Rename test.pl's skip_all_without_extension to *_dynamic_extension().

[perl5.git] / pod / perlre.pod

=head1 NAME
X<regular expression> X<regex> X<regexp>

perlre - Perl regular expressions

=head1 DESCRIPTION

This page describes the syntax of regular expressions in Perl.

If you haven't used regular expressions before, a quick-start
introduction is available in L<perlrequick>, and a longer tutorial
introduction is available in L<perlretut>.

For reference on how regular expressions are used in matching
operations, plus various examples of the same, see discussions of
C<m//>, C<s///>, C<qr//> and C<??> in L<perlop/"Regexp Quote-Like
Operators">.


=head2 Modifiers

Matching operations can have various modifiers.  Modifiers
that relate to the interpretation of the regular expression inside
are listed below.  Modifiers that alter the way a regular expression
is used by Perl are detailed in L<perlop/"Regexp Quote-Like Operators"> and
L<perlop/"Gory details of parsing quoted constructs">.

=over 4

=item m
X</m> X<regex, multiline> X<regexp, multiline> X<regular expression, multiline>

Treat string as multiple lines.  That is, change "^" and "$" from matching
the start or end of the string to matching the start or end of any
line anywhere within the string.

=item s
X</s> X<regex, single-line> X<regexp, single-line>
X<regular expression, single-line>

Treat string as single line.  That is, change "." to match any character
whatsoever, even a newline, which normally it would not match.

Used together, as C</ms>, they let the "." match any character whatsoever,
while still allowing "^" and "$" to match, respectively, just after
and just before newlines within the string.

=item i
X</i> X<regex, case-insensitive> X<regexp, case-insensitive>
X<regular expression, case-insensitive>

Do case-insensitive pattern matching.

If C<use locale> is in effect, the case map is taken from the current
locale for code points less than 255, and from Unicode rules for larger
code points.  See L<perllocale>.

=item x
X</x>

Extend your pattern's legibility by permitting whitespace and comments.

=item p
X</p> X<regex, preserve> X<regexp, preserve>

Preserve the string matched such that ${^PREMATCH}, ${^MATCH}, and
${^POSTMATCH} are available for use after matching.

=item g and c
X</g> X</c>

Global matching, and keep the Current position after failed matching.
Unlike i, m, s and x, these two flags affect the way the regex is used
rather than the regex itself. See
L<perlretut/"Using regular expressions in Perl"> for further explanation
of the g and c modifiers.

=item a, d, l and u
X</a> X</d> X</l> X</u>

These modifiers, new in 5.14, affect which character-set semantics
(Unicode, ASCII, etc.) are used, as described below.

=back

These are usually written as "the C</x> modifier", even though the delimiter
in question might not really be a slash.  The modifiers C</imsxadlup>
may also be embedded within the regular expression itself using
the C<(?...)> construct.

The C</x>, C</l>, C</u>, C</a> and C</d> modifiers need a little more
explanation.

C</x> tells
the regular expression parser to ignore most whitespace that is neither
backslashed nor within a character class.  You can use this to break up
your regular expression into (slightly) more readable parts.  The C<#>
character is also treated as a metacharacter introducing a comment,
just as in ordinary Perl code.  This also means that if you want real
whitespace or C<#> characters in the pattern (outside a character
class, where they are unaffected by C</x>), then you'll either have to
escape them (using backslashes or C<\Q...\E>) or encode them using octal,
hex, or C<\N{}> escapes.  Taken together, these features go a long way towards
making Perl's regular expressions more readable.  Note that you have to
be careful not to include the pattern delimiter in the comment--perl has
no way of knowing you did not intend to close the pattern early.  See
the C-comment deletion code in L<perlop>.  Also note that anything inside
a C<\Q...\E> stays unaffected by C</x>.  And note that C</x> doesn't affect
space interpretation within a single multi-character construct.  For
example in C<\x{...}>, regardless of the C</x> modifier, there can be no
spaces.  Same for a L<quantifier|/Quantifiers> such as C<{3}> or
C<{5,}>.  Similarly, C<(?:...)> can't have a space between the C<?> and C<:>,
but can between the C<(> and C<?>.  Within any delimiters for such a
construct, allowed spaces are not affected by C</x>, and depend on the
construct.  For example, C<\x{...}> can't have spaces because hexadecimal
numbers don't have spaces in them.  But, Unicode properties can have spaces, so
in C<\p{...}> there can be spaces that follow the Unicode rules, for which see
L<perluniprops/Properties accessible through \p{} and \P{}>.
X</x>

C</l> means to use a locale (see L<perllocale>) when pattern matching.
The locale used will be the one in effect at the time of execution of
the pattern match.  This may not be the same as the compilation-time
locale, and can differ from one match to another if there is an
intervening call of the
L<setlocale() function|perllocale/The setlocale function>.
This modifier is automatically set if the regular expression is compiled
within the scope of a C<"use locale"> pragma.
Perl only allows single-byte locales.  This means that code points above
255 are treated as Unicode no matter what locale is in effect.
Under Unicode rules, there are a few case-insensitive matches that cross the
255/256 boundary.  These are disallowed.  For example,
0xFF does not caselessly match the character at 0x178, LATIN CAPITAL
LETTER Y WITH DIAERESIS, because 0xFF may not be LATIN SMALL LETTER Y
in the current locale, and Perl has no way of knowing if that character
even exists in the locale, much less what code point it is.
X</l>

C</u> means to use Unicode semantics when pattern matching.  It is
automatically set if the regular expression is encoded in utf8 internally,
or is compiled within the scope of a
L<C<"use feature 'unicode_strings">|feature> pragma (and isn't also in
the scope of the L<C<"use locale">|locale> or the L<C<"use bytes">|bytes>
pragma).  On ASCII platforms, the code points between 128 and 255 take on their
Latin-1 (ISO-8859-1) meanings (which are the same as Unicode's), whereas
in strict ASCII their meanings are undefined.  Thus the platform
effectively becomes a Unicode platform.  The ASCII characters remain as
ASCII characters (since ASCII is a subset of Latin-1 and Unicode).  For
example, when this option is not on, on a non-utf8 string, C<"\w">
matches precisely C<[A-Za-z0-9_]>.  When the option is on, it matches
not just those, but all the Latin-1 word characters (such as an "n" with
a tilde).  On EBCDIC platforms, which already are equivalent to Latin-1,
this modifier changes behavior only when the C<"/i"> modifier is also
specified, and affects only two characters, giving them full Unicode
semantics: the C<MICRO SIGN> will match the Greek capital and
small letters C<MU>; otherwise not; and the C<LATIN CAPITAL LETTER SHARP
S> will match any of C<SS>, C<Ss>, C<sS>, and C<ss>, otherwise not.
(This last case is buggy, however.)
X</u>

C</a> is the same as C</u>, except that C<\d>, C<\s>, C<\w>, and the
Posix character classes are restricted to matching in the ASCII range
only.  That is, with this modifier, C<\d> always means precisely the
digits C<"0"> to C<"9">; C<\s> means the five characters C<[ \f\n\r\t]>;
C<\w> means the 63 characters C<[A-Za-z0-9_]>; and likewise, all the
Posix classes such as C<[[:print:]]> match only the appropriate
ASCII-range characters.  As you would expect, this modifier causes, for
example, C<\D> to mean the same thing as C<[^0-9]>; in fact, all
non-ASCII characters match C<\D>, C<\S>, and C<\W>.  C<\b> still means
to match at the boundary between C<\w> and C<\W>, using the C<"a">
definitions of them (similarly for C<\B>).  Otherwise, C<"a"> behaves
like the C<"u"> modifier, in that case-insensitive matching uses Unicode
semantics; for example, "k" will match the Unicode C<\N{KELVIN SIGN}>
under C</i> matching, and code points in the Latin1 range, above ASCII
will have Unicode semantics when it comes to case-insensitive matching.
But writing two in "a"'s in a row will increase its effect, causing the
Kelvin sign and all other non-ASCII characters not to match any ASCII
character under C</i> matching.
X</a>

C</d> means to use the traditional Perl pattern-matching behavior.
This is dualistic (hence the name C</d>, which also could stand for
"depends").  When this is in effect, Perl matches according to the
platform's native character set rules unless there is something that
indicates to use Unicode rules.  If either the target string or the
pattern itself is encoded in UTF-8, Unicode rules are used.  Also, if
the pattern contains Unicode-only features, such as code points above
255, C<\p()> Unicode properties or C<\N{}> Unicode names, Unicode rules
will be used.  It is automatically selected by default if the regular
expression is compiled neither within the scope of a C<"use locale">
pragma nor a <C<"use feature 'unicode_strings"> pragma.
This behavior causes a number of glitches, see
L<perlunicode/The "Unicode Bug">.
X</d>

=head2 Regular Expressions

=head3 Metacharacters

The patterns used in Perl pattern matching evolved from those supplied in
the Version 8 regex routines.  (The routines are derived
(distantly) from Henry Spencer's freely redistributable reimplementation
of the V8 routines.)  See L<Version 8 Regular Expressions> for
details.

