the pattern is to be matched using one of the traditional matching functions, because those use a backtracking algorithm. with the exception of (*FAIL), which behaves like a failing negative assertion, the backtracking control verbs cause an error if encountered by a DFA matching function. 💥
the behaviour of these verbs in repeated groups, assertions, and in subpatterns called as subroutines (whether or not recursively) is documented below. 👇
💫 Optimizations that affect backtracking verbs 💫
PCRE contains some optimizations that are used to speed up matching by running some checks at the start of each match attempt. for example, it may know the minimum length of a matching subject, or that a particular character must be present. when one of these optimizations bypasses the running of a match, any included backtracking verbs will, of course, not be processed. 😭
you can suppress the start-of-match optimizations by setting the PCRE_NO_START_OPTIMIZE option when calling pcre_compile() or pcre_exec(), or by starting the pattern with (*NO_START_OPT). there's more discussion of this option in the section entitled "Option bits for pcre_exec()" in the pcreapi documentation.
experiments with Perl suggest that it too has similar optimizations, sometimes leading to anomalous results. so it's not just us, king. 🤷♀️
⭐ Verbs that act immediately ⭐
the following verbs act as soon as they are encountered. they may not be followed by a name.
(*ACCEPT)— this verb causes the match to end successfully, skipping the remainder of the pattern. however, when it's inside a subpattern that is called as a subroutine, only that subpattern is ended successfully, and matching then continues at the outer level. if(*ACCEPT)is triggered in a positive assertion, the assertion succeeds; in a negative assertion, the assertion fails. 🎯
if (*ACCEPT) is inside capturing parentheses, the data so far is captured. for example:
A((?:A|B(*ACCEPT)|C)D)this matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is captured by the outer parentheses. cute. 🎀
(*FAIL)or(*F)— this verb causes a matching failure, forcing backtracking to occur. it is equivalent to(?!)but easier to read. the Perl documentation notes that it is probably useful only when combined with(?{})or(??{})— those are, of course, Perl features that are not present in PCRE. the nearest equivalent is the callout feature, as for example in this pattern:
a+(?C)(*FAIL)a match with the string "aaaa" always fails, but the callout is taken before each backtrack happens (in this example, 10 times). 💀
🎀 Recording which path was taken 🎀
there is one verb whose main purpose is to track how a match was arrived at, though it also has a secondary use in conjunction with advancing the match starting point (see (*SKIP) below).
(*MARK:NAME)or(*:NAME)— a name is always required with this verb. there may be as many instances of(*MARK)as you like in a pattern, and their names do not have to be unique. 🙌
when a match succeeds, the name of the last-encountered (*MARK:NAME), (*PRUNE:NAME), or (*THEN:NAME) on the matching path is passed back to the caller as described in the section entitled "Extra data for pcre_exec()" in the pcreapi documentation. here's an example of pcretest output, where the /K modifier requests the retrieval and outputting of (*MARK) data:
re> /X(*MARK:A)Y|X(*MARK:B)Z/K
data> XY
0: XY
MK: A
XZ
0: XZ
MK: Bthe (*MARK) name is tagged with "MK:" in this output, and in this example it indicates which of the two alternatives matched. this is a more efficient way of obtaining this information than putting each alternative in its own capturing parentheses. efficiency queen. 👑
if a verb with a name is encountered in a positive assertion that is true, the name is recorded and passed back if it is the last-encountered. this does not happen for negative assertions or failing positive assertions. noted. 📌
after a partial match or a failed match, the last encountered name in the entire match process is returned. for example:
re> /X(*MARK:A)Y|X(*MARK:B)Z/K
data> XP
No match, mark = Bnote that in this unanchored example the mark is retained from the match attempt that started at the letter "X" in the subject. subsequent match attempts starting at "P" and then with an empty string do not get as far as the (*MARK) item, but nevertheless do not reset it. 🫶
if you're interested in (*MARK) values after failed matches, you should probably set the PCRE_NO_START_OPTIMIZE option (see above) to ensure that the match is always attempted. don't get caught slipping. 💅
🌈 Verbs that act after backtracking 🌈
The following verbs do nothing when they're just, like, sitting there chilling. Matching keeps going with whatever comes next, but if there's no subsequent match and it causes a backtrack to the verb? Boom, failure is forced. Which means backtracking literally cannot pass to the left of the verb. 💀
But wait — when one of these verbs is vibing inside an atomic group or an assertion that is true, its effect is confined to that group, because once the group has matched, there's never any backtracking back into it. In that situation, backtracking can "jump back" to the left of the entire atomic group or assertion. (Also remember, like we said earlier, this localization thing also applies in subroutine calls, bestie.)