In particular the following metacharacters have their standard I<egrep>-ish
meanings:
X<metacharacter>
X<\> X<^> X<.> X<$> X<|> X<(> X<()> X<[> X<[]>


    \        Quote the next metacharacter
    ^        Match the beginning of the line
    .        Match any character (except newline)
    $        Match the end of the line (or before newline at the end)
    |        Alternation
    ()       Grouping
    []       Bracketed Character class

By default, the "^" character is guaranteed to match only the
beginning of the string, the "$" character only the end (or before the
newline at the end), and Perl does certain optimizations with the
assumption that the string contains only one line.  Embedded newlines
will not be matched by "^" or "$".  You may, however, wish to treat a
string as a multi-line buffer, such that the "^" will match after any
newline within the string (except if the newline is the last character in
the string), and "$" will match before any newline.  At the
cost of a little more overhead, you can do this by using the /m modifier
on the pattern match operator.  (Older programs did this by setting C<$*>,
but this option was removed in perl 5.9.)
X<^> X<$> X</m>

To simplify multi-line substitutions, the "." character never matches a
newline unless you use the C</s> modifier, which in effect tells Perl to pretend
the string is a single line--even if it isn't.
X<.> X</s>

=head3 Quantifiers

The following standard quantifiers are recognized:
X<metacharacter> X<quantifier> X<*> X<+> X<?> X<{n}> X<{n,}> X<{n,m}>

    *           Match 0 or more times
    +           Match 1 or more times
    ?           Match 1 or 0 times
    {n}         Match exactly n times
    {n,}        Match at least n times
    {n,m}       Match at least n but not more than m times

(If a curly bracket occurs in any other context and does not form part of
a backslashed sequence like C<\x{...}>, it is treated
as a regular character.  In particular, the lower bound
is not optional.)  The "*" quantifier is equivalent to C<{0,}>, the "+"
quantifier to C<{1,}>, and the "?" quantifier to C<{0,1}>.  n and m are limited
to non-negative integral values less than a preset limit defined when perl is built.
This is usually 32766 on the most common platforms.  The actual limit can
be seen in the error message generated by code such as this:

    $_ **= $_ , / {$_} / for 2 .. 42;

By default, a quantified subpattern is "greedy", that is, it will match as
many times as possible (given a particular starting location) while still
allowing the rest of the pattern to match.  If you want it to match the
minimum number of times possible, follow the quantifier with a "?".  Note
that the meanings don't change, just the "greediness":
X<metacharacter> X<greedy> X<greediness>
X<?> X<*?> X<+?> X<??> X<{n}?> X<{n,}?> X<{n,m}?>

    *?        Match 0 or more times, not greedily
    +?        Match 1 or more times, not greedily
    ??        Match 0 or 1 time, not greedily
    {n}?      Match exactly n times, not greedily (redundant)
    {n,}?     Match at least n times, not greedily
    {n,m}?    Match at least n but not more than m times, not greedily

By default, when a quantified subpattern does not allow the rest of the
overall pattern to match, Perl will backtrack. However, this behaviour is
sometimes undesirable. Thus Perl provides the "possessive" quantifier form
as well.

 *+     Match 0 or more times and give nothing back
 ++     Match 1 or more times and give nothing back
 ?+     Match 0 or 1 time and give nothing back
 {n}+   Match exactly n times and give nothing back (redundant)
 {n,}+  Match at least n times and give nothing back
 {n,m}+ Match at least n but not more than m times and give nothing back

For instance,

   'aaaa' =~ /a++a/

will never match, as the C<a++> will gobble up all the C<a>'s in the
string and won't leave any for the remaining part of the pattern. This
feature can be extremely useful to give perl hints about where it
shouldn't backtrack. For instance, the typical "match a double-quoted
string" problem can be most efficiently performed when written as:

   /"(?:[^"\\]++|\\.)*+"/

as we know that if the final quote does not match, backtracking will not
help. See the independent subexpression
L</C<< (?>pattern) >>> for more details;
possessive quantifiers are just syntactic sugar for that construct. For
instance the above example could also be written as follows:

   /"(?>(?:(?>[^"\\]+)|\\.)*)"/

=head3 Escape sequences

Because patterns are processed as double-quoted strings, the following
also work:

 \t          tab                   (HT, TAB)
 \n          newline               (LF, NL)
 \r          return                (CR)
 \f          form feed             (FF)
 \a          alarm (bell)          (BEL)
 \e          escape (think troff)  (ESC)
 \cK         control char          (example: VT)
 \x{}, \x00  character whose ordinal is the given hexadecimal number
 \N{name}    named Unicode character or character sequence
 \N{U+263D}  Unicode character     (example: FIRST QUARTER MOON)
 \o{}, \000  character whose ordinal is the given octal number
 \l          lowercase next char (think vi)
 \u          uppercase next char (think vi)
 \L          lowercase till \E (think vi)
 \U          uppercase till \E (think vi)
 \Q          quote (disable) pattern metacharacters till \E
 \E          end either case modification or quoted section, think vi

Details are in L<perlop/Quote and Quote-like Operators>.

=head3 Character Classes and other Special Escapes

In addition, Perl defines the following:
X<\g> X<\k> X<\K> X<backreference>

 Sequence   Note    Description
  [...]     [1]  Match a character according to the rules of the
                   bracketed character class defined by the "...".
                   Example: [a-z] matches "a" or "b" or "c" ... or "z"
  [[:...:]] [2]  Match a character according to the rules of the POSIX
                   character class "..." within the outer bracketed
                   character class.  Example: [[:upper:]] matches any
                   uppercase character.
  \w        [3]  Match a "word" character (alphanumeric plus "_", plus
                   other connector punctuation chars plus Unicode
                   marks)
  \W        [3]  Match a non-"word" character
  \s        [3]  Match a whitespace character
  \S        [3]  Match a non-whitespace character
  \d        [3]  Match a decimal digit character
  \D        [3]  Match a non-digit character
  \pP       [3]  Match P, named property.  Use \p{Prop} for longer names
  \PP       [3]  Match non-P
  \X        [4]  Match Unicode "eXtended grapheme cluster"
  \C             Match a single C-language char (octet) even if that is
                   part of a larger UTF-8 character.  Thus it breaks up
                   characters into their UTF-8 bytes, so you may end up
                   with malformed pieces of UTF-8.  Unsupported in
                   lookbehind.
  \1        [5]  Backreference to a specific capture group or buffer.
                   '1' may actually be any positive integer.
  \g1       [5]  Backreference to a specific or previous group,
  \g{-1}    [5]  The number may be negative indicating a relative
                   previous group and may optionally be wrapped in
                   curly brackets for safer parsing.
  \g{name}  [5]  Named backreference
  \k<name>  [5]  Named backreference
  \K        [6]  Keep the stuff left of the \K, don't include it in $&
  \N        [7]  Any character but \n (experimental).  Not affected by
                   /s modifier
  \v        [3]  Vertical whitespace
  \V        [3]  Not vertical whitespace
  \h        [3]  Horizontal whitespace
  \H        [3]  Not horizontal whitespace
  \R        [4]  Linebreak

=over 4

=item [1]

See L<perlrecharclass/Bracketed Character Classes> for details.

=item [2]

See L<perlrecharclass/POSIX Character Classes> for details.

=item [3]

See L<perlrecharclass/Backslash sequences> for details.

=item [4]

See L<perlrebackslash/Misc> for details.

=item [5]

See L</Capture groups> below for details.

=item [6]

See L</Extended Patterns> below for details.

=item [7]

Note that C<\N> has two meanings.  When of the form C<\N{NAME}>, it matches the
character or character sequence whose name is C<NAME>; and similarly
when of the form C<\N{U+I<hex>}>, it matches the character whose Unicode
code point is I<hex>.  Otherwise it matches any character but C<\n>.

=back

=head3 Assertions

Perl defines the following zero-width assertions:
X<zero-width assertion> X<assertion> X<regex, zero-width assertion>
X<regexp, zero-width assertion>
X<regular expression, zero-width assertion>
X<\b> X<\B> X<\A> X<\Z> X<\z> X<\G>

    \b  Match a word boundary
    \B  Match except at a word boundary
    \A  Match only at beginning of string
    \Z  Match only at end of string, or before newline at the end
    \z  Match only at end of string
    \G  Match only at pos() (e.g. at the end-of-match position
        of prior m//g)

A word boundary (C<\b>) is a spot between two characters
that has a C<\w> on one side of it and a C<\W> on the other side
of it (in either order), counting the imaginary characters off the
beginning and end of the string as matching a C<\W>.  (Within
character classes C<\b> represents backspace rather than a word
boundary, just as it normally does in any double-quoted string.)
The C<\A> and C<\Z> are just like "^" and "$", except that they
won't match multiple times when the C</m> modifier is used, while
"^" and "$" will match at every internal line boundary.  To match
the actual end of the string and not ignore an optional trailing
newline, use C<\z>.
X<\b> X<\A> X<\Z> X<\z> X</m>

The C<\G> assertion can be used to chain global matches (using
C<m//g>), as described in L<perlop/"Regexp Quote-Like Operators">.
It is also useful when writing C<lex>-like scanners, when you have
several patterns that you want to match against consequent substrings
of your string; see the previous reference.  The actual location
where C<\G> will match can also be influenced by using C<pos()> as
an lvalue: see L<perlfunc/pos>. Note that the rule for zero-length
matches (see L</"Repeated Patterns Matching a Zero-length Substring">)
is modified somewhat, in that contents to the left of C<\G> are
not counted when determining the length of the match. Thus the following
will not match forever:
X<\G>

     my $string = 'ABC';
     pos($string) = 1;
     while ($string =~ /(.\G)/g) {
         print $1;
     }

It will print 'A' and then terminate, as it considers the match to
be zero-width, and thus will not match at the same position twice in a
row.

It is worth noting that C<\G> improperly used can result in an infinite
loop. Take care when using patterns that include C<\G> in an alternation.

=head3 Capture groups

The bracketing construct C<( ... )> creates capture groups (also referred to as
capture buffers). To refer to the current contents of a group later on, within
the same pattern, use C<\g1> (or C<\g{1}>) for the first, C<\g2> (or C<\g{2}>)
for the second, and so on.
This is called a I<backreference>.
X<regex, capture buffer> X<regexp, capture buffer>
X<regex, capture group> X<regexp, capture group>
X<regular expression, capture buffer> X<backreference>
X<regular expression, capture group> X<backreference>
X<\g{1}> X<\g{-1}> X<\g{name}> X<relative backreference> X<named backreference>
X<named capture buffer> X<regular expression, named capture buffer>
X<named capture group> X<regular expression, named capture group>
X<%+> X<$+{name}> X<< \k<name> >>
There is no limit to the number of captured substrings that you may use.
Groups are numbered with the leftmost open parenthesis being number 1, etc.  If
a group did not match, the associated backreference won't match either. (This
can happen if the group is optional, or in a different branch of an
alternation.)
You can omit the C<"g">, and write C<"\1">, etc, but there are some issues with
this form, described below.

You can also refer to capture groups relatively, by using a negative number, so
that C<\g-1> and C<\g{-1}> both refer to the immediately preceding capture
group, and C<\g-2> and C<\g{-2}> both refer to the group before it.  For
example:

        /
         (Y)            # group 1
         (              # group 2
            (X)         # group 3
            \g{-1}      # backref to group 3
            \g{-3}      # backref to group 1
         )
        /x

would match the same as C</(Y) ( (X) \g3 \g1 )/x>.  This allows you to
interpolate regexes into larger regexes and not have to worry about the
capture groups being renumbered.

You can dispense with numbers altogether and create named capture groups.
The notation is C<(?E<lt>I<name>E<gt>...)> to declare and C<\g{I<name>}> to
reference.  (To be compatible with .Net regular expressions, C<\g{I<name>}> may
also be written as C<\k{I<name>}>, C<\kE<lt>I<name>E<gt>> or C<\k'I<name>'>.)
I<name> must not begin with a number, nor contain hyphens.
When different groups within the same pattern have the same name, any reference
to that name assumes the leftmost defined group.  Named groups count in
absolute and relative numbering, and so can also be referred to by those
numbers.
(It's possible to do things with named capture groups that would otherwise
require C<(??{})>.)

Capture group contents are dynamically scoped and available to you outside the
pattern until the end of the enclosing block or until the next successful
match, whichever comes first.  (See L<perlsyn/"Compound Statements">.)
You can refer to them by absolute number (using C<"$1"> instead of C<"\g1">,
etc); or by name via the C<%+> hash, using C<"$+{I<name>}">.

Braces are required in referring to named capture groups, but are optional for
absolute or relative numbered ones.  Braces are safer when creating a regex by
concatenating smaller strings.  For example if you have C<qr/$a$b/>, and C<$a>
contained C<"\g1">, and C<$b> contained C<"37">, you would get C</\g137/> which
is probably not what you intended.

The C<\g> and C<\k> notations were introduced in Perl 5.10.0.  Prior to that
there were no named nor relative numbered capture groups.  Absolute numbered
groups were referred to using C<\1>,
C<\2>, etc., and this notation is still
accepted (and likely always will be).  But it leads to some ambiguities if
there are more than 9 capture groups, as C<\10> could mean either the tenth
capture group, or the character whose ordinal in octal is 010 (a backspace in
ASCII).  Perl resolves this ambiguity by interpreting C<\10> as a backreference
only if at least 10 left parentheses have opened before it.  Likewise C<\11> is
a backreference only if at least 11 left parentheses have opened before it.
And so on.  C<\1> through C<\9> are always interpreted as backreferences.
There are several examples below that illustrate these perils.  You can avoid
the ambiguity by always using C<\g{}> or C<\g> if you mean capturing groups;
and for octal constants always using C<\o{}>, or for C<\077> and below, using 3
digits padded with leading zeros, since a leading zero implies an octal
constant.

The C<\I<digit>> notation also works in certain circumstances outside
the pattern.  See L</Warning on \1 Instead of $1> below for details.)

Examples:

    s/^([^ ]*) *([^ ]*)/$2 $1/;     # swap first two words

    /(.)\g1/                        # find first doubled char
         and print "'$1' is the first doubled character\n";

    /(?<char>.)\k<char>/            # ... a different way
         and print "'$+{char}' is the first doubled character\n";

    /(?'char'.)\g1/                 # ... mix and match
         and print "'$1' is the first doubled character\n";

    if (/Time: (..):(..):(..)/) {   # parse out values
        $hours = $1;
        $minutes = $2;
        $seconds = $3;
    }

    /(.)(.)(.)(.)(.)(.)(.)(.)(.)\g10/   # \g10 is a backreference
    /(.)(.)(.)(.)(.)(.)(.)(.)(.)\10/    # \10 is octal
    /((.)(.)(.)(.)(.)(.)(.)(.)(.))\10/  # \10 is a backreference
    /((.)(.)(.)(.)(.)(.)(.)(.)(.))\010/ # \010 is octal

    $a = '(.)\1';        # Creates problems when concatenated.
    $b = '(.)\g{1}';     # Avoids the problems.
    "aa" =~ /${a}/;      # True
    "aa" =~ /${b}/;      # True
    "aa0" =~ /${a}0/;    # False!
    "aa0" =~ /${b}0/;    # True
    "aa\x08" =~ /${a}0/;  # True!
    "aa\x08" =~ /${b}0/;  # False

Several special variables also refer back to portions of the previous
match.  C<$+> returns whatever the last bracket match matched.
C<$&> returns the entire matched string.  (At one point C<$0> did
also, but now it returns the name of the program.)  C<$`> returns
everything before the matched string.  C<$'> returns everything
after the matched string. And C<$^N> contains whatever was matched by
the most-recently closed group (submatch). C<$^N> can be used in
extended patterns (see below), for example to assign a submatch to a
variable.
X<$+> X<$^N> X<$&> X<$`> X<$'>

These special variables, like the C<%+> hash and the numbered match variables
(C<$1>, C<$2>, C<$3>, etc.) are dynamically scoped
until the end of the enclosing block or until the next successful
match, whichever comes first.  (See L<perlsyn/"Compound Statements">.)
X<$+> X<$^N> X<$&> X<$`> X<$'>
X<$1> X<$2> X<$3> X<$4> X<$5> X<$6> X<$7> X<$8> X<$9>

B<NOTE>: Failed matches in Perl do not reset the match variables,
which makes it easier to write code that tests for a series of more
specific cases and remembers the best match.