These verbs differ in exactly what kind of failure happens when backtracking reaches them. The behaviour below is what goes down when the verb is not in a subroutine or an assertion. The later sections cover those special cases. 🎀
(*COMMIT)
This verb, which may not be followed by a name, causes the whole match to fail outright if there's a later matching failure that causes backtracking to reach it. Even if the pattern is unanchored, no further attempts to find a match by advancing the starting point take place. If (*COMMIT) is the only backtracking verb that's encountered, once it's been passed pcre_exec() is committed to finding a match at the current starting point, or not at all. No in-between, periodt. For example:
a+(*COMMIT)bThis matches "xxaab" but not "aacaab". It can be thought of as a kind of dynamic anchor, or "I've started, so I must finish." The name of the most recently passed (*MARK) in the path is passed back when (*COMMIT) forces a match failure.
If there's more than one backtracking verb in a pattern, a different one that follows (*COMMIT) may be triggered first, so merely passing (*COMMIT) during a match does not always guarantee that a match must be at this starting point. Lowkey sneaky, but that's how it is.
Note that (*COMMIT) at the start of a pattern is not the same as an anchor, unless PCRE's start-of-match optimizations are turned off, as shown in this output from pcretest:
re> /(*COMMIT)abc/
data> xyzabc
0: abc
data> xyzabc\Y
No matchFor this pattern, PCRE knows that any match must start with "a", so the optimization skips along the subject to "a" before applying the pattern to the first set of data. The match attempt then succeeds. In the second set of data, the escape sequence \Y is interpreted by the pcretest program. It causes the PCRE_NO_START_OPTIMIZE option to be set when pcre_exec() is called. This disables the optimization that skips along to the first character. The pattern is now applied starting at "x", and so the (*COMMIT) causes the match to fail without trying any other starting points. 🦋
(*PRUNE) or (*PRUNE:NAME)
This verb causes the match to fail at the current starting position in the subject if there's a later matching failure that causes backtracking to reach it. If the pattern is unanchored, the normal "bumpalong" advance to the next starting character then happens. Backtracking can occur as usual to the left of (*PRUNE), before it's reached, or when matching to the right of (*PRUNE), but if there's no match to the right, backtracking cannot cross (*PRUNE). In simple cases, using (*PRUNE) is just an alternative to an atomic group or possessive quantifier, but there are some uses of (*PRUNE) that can't be expressed in any other way. In an anchored pattern (*PRUNE) has the same effect as (*COMMIT).
The behaviour of (*PRUNE:NAME) is the not the same as (*MARK:NAME)(*PRUNE). It is like (*MARK:NAME) in that the name is remembered for passing back to the caller. However, (*SKIP:NAME) searches only for names set with (*MARK). 👑
(*SKIP)
This verb, when given without a name, is like (*PRUNE), except that if the pattern is unanchored, the "bumpalong" advance is not to the next character, but to the position in the subject where (*SKIP) was encountered. (*SKIP) signifies that whatever text was matched leading up to it cannot be part of a successful match. Consider:
a+(*SKIP)bIf the subject is "aaaac...", after the first match attempt fails (starting at the first character in the string), the starting point skips on to start the next attempt at "c". Note that a possessive quantifier does not have the same effect as this example; although it would suppress backtracking during the first match attempt, the second attempt would start at the second character instead of skipping on to "c". Not the same vibe at all. 💅
(*SKIP:NAME)
When (*SKIP) has an associated name, its behaviour is modified. When it's triggered, the previous path through the pattern is searched for the most recent (*MARK) that has the same name. If one is found, the "bumpalong" advance is to the subject position that corresponds to that (*MARK) instead of to where (*SKIP) was encountered. If no (*MARK) with a matching name is found, the (*SKIP) is ignored.