B<WARNING>: Once Perl sees that you need one of C<$&>, C<$`>, or
C<$'> anywhere in the program, it has to provide them for every
pattern match.  This may substantially slow your program.  Perl
uses the same mechanism to produce C<$1>, C<$2>, etc, so you also pay a
price for each pattern that contains capturing parentheses.  (To
avoid this cost while retaining the grouping behaviour, use the
extended regular expression C<(?: ... )> instead.)  But if you never
use C<$&>, C<$`> or C<$'>, then patterns I<without> capturing
parentheses will not be penalized.  So avoid C<$&>, C<$'>, and C<$`>
if you can, but if you can't (and some algorithms really appreciate
them), once you've used them once, use them at will, because you've
already paid the price.  As of 5.005, C<$&> is not so costly as the
other two.
X<$&> X<$`> X<$'>

As a workaround for this problem, Perl 5.10.0 introduces C<${^PREMATCH}>,
C<${^MATCH}> and C<${^POSTMATCH}>, which are equivalent to C<$`>, C<$&>
and C<$'>, B<except> that they are only guaranteed to be defined after a
successful match that was executed with the C</p> (preserve) modifier.
The use of these variables incurs no global performance penalty, unlike
their punctuation char equivalents, however at the trade-off that you
have to tell perl when you want to use them.
X</p> X<p modifier>

=head2 Quoting metacharacters

Backslashed metacharacters in Perl are alphanumeric, such as C<\b>,
C<\w>, C<\n>.  Unlike some other regular expression languages, there
are no backslashed symbols that aren't alphanumeric.  So anything
that looks like \\, \(, \), \<, \>, \{, or \} is always
interpreted as a literal character, not a metacharacter.  This was
once used in a common idiom to disable or quote the special meanings
of regular expression metacharacters in a string that you want to
use for a pattern. Simply quote all non-"word" characters:

    $pattern =~ s/(\W)/\\$1/g;

(If C<use locale> is set, then this depends on the current locale.)
Today it is more common to use the quotemeta() function or the C<\Q>
metaquoting escape sequence to disable all metacharacters' special
meanings like this:

    /$unquoted\Q$quoted\E$unquoted/

Beware that if you put literal backslashes (those not inside
interpolated variables) between C<\Q> and C<\E>, double-quotish
backslash interpolation may lead to confusing results.  If you
I<need> to use literal backslashes within C<\Q...\E>,
consult L<perlop/"Gory details of parsing quoted constructs">.

=head2 Extended Patterns

Perl also defines a consistent extension syntax for features not
found in standard tools like B<awk> and
B<lex>.  The syntax for most of these is a
pair of parentheses with a question mark as the first thing within
the parentheses.  The character after the question mark indicates
the extension.

The stability of these extensions varies widely.  Some have been
part of the core language for many years.  Others are experimental
and may change without warning or be completely removed.  Check
the documentation on an individual feature to verify its current
status.

A question mark was chosen for this and for the minimal-matching
construct because 1) question marks are rare in older regular
expressions, and 2) whenever you see one, you should stop and
"question" exactly what is going on.  That's psychology....

=over 10

=item C<(?#text)>
X<(?#)>

A comment.  The text is ignored.  If the C</x> modifier enables
whitespace formatting, a simple C<#> will suffice.  Note that Perl closes
the comment as soon as it sees a C<)>, so there is no way to put a literal
C<)> in the comment.

=item C<(?adlupimsx-imsx)>

=item C<(?^alupimsx)>
X<(?)> X<(?^)>

One or more embedded pattern-match modifiers, to be turned on (or
turned off, if preceded by C<->) for the remainder of the pattern or
the remainder of the enclosing pattern group (if any).

This is particularly useful for dynamic patterns, such as those read in from a
configuration file, taken from an argument, or specified in a table
somewhere.  Consider the case where some patterns want to be
case-sensitive and some do not:  The case-insensitive ones merely need to
include C<(?i)> at the front of the pattern.  For example:

    $pattern = "foobar";
    if ( /$pattern/i ) { }

    # more flexible:

    $pattern = "(?i)foobar";
    if ( /$pattern/ ) { }

These modifiers are restored at the end of the enclosing group. For example,

    ( (?i) blah ) \s+ \g1

will match C<blah> in any case, some spaces, and an exact (I<including the case>!)
repetition of the previous word, assuming the C</x> modifier, and no C</i>
modifier outside this group.

These modifiers do not carry over into named subpatterns called in the
enclosing group. In other words, a pattern such as C<((?i)(&NAME))> does not
change the case-sensitivity of the "NAME" pattern.

Any of these modifiers can be set to apply globally to all regular
expressions compiled within the scope of a C<use re>.  See
L<re/"'/flags' mode">.

Starting in Perl 5.14, a C<"^"> (caret or circumflex accent) immediately
after the C<"?"> is a shorthand equivalent to C<d-imsx>.  Flags (except
C<"d">) may follow the caret to override it.
But a minus sign is not legal with it.

Note that the C<a>, C<d>, C<l>, C<p>, and C<u> modifiers are special in
that they can only be enabled, not disabled, and the C<a>, C<d>, C<l>, and
C<u> modifiers are mutually exclusive: specifying one de-specifies the
others, and a maximum of one may appear in the construct.  Thus, for
example, C<(?-p)> will warn when compiled under C<use warnings>;
C<(?-d:...)> and C<(?dl:...)> are fatal errors.

Note also that the C<p> modifier is special in that its presence
anywhere in a pattern has a global effect.

=item C<(?:pattern)>
X<(?:)>

=item C<(?adluimsx-imsx:pattern)>

=item C<(?^aluimsx:pattern)>
X<(?^:)>

This is for clustering, not capturing; it groups subexpressions like
"()", but doesn't make backreferences as "()" does.  So

    @fields = split(/\b(?:a|b|c)\b/)

is like

    @fields = split(/\b(a|b|c)\b/)

but doesn't spit out extra fields.  It's also cheaper not to capture
characters if you don't need to.

Any letters between C<?> and C<:> act as flags modifiers as with
C<(?adluimsx-imsx)>.  For example,

    /(?s-i:more.*than).*million/i

is equivalent to the more verbose

    /(?:(?s-i)more.*than).*million/i

Starting in Perl 5.14, a C<"^"> (caret or circumflex accent) immediately
after the C<"?"> is a shorthand equivalent to C<d-imsx>.  Any positive
flags (except C<"d">) may follow the caret, so

    (?^x:foo)

is equivalent to

    (?x-ims:foo)

The caret tells Perl that this cluster doesn't inherit the flags of any
surrounding pattern, but uses the system defaults (C<d-imsx>),
modified by any flags specified.

The caret allows for simpler stringification of compiled regular
expressions.  These look like

    (?^:pattern)

with any non-default flags appearing between the caret and the colon.
A test that looks at such stringification thus doesn't need to have the
system default flags hard-coded in it, just the caret.  If new flags are
added to Perl, the meaning of the caret's expansion will change to include
the default for those flags, so the test will still work, unchanged.

Specifying a negative flag after the caret is an error, as the flag is
redundant.

Mnemonic for C<(?^...)>:  A fresh beginning since the usual use of a caret is
to match at the beginning.

=item C<(?|pattern)>
X<(?|)> X<Branch reset>

This is the "branch reset" pattern, which has the special property
that the capture groups are numbered from the same starting point
in each alternation branch. It is available starting from perl 5.10.0.

Capture groups are numbered from left to right, but inside this
construct the numbering is restarted for each branch.

The numbering within each branch will be as normal, and any groups
following this construct will be numbered as though the construct
contained only one branch, that being the one with the most capture
groups in it.

This construct is useful when you want to capture one of a
number of alternative matches.

Consider the following pattern.  The numbers underneath show in
which group 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     4  

Be careful when using the branch reset pattern in combination with 
named captures. Named captures are implemented as being aliases to 
numbered groups holding the captures, and that interferes with the
implementation of the branch reset pattern. If you are using named
captures in a branch reset pattern, it's best to use the same names,
in the same order, in each of the alternations:

   /(?|  (?<a> x ) (?<b> y )
      |  (?<a> z ) (?<b> w )) /x

Not doing so may lead to surprises:

  "12" =~ /(?| (?<a> \d+ ) | (?<b> \D+))/x;
  say $+ {a};   # Prints '12'
  say $+ {b};   # *Also* prints '12'.

The problem here is that both the group named C<< a >> and the group
named C<< b >> are aliases for the group belonging to C<< $1 >>.

=item Look-Around Assertions
X<look-around assertion> X<lookaround assertion> X<look-around> X<lookaround>

Look-around assertions are zero-width patterns which match a specific
pattern without including it in C<$&>. Positive assertions match when
their subpattern matches, negative assertions match when their subpattern
fails. Look-behind matches text up to the current match position,
look-ahead matches text following the current match position.

=over 4

=item C<(?=pattern)>
X<(?=)> X<look-ahead, positive> X<lookahead, positive>

A zero-width positive look-ahead assertion.  For example, C</\w+(?=\t)/>
matches a word followed by a tab, without including the tab in C<$&>.

=item C<(?!pattern)>
X<(?!)> X<look-ahead, negative> X<lookahead, negative>

A zero-width negative look-ahead assertion.  For example C</foo(?!bar)/>
matches any occurrence of "foo" that isn't followed by "bar".  Note
however that look-ahead and look-behind are NOT the same thing.  You cannot
use this for look-behind.

If you are looking for a "bar" that isn't preceded by a "foo", C</(?!foo)bar/>
will not do what you want.  That's because the C<(?!foo)> is just saying that
the next thing cannot be "foo"--and it's not, it's a "bar", so "foobar" will
match.  Use look-behind instead (see below).

=item C<(?<=pattern)> C<\K>
X<(?<=)> X<look-behind, positive> X<lookbehind, positive> X<\K>

A zero-width positive look-behind assertion.  For example, C</(?<=\t)\w+/>
matches a word that follows a tab, without including the tab in C<$&>.
Works only for fixed-width look-behind.

There is a special form of this construct, called C<\K>, which causes the
regex engine to "keep" everything it had matched prior to the C<\K> and
not include it in C<$&>. This effectively provides variable-length
look-behind. The use of C<\K> inside of another look-around assertion
is allowed, but the behaviour is currently not well defined.

For various reasons C<\K> may be significantly more efficient than the
equivalent C<< (?<=...) >> construct, and it is especially useful in
situations where you want to efficiently remove something following
something else in a string. For instance

  s/(foo)bar/$1/g;

can be rewritten as the much more efficient

  s/foo\Kbar//g;

=item C<(?<!pattern)>
X<(?<!)> X<look-behind, negative> X<lookbehind, negative>

A zero-width negative look-behind assertion.  For example C</(?<!bar)foo/>
matches any occurrence of "foo" that does not follow "bar".  Works
only for fixed-width look-behind.

=back

=item C<(?'NAME'pattern)>

=item C<< (?<NAME>pattern) >>
X<< (?<NAME>) >> X<(?'NAME')> X<named capture> X<capture>

A named capture group. Identical in every respect to normal capturing
parentheses C<()> but for the additional fact that the group
can be referred to by name in various regular expression
constructs (like C<\g{NAME}>) and can be accessed by name
after a successful match via C<%+> or C<%->. See L<perlvar>
for more details on the C<%+> and C<%-> hashes.

If multiple distinct capture groups have the same name then the
$+{NAME} will refer to the leftmost defined group in the match.

The forms C<(?'NAME'pattern)> and C<< (?<NAME>pattern) >> are equivalent.

B<NOTE:> While the notation of this construct is the same as the similar
function in .NET regexes, the behavior is not. In Perl the groups are
numbered sequentially regardless of being named or not. Thus in the
pattern

  /(x)(?<foo>y)(z)/

$+{foo} will be the same as $2, and $3 will contain 'z' instead of
the opposite which is what a .NET regex hacker might expect.

Currently NAME is restricted to simple identifiers only.
In other words, it must match C</^[_A-Za-z][_A-Za-z0-9]*\z/> or
its Unicode extension (see L<utf8>),
though it isn't extended by the locale (see L<perllocale>).

B<NOTE:> In order to make things easier for programmers with experience
with the Python or PCRE regex engines, the pattern C<< (?PE<lt>NAMEE<gt>pattern) >>
may be used instead of C<< (?<NAME>pattern) >>; however this form does not
support the use of single quotes as a delimiter for the name.

=item C<< \k<NAME> >>

=item C<< \k'NAME' >>

Named backreference. Similar to numeric backreferences, except that
the group is designated by name and not number. If multiple groups
have the same name then it refers to the leftmost defined group in
the current match.

It is an error to refer to a name not defined by a C<< (?<NAME>) >>
earlier in the pattern.

Both forms are equivalent.

B<NOTE:> In order to make things easier for programmers with experience
with the Python or PCRE regex engines, the pattern C<< (?P=NAME) >>
may be used instead of C<< \k<NAME> >>.

=item C<(?{ code })>
X<(?{})> X<regex, code in> X<regexp, code in> X<regular expression, code in>

B<WARNING>: This extended regular expression feature is considered
experimental, and may be changed without notice. Code executed that
has side effects may not perform identically from version to version
due to the effect of future optimisations in the regex engine.

This zero-width assertion evaluates any embedded Perl code.  It
always succeeds, and its C<code> is not interpolated.  Currently,
the rules to determine where the C<code> ends are somewhat convoluted.

This feature can be used together with the special variable C<$^N> to
capture the results of submatches in variables without having to keep
track of the number of nested parentheses. For example:

  $_ = "The brown fox jumps over the lazy dog";
  /the (\S+)(?{ $color = $^N }) (\S+)(?{ $animal = $^N })/i;
  print "color = $color, animal = $animal\n";

Inside the C<(?{...})> block, C<$_> refers to the string the regular
expression is matching against. You can also use C<pos()> to know what is
the current position of matching within this string.

The C<code> is properly scoped in the following sense: If the assertion
is backtracked (compare L<"Backtracking">), all changes introduced after
C<local>ization are undone, so that

  $_ = 'a' x 8;
  m<
     (?{ $cnt = 0 })                   # Initialize $cnt.
     (
       a
       (?{
           local $cnt = $cnt + 1;      # Update $cnt, backtracking-safe.
       })
     )*
     aaaa
     (?{ $res = $cnt })                # On success copy to
                                       # non-localized location.
   >x;

will set C<$res = 4>.  Note that after the match, C<$cnt> returns to the globally
introduced value, because the scopes that restrict C<local> operators
are unwound.

This assertion may be used as a C<(?(condition)yes-pattern|no-pattern)>
switch.  If I<not> used in this way, the result of evaluation of
C<code> is put into the special variable C<$^R>.  This happens
immediately, so C<$^R> can be used from other C<(?{ code })> assertions
inside the same regular expression.

The assignment to C<$^R> above is properly localized, so the old
value of C<$^R> is restored if the assertion is backtracked; compare
L<"Backtracking">.