Note that (*SKIP:NAME) searches only for names set by (*MARK:NAME). It ignores names that are set by (*PRUNE:NAME) or (*THEN:NAME). Rude, but that's the rule. 🙈
(*THEN) or (*THEN:NAME)
This verb causes a skip to the next innermost alternative when backtracking reaches it. That is, it cancels any further backtracking within the current alternative. Its name comes from the observation that it can be used for a pattern-based if-then-else block:
( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...If the COND1 pattern matches, FOO is tried (and possibly further items after the end of the group if FOO succeeds); on failure, the matcher skips to the second alternative and tries COND2, without backtracking into COND1. If that succeeds and BAR fails, COND3 is tried. If subsequently BAZ fails, there are no more alternatives, so there's a backtrack to whatever came before the entire group. If (*THEN) is not inside an alternation, it acts like (*PRUNE). ✨
The behaviour of (*THEN:NAME) is the not the same as (*MARK:NAME)(*THEN). It is like (*MARK:NAME) in that the name is remembered for passing back to the caller. However, (*SKIP:NAME) searches only for names set with (*MARK).
A subpattern that does not contain a | character is just a part of the enclosing alternative; it is not a nested alternation with only one alternative. The effect of (*THEN) extends beyond such a subpattern to the enclosing alternative. Consider this pattern, where A, B, etc. are complex pattern fragments that do not contain any | characters at this level:
A (B(*THEN)C) | DIf A and B are matched, but there's a failure in C, matching does not backtrack into A; instead it moves to the next alternative, that is, D. However, if the subpattern containing (*THEN) is given an alternative, it behaves differently:
A (B(*THEN)C | (*FAIL)) | DThe effect of (*THEN) is now confined to the inner subpattern. After a failure in C, matching moves to (*FAIL), which causes the whole subpattern to fail because there are no more alternatives to try. In this case, matching does now backtrack into A. 🌸
Note that a conditional subpattern is not considered as having two alternatives, because only one is ever used. In other words, the | character in a conditional subpattern has a different meaning. Ignoring white space, consider:
^.*? (?(?=a) a | b(*THEN)c )If the subject is "ba", this pattern does not match. Because .*? is ungreedy, it initially matches zero characters. The condition (?=a) then fails, the character "b" is matched, but "c" is not. At this point, matching does not backtrack to .*? as might perhaps be expected from the presence of the | character. The conditional subpattern is part of the single alternative that comprises the whole pattern, and so the match fails. (If there was a backtrack into .*?, allowing it to match "b", the match would succeed.) 😭
The verbs just described provide four different "strengths" of control when subsequent matching fails:
(*THEN)is the weakest, carrying on the match at the next alternative.(*PRUNE)comes next, failing the match at the current starting position, but allowing an advance to the next character (for an unanchored pattern).(*SKIP)is similar, except that the advance may be more than one character.(*COMMIT)is the strongest, causing the entire match to fail.
More than one backtracking verb
If more than one backtracking verb is present in a pattern, the one that is backtracked onto first acts. For example, consider this pattern, where A, B, etc. are complex pattern fragments:
(A(*COMMIT)B(*THEN)C|ABD)If A matches but B fails, the backtrack to (*COMMIT) causes the entire match to fail. However, if A and B match, but C fails, the backtrack to (*THEN) causes the next alternative (ABD) to be tried. This behaviour is consistent, but is not always the same as Perl's. It means that if two or more backtracking verbs appear in succession, all the the last of them has no effect. Consider this example:
...(*COMMIT)(*PRUNE)...If there's a matching failure to the right, backtracking onto (*PRUNE) causes it to be triggered, and its action is taken. There can never be a backtrack onto (*COMMIT). 💀
Backtracking verbs in repeated groups
PCRE differs from Perl in its handling of backtracking verbs in repeated groups. For example, consider:
/(a(*COMMIT)b)+ac/If the subject is "abac", Perl matches, but PCRE fails because the (*COMMIT) in the second repeat of the group acts. Not me picking sides but okay. 🙈
Backtracking verbs in assertions
(*FAIL) in an assertion has its normal effect: it forces an immediate backtrack.