For reasons of security, this construct is forbidden if the regular
expression involves run-time interpolation of variables, unless the
perilous C<use re 'eval'> pragma has been used (see L<re>), or the
variables contain results of the C<qr//> operator (see
L<perlop/"qr/STRINGE<sol>msixpodual">).

This restriction is due to the wide-spread and remarkably convenient
custom of using run-time determined strings as patterns.  For example:

    $re = <>;
    chomp $re;
    $string =~ /$re/;

Before Perl knew how to execute interpolated code within a pattern,
this operation was completely safe from a security point of view,
although it could raise an exception from an illegal pattern.  If
you turn on the C<use re 'eval'>, though, it is no longer secure,
so you should only do so if you are also using taint checking.
Better yet, use the carefully constrained evaluation within a Safe
compartment.  See L<perlsec> for details about both these mechanisms.

B<WARNING>: Use of lexical (C<my>) variables in these blocks is
broken. The result is unpredictable and will make perl unstable. The
workaround is to use global (C<our>) variables.

B<WARNING>: In perl 5.12.x and earlier, the regex engine
was not re-entrant, so interpolated code could not
safely invoke the regex engine either directly with
C<m//> or C<s///>), or indirectly with functions such as
C<split>. Invoking the regex engine in these blocks would make perl
unstable.

=item C<(??{ code })>
X<(??{})>
X<regex, postponed> X<regexp, postponed> X<regular expression, postponed>

B<WARNING>: This extended regular expression feature is considered
experimental, and may be changed without notice. Code executed that
has side effects may not perform identically from version to version
due to the effect of future optimisations in the regex engine.

This is a "postponed" regular subexpression.  The C<code> is evaluated
at run time, at the moment this subexpression may match.  The result
of evaluation is considered a regular expression and matched as
if it were inserted instead of this construct.  Note that this means
that the contents of capture groups defined inside an eval'ed pattern
are not available outside of the pattern, and vice versa, there is no
way for the inner pattern to refer to a capture group defined outside.
Thus,

    ('a' x 100)=~/(??{'(.)' x 100})/

B<will> match, it will B<not> set $1.

The C<code> is not interpolated.  As before, the rules to determine
where the C<code> ends are currently somewhat convoluted.

The following pattern matches a parenthesized group:

  $re = qr{
             \(
             (?:
                (?> [^()]+ )       # Non-parens without backtracking
              |
                (??{ $re })        # Group with matching parens
             )*
             \)
          }x;

See also C<(?PARNO)> for a different, more efficient way to accomplish
the same task.

For reasons of security, this construct is forbidden if the regular
expression involves run-time interpolation of variables, unless the
perilous C<use re 'eval'> pragma has been used (see L<re>), or the
variables contain results of the C<qr//> operator (see
L<perlop/"qrE<sol>STRINGE<sol>msixpodual">).

In perl 5.12.x and earlier, because the regex engine was not re-entrant,
delayed code could not safely invoke the regex engine either directly with
C<m//> or C<s///>), or indirectly with functions such as C<split>.

Recursing deeper than 50 times without consuming any input string will
result in a fatal error.  The maximum depth is compiled into perl, so
changing it requires a custom build.

=item C<(?PARNO)> C<(?-PARNO)> C<(?+PARNO)> C<(?R)> C<(?0)>
X<(?PARNO)> X<(?1)> X<(?R)> X<(?0)> X<(?-1)> X<(?+1)> X<(?-PARNO)> X<(?+PARNO)>
X<regex, recursive> X<regexp, recursive> X<regular expression, recursive>
X<regex, relative recursion>

Similar to C<(??{ code })> except it does not involve compiling any code,
instead it treats the contents of a capture group as an independent
pattern that must match at the current position.  Capture groups
contained by the pattern will have the value as determined by the
outermost recursion.

PARNO is a sequence of digits (not starting with 0) whose value reflects
the paren-number of the capture group to recurse to. C<(?R)> recurses to
the beginning of the whole pattern. C<(?0)> is an alternate syntax for
C<(?R)>. If PARNO is preceded by a plus or minus sign then it is assumed
to be relative, with negative numbers indicating preceding capture groups
and positive ones following. Thus C<(?-1)> refers to the most recently
declared group, and C<(?+1)> indicates the next group to be declared.
Note that the counting for relative recursion differs from that of
relative backreferences, in that with recursion unclosed groups B<are>
included.

The following pattern matches a function foo() which may contain
balanced parentheses as the argument.

  $re = qr{ (                    # paren group 1 (full function)
              foo
              (                  # paren group 2 (parens)
                \(
                  (              # paren group 3 (contents of parens)
                  (?:
                   (?> [^()]+ )  # Non-parens without backtracking
                  |
                   (?2)          # Recurse to start of paren group 2
                  )*
                  )
                \)
              )
            )
          }x;

If the pattern was used as follows

    'foo(bar(baz)+baz(bop))'=~/$re/
        and print "\$1 = $1\n",
                  "\$2 = $2\n",
                  "\$3 = $3\n";

the output produced should be the following:

    $1 = foo(bar(baz)+baz(bop))
    $2 = (bar(baz)+baz(bop))
    $3 = bar(baz)+baz(bop)

If there is no corresponding capture group defined, then it is a
fatal error.  Recursing deeper than 50 times without consuming any input
string will also result in a fatal error.  The maximum depth is compiled
into perl, so changing it requires a custom build.

The following shows how using negative indexing can make it
easier to embed recursive patterns inside of a C<qr//> construct
for later use:

    my $parens = qr/(\((?:[^()]++|(?-1))*+\))/;
    if (/foo $parens \s+ + \s+ bar $parens/x) {
       # do something here...
    }

B<Note> that this pattern does not behave the same way as the equivalent
PCRE or Python construct of the same form. In Perl you can backtrack into
a recursed group, in PCRE and Python the recursed into group is treated
as atomic. Also, modifiers are resolved at compile time, so constructs
like (?i:(?1)) or (?:(?i)(?1)) do not affect how the sub-pattern will
be processed.

=item C<(?&NAME)>
X<(?&NAME)>

Recurse to a named subpattern. Identical to C<(?PARNO)> except that the
parenthesis to recurse to is determined by name. If multiple parentheses have
the same name, then it recurses to the leftmost.

It is an error to refer to a name that is not declared somewhere in the
pattern.

B<NOTE:> In order to make things easier for programmers with experience
with the Python or PCRE regex engines the pattern C<< (?P>NAME) >>
may be used instead of C<< (?&NAME) >>.

=item C<(?(condition)yes-pattern|no-pattern)>
X<(?()>

=item C<(?(condition)yes-pattern)>

Conditional expression.  C<(condition)> should be either an integer in
parentheses (which is valid if the corresponding pair of parentheses
matched), a look-ahead/look-behind/evaluate zero-width assertion, a
name in angle brackets or single quotes (which is valid if a group
with the given name matched), or the special symbol (R) (true when
evaluated inside of recursion or eval). Additionally the R may be
followed by a number, (which will be true when evaluated when recursing
inside of the appropriate group), or by C<&NAME>, in which case it will
be true only when evaluated during recursion in the named group.

Here's a summary of the possible predicates:

=over 4

=item (1) (2) ...

Checks if the numbered capturing group has matched something.

=item (<NAME>) ('NAME')

Checks if a group with the given name has matched something.

=item (?=...) (?!...) (?<=...) (?<!...)

Checks whether the pattern matches (or does not match, for the '!'
variants).

=item (?{ CODE })

Treats the return value of the code block as the condition.

=item (R)

Checks if the expression has been evaluated inside of recursion.

=item (R1) (R2) ...

Checks if the expression has been evaluated while executing directly
inside of the n-th capture group. This check is the regex equivalent of

  if ((caller(0))[3] eq 'subname') { ... }

In other words, it does not check the full recursion stack.

=item (R&NAME)

Similar to C<(R1)>, this predicate checks to see if we're executing
directly inside of the leftmost group with a given name (this is the same
logic used by C<(?&NAME)> to disambiguate). It does not check the full
stack, but only the name of the innermost active recursion.

=item (DEFINE)

In this case, the yes-pattern is never directly executed, and no
no-pattern is allowed. Similar in spirit to C<(?{0})> but more efficient.
See below for details.

=back

For example:

    m{ ( \( )?
       [^()]+
       (?(1) \) )
     }x

matches a chunk of non-parentheses, possibly included in parentheses
themselves.

A special form is the C<(DEFINE)> predicate, which never executes its
yes-pattern directly, and does not allow a no-pattern. This allows one to
define subpatterns which will be executed only by the recursion mechanism.
This way, you can define a set of regular expression rules that can be
bundled into any pattern you choose.

It is recommended that for this usage you put the DEFINE block at the
end of the pattern, and that you name any subpatterns defined within it.

Also, it's worth noting that patterns defined this way probably will
not be as efficient, as the optimiser is not very clever about
handling them.

An example of how this might be used is as follows:

  /(?<NAME>(?&NAME_PAT))(?<ADDR>(?&ADDRESS_PAT))
   (?(DEFINE)
     (?<NAME_PAT>....)
     (?<ADRESS_PAT>....)
   )/x

Note that capture groups matched inside of recursion are not accessible
after the recursion returns, so the extra layer of capturing groups is
necessary. Thus C<$+{NAME_PAT}> would not be defined even though
C<$+{NAME}> would be.

=item C<< (?>pattern) >>
X<backtrack> X<backtracking> X<atomic> X<possessive>

An "independent" subexpression, one which matches the substring
that a I<standalone> C<pattern> would match if anchored at the given
position, and it matches I<nothing other than this substring>.  This
construct is useful for optimizations of what would otherwise be
"eternal" matches, because it will not backtrack (see L<"Backtracking">).
It may also be useful in places where the "grab all you can, and do not
give anything back" semantic is desirable.

For example: C<< ^(?>a*)ab >> will never match, since C<< (?>a*) >>
(anchored at the beginning of string, as above) will match I<all>
characters C<a> at the beginning of string, leaving no C<a> for
C<ab> to match.  In contrast, C<a*ab> will match the same as C<a+b>,
since the match of the subgroup C<a*> is influenced by the following
group C<ab> (see L<"Backtracking">).  In particular, C<a*> inside
C<a*ab> will match fewer characters than a standalone C<a*>, since
this makes the tail match.

C<< (?>pattern) >> does not disable backtracking altogether once it has
matched. It is still possible to backtrack past the construct, but not
into it. So C<< ((?>a*)|(?>b*))ar >> will still match "bar".

An effect similar to C<< (?>pattern) >> may be achieved by writing
C<(?=(pattern))\g{-1}>.  This matches the same substring as a standalone
C<a+>, and the following C<\g{-1}> eats the matched string; it therefore
makes a zero-length assertion into an analogue of C<< (?>...) >>.
(The difference between these two constructs is that the second one
uses a capturing group, thus shifting ordinals of backreferences
in the rest of a regular expression.)

Consider this pattern:

    m{ \(
          (
            [^()]+           # x+
          |
            \( [^()]* \)
          )+
       \)
     }x

That will efficiently match a nonempty group with matching parentheses
two levels deep or less.  However, if there is no such group, it
will take virtually forever on a long string.  That's because there
are so many different ways to split a long string into several
substrings.  This is what C<(.+)+> is doing, and C<(.+)+> is similar
to a subpattern of the above pattern.  Consider how the pattern
above detects no-match on C<((()aaaaaaaaaaaaaaaaaa> in several
seconds, but that each extra letter doubles this time.  This
exponential performance will make it appear that your program has
hung.  However, a tiny change to this pattern

    m{ \(
          (
            (?> [^()]+ )        # change x+ above to (?> x+ )
          |
            \( [^()]* \)
          )+
       \)
     }x

which uses C<< (?>...) >> matches exactly when the one above does (verifying
this yourself would be a productive exercise), but finishes in a fourth
the time when used on a similar string with 1000000 C<a>s.  Be aware,
however, that, when this construct is followed by a
quantifier, it currently triggers a warning message under
the C<use warnings> pragma or B<-w> switch saying it
C<"matches null string many times in regex">.

On simple groups, such as the pattern C<< (?> [^()]+ ) >>, a comparable
effect may be achieved by negative look-ahead, as in C<[^()]+ (?! [^()] )>.
This was only 4 times slower on a string with 1000000 C<a>s.

The "grab all you can, and do not give anything back" semantic is desirable
in many situations where on the first sight a simple C<()*> looks like
the correct solution.  Suppose we parse text with comments being delimited
by C<#> followed by some optional (horizontal) whitespace.  Contrary to
its appearance, C<#[ \t]*> I<is not> the correct subexpression to match
the comment delimiter, because it may "give up" some whitespace if
the remainder of the pattern can be made to match that way.  The correct
answer is either one of these:

    (?>#[ \t]*)
    #[ \t]*(?![ \t])

For example, to grab non-empty comments into $1, one should use either
one of these:

    / (?> \# [ \t]* ) (        .+ ) /x;
    /     \# [ \t]*   ( [^ \t] .* ) /x;

Which one you pick depends on which of these expressions better reflects
the above specification of comments.

In some literature this construct is called "atomic matching" or
"possessive matching".

Possessive quantifiers are equivalent to putting the item they are applied
to inside of one of these constructs. The following equivalences apply:

    Quantifier Form     Bracketing Form
    ---------------     ---------------
    PAT*+               (?>PAT*)
    PAT++               (?>PAT+)
    PAT?+               (?>PAT?)
    PAT{min,max}+       (?>PAT{min,max})

=back

=head2 Special Backtracking Control Verbs

B<WARNING:> These patterns are experimental and subject to change or
removal in a future version of Perl. Their usage in production code should
be noted to avoid problems during upgrades.

These special patterns are generally of the form C<(*VERB:ARG)>. Unless
otherwise stated the ARG argument is optional; in some cases, it is
forbidden.

Any pattern containing a special backtracking verb that allows an argument
has the special behaviour that when executed it sets the current package's
C<$REGERROR> and C<$REGMARK> variables. When doing so the following
rules apply:

On failure, the C<$REGERROR> variable will be set to the ARG value of the
verb pattern, if the verb was involved in the failure of the match. If the
ARG part of the pattern was omitted, then C<$REGERROR> will be set to the
name of the last C<(*MARK:NAME)> pattern executed, or to TRUE if there was
none. Also, the C<$REGMARK> variable will be set to FALSE.

On a successful match, the C<$REGERROR> variable will be set to FALSE, and
the C<$REGMARK> variable will be set to the name of the last
C<(*MARK:NAME)> pattern executed.  See the explanation for the
C<(*MARK:NAME)> verb below for more details.

B<NOTE:> C<$REGERROR> and C<$REGMARK> are not magic variables like C<$1>
and most other regex-related variables. They are not local to a scope, nor
readonly, but instead are volatile package variables similar to C<$AUTOLOAD>.
Use C<local> to localize changes to them to a specific scope if necessary.

If a pattern does not contain a special backtracking verb that allows an
argument, then C<$REGERROR> and C<$REGMARK> are not touched at all.

=over 4

=item Verbs that take an argument

=over 4

=item C<(*PRUNE)> C<(*PRUNE:NAME)>
X<(*PRUNE)> X<(*PRUNE:NAME)>

This zero-width pattern prunes the backtracking tree at the current point
when backtracked into on failure. Consider the pattern C<A (*PRUNE) B>,
where A and B are complex patterns. Until the C<(*PRUNE)> verb is reached,
A may backtrack as necessary to match. Once it is reached, matching
continues in B, which may also backtrack as necessary; however, should B
not match, then no further backtracking will take place, and the pattern
will fail outright at the current starting position.

The following example counts all the possible matching strings in a
pattern (without actually matching any of them).

    'aaab' =~ /a+b?(?{print "$&\n"; $count++})(*FAIL)/;
    print "Count=$count\n";

which produces:

    aaab
    aaa
    aa
    a
    aab
    aa
    a
    ab
    a
    Count=9

If we add a C<(*PRUNE)> before the count like the following

    'aaab' =~ /a+b?(*PRUNE)(?{print "$&\n"; $count++})(*FAIL)/;
    print "Count=$count\n";

we prevent backtracking and find the count of the longest matching string
at each matching starting point like so:

    aaab
    aab
    ab
    Count=3

Any number of C<(*PRUNE)> assertions may be used in a pattern.