(*ACCEPT) in a positive assertion causes the assertion to succeed without any further processing. In a negative assertion, (*ACCEPT) causes the assertion to fail without any further processing.
The other backtracking verbs are not treated specially if they appear in a positive assertion. In particular, (*THEN) skips to the next alternative in the innermost enclosing group that has alternations, whether or not this is within the assertion.
Negative assertions are, however, different, in order to ensure that changing a positive assertion into a negative assertion changes its result. Backtracking into (*COMMIT), (*SKIP), or (*PRUNE) causes a negative assertion to be true, without considering any further alternative branches in the assertion. Backtracking into (*THEN) causes it to skip to the next enclosing alternative within the assertion (the normal behaviour), but if the assertion does not have such an alternative, (*THEN) behaves like (*PRUNE). 🧚
Backtracking verbs in subroutines
These behaviours occur whether or not the subpattern is called recursively. Perl's treatment of subroutines is different in some cases.
(*FAIL) in a subpattern called as a subroutine has its normal effect: it forces an immediate backtrack.
okay so when these backtracking-control besties show up inside a subpattern that got called as a subroutine, they act a lil different, and you need to know this:
(*ACCEPT)in a subpattern called as a subroutine → the subroutine match just succeeds, no further processing, done, slay. 💅 Then matching carries on right after the subroutine call.(*COMMIT),(*SKIP), and(*PRUNE)in a subpattern called as a subroutine → they make the subroutine match fail. That's it, that's the whole vibe. 💀(*THEN)skips to the next alternative in the innermost enclosing group within the subpattern that has alternatives. And if there's, like, no such group inside the subpattern? Then(*THEN)makes the subroutine match fail. periodt. 🎀
✨ SEE ALSO ✨
go read your sisters, bestie, they're all iconic in their own way 🫶
👑 AUTHOR 👑
the literal main character who built all this:
Philip Hazel
University Computing Service
Cambridge CB2 3QH, England.🌸 REVISION 🌸
- Last updated: 23 October 2016
- Copyright (c) 1997-2016 University of Cambridge.
and that's a wrap on the whole page, bestie!! 🎉 this was PCRE 8.40, dated 23 October 2016, and the page itself is PCREPATTERN(3) — go forth and regex responsibly, mwah 💖🦋✨
📦 where this page comes from
- alpine 3.22: pcre-doc 8.45-r4 · BSD-3-Clause
- alpine 3.23: pcre-doc 8.45-r4 · BSD-3-Clause
- alpine 3.24: pcre-doc 8.45-r4 · BSD-3-Clause
- arch rolling: pcre 8.45-5 · BSD
- fedora 43: pcre-devel 8.45-1.fc43.9 · BSD-3-Clause
- fedora 43: pcre-devel 8.45-1.fc43.9 · BSD-3-Clause
- fedora 44: pcre-devel 8.45-1.fc43.9 · BSD-3-Clause
- fedora 44: pcre-devel 8.45-1.fc43.9 · BSD-3-Clause
- nixos 26.05: pcre 8.45
- nixos 26.05: pcre-cpp 8.45
- opensuse-leap 16.0: pcre-devel 8.45-160000.2.2 · BSD-3-Clause
- opensuse-leap 16.0: pcre-devel 8.45-160000.2.2 · BSD-3-Clause
- opensuse-leap 16.0: pcre-devel 8.45-160000.2.2 · BSD-3-Clause
- opensuse-leap 16.0: pcre-devel 8.45-160000.2.2 · BSD-3-Clause
- 💖 girlified by deepseek-ai/DeepSeek-V4.1-Flash on : a language model's rewrite of the original, under its terms