See also C<< (?>pattern) >> and possessive quantifiers for other ways to
control backtracking. In some cases, the use of C<(*PRUNE)> can be
replaced with a C<< (?>pattern) >> with no functional difference; however,
C<(*PRUNE)> can be used to handle cases that cannot be expressed using a
C<< (?>pattern) >> alone.


=item C<(*SKIP)> C<(*SKIP:NAME)>
X<(*SKIP)>

This zero-width pattern is similar to C<(*PRUNE)>, except that on
failure it also signifies that whatever text that was matched leading up
to the C<(*SKIP)> pattern being executed cannot be part of I<any> match
of this pattern. This effectively means that the regex engine "skips" forward
to this position on failure and tries to match again, (assuming that
there is sufficient room to match).

The name of the C<(*SKIP:NAME)> pattern has special significance. If a
C<(*MARK:NAME)> was encountered while matching, then it is that position
which is used as the "skip point". If no C<(*MARK)> of that name was
encountered, then the C<(*SKIP)> operator has no effect. When used
without a name the "skip point" is where the match point was when
executing the (*SKIP) pattern.

Compare the following to the examples in C<(*PRUNE)>; note the string
is twice as long:

    'aaabaaab' =~ /a+b?(*SKIP)(?{print "$&\n"; $count++})(*FAIL)/;
    print "Count=$count\n";

outputs

    aaab
    aaab
    Count=2

Once the 'aaab' at the start of the string has matched, and the C<(*SKIP)>
executed, the next starting point will be where the cursor was when the
C<(*SKIP)> was executed.

=item C<(*MARK:NAME)> C<(*:NAME)>
X<(*MARK)> C<(*MARK:NAME)> C<(*:NAME)>

This zero-width pattern can be used to mark the point reached in a string
when a certain part of the pattern has been successfully matched. This
mark may be given a name. A later C<(*SKIP)> pattern will then skip
forward to that point if backtracked into on failure. Any number of
C<(*MARK)> patterns are allowed, and the NAME portion may be duplicated.

In addition to interacting with the C<(*SKIP)> pattern, C<(*MARK:NAME)>
can be used to "label" a pattern branch, so that after matching, the
program can determine which branches of the pattern were involved in the
match.

When a match is successful, the C<$REGMARK> variable will be set to the
name of the most recently executed C<(*MARK:NAME)> that was involved
in the match.

This can be used to determine which branch of a pattern was matched
without using a separate capture group for each branch, which in turn
can result in a performance improvement, as perl cannot optimize
C</(?:(x)|(y)|(z))/> as efficiently as something like
C</(?:x(*MARK:x)|y(*MARK:y)|z(*MARK:z))/>.

When a match has failed, and unless another verb has been involved in
failing the match and has provided its own name to use, the C<$REGERROR>
variable will be set to the name of the most recently executed
C<(*MARK:NAME)>.

See C<(*SKIP)> for more details.

As a shortcut C<(*MARK:NAME)> can be written C<(*:NAME)>.

=item C<(*THEN)> C<(*THEN:NAME)>

This is similar to the "cut group" operator C<::> from Perl 6. Like
C<(*PRUNE)>, this verb always matches, and when backtracked into on
failure, it causes the regex engine to try the next alternation in the
innermost enclosing group (capturing or otherwise).

Its name comes from the observation that this operation combined with the
alternation operator (C<|>) can be used to create what is essentially a
pattern-based if/then/else block:

  ( COND (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ )

Note that if this operator is used and NOT inside of an alternation then
it acts exactly like the C<(*PRUNE)> operator.

  / A (*PRUNE) B /

is the same as

  / A (*THEN) B /

but

  / ( A (*THEN) B | C (*THEN) D ) /

is not the same as

  / ( A (*PRUNE) B | C (*PRUNE) D ) /

as after matching the A but failing on the B the C<(*THEN)> verb will
backtrack and try C; but the C<(*PRUNE)> verb will simply fail.

=item C<(*COMMIT)>
X<(*COMMIT)>

This is the Perl 6 "commit pattern" C<< <commit> >> or C<:::>. It's a
zero-width pattern similar to C<(*SKIP)>, except that when backtracked
into on failure it causes the match to fail outright. No further attempts
to find a valid match by advancing the start pointer will occur again.
For example,

    'aaabaaab' =~ /a+b?(*COMMIT)(?{print "$&\n"; $count++})(*FAIL)/;
    print "Count=$count\n";

outputs

    aaab
    Count=1

In other words, once the C<(*COMMIT)> has been entered, and if the pattern
does not match, the regex engine will not try any further matching on the
rest of the string.

=back

=item Verbs without an argument

=over 4

=item C<(*FAIL)> C<(*F)>
X<(*FAIL)> X<(*F)>

This pattern matches nothing and always fails. It can be used to force the
engine to backtrack. It is equivalent to C<(?!)>, but easier to read. In
fact, C<(?!)> gets optimised into C<(*FAIL)> internally.

It is probably useful only when combined with C<(?{})> or C<(??{})>.

=item C<(*ACCEPT)>
X<(*ACCEPT)>

B<WARNING:> This feature is highly experimental. It is not recommended
for production code.

This pattern matches nothing and causes the end of successful matching at
the point at which the C<(*ACCEPT)> pattern was encountered, regardless of
whether there is actually more to match in the string. When inside of a
nested pattern, such as recursion, or in a subpattern dynamically generated
via C<(??{})>, only the innermost pattern is ended immediately.

If the C<(*ACCEPT)> is inside of capturing groups then the groups are
marked as ended at the point at which the C<(*ACCEPT)> was encountered.
For instance:

  'AB' =~ /(A (A|B(*ACCEPT)|C) D)(E)/x;

will match, and C<$1> will be C<AB> and C<$2> will be C<B>, C<$3> will not
be set. If another branch in the inner parentheses was matched, such as in the
string 'ACDE', then the C<D> and C<E> would have to be matched as well.

=back

=back

=head2 Backtracking
X<backtrack> X<backtracking>

NOTE: This section presents an abstract approximation of regular
expression behavior.  For a more rigorous (and complicated) view of
the rules involved in selecting a match among possible alternatives,
see L<Combining RE Pieces>.

A fundamental feature of regular expression matching involves the
notion called I<backtracking>, which is currently used (when needed)
by all regular non-possessive expression quantifiers, namely C<*>, C<*?>, C<+>,
C<+?>, C<{n,m}>, and C<{n,m}?>.  Backtracking is often optimized
internally, but the general principle outlined here is valid.

For a regular expression to match, the I<entire> regular expression must
match, not just part of it.  So if the beginning of a pattern containing a
quantifier succeeds in a way that causes later parts in the pattern to
fail, the matching engine backs up and recalculates the beginning
part--that's why it's called backtracking.

Here is an example of backtracking:  Let's say you want to find the
word following "foo" in the string "Food is on the foo table.":

    $_ = "Food is on the foo table.";
    if ( /\b(foo)\s+(\w+)/i ) {
        print "$2 follows $1.\n";
    }

When the match runs, the first part of the regular expression (C<\b(foo)>)
finds a possible match right at the beginning of the string, and loads up
$1 with "Foo".  However, as soon as the matching engine sees that there's
no whitespace following the "Foo" that it had saved in $1, it realizes its
mistake and starts over again one character after where it had the
tentative match.  This time it goes all the way until the next occurrence
of "foo". The complete regular expression matches this time, and you get
the expected output of "table follows foo."

Sometimes minimal matching can help a lot.  Imagine you'd like to match
everything between "foo" and "bar".  Initially, you write something
like this:

    $_ =  "The food is under the bar in the barn.";
    if ( /foo(.*)bar/ ) {
        print "got <$1>\n";
    }

Which perhaps unexpectedly yields:

  got <d is under the bar in the >

That's because C<.*> was greedy, so you get everything between the
I<first> "foo" and the I<last> "bar".  Here it's more effective
to use minimal matching to make sure you get the text between a "foo"
and the first "bar" thereafter.

    if ( /foo(.*?)bar/ ) { print "got <$1>\n" }
  got <d is under the >

Here's another example. Let's say you'd like to match a number at the end
of a string, and you also want to keep the preceding part of the match.
So you write this:

    $_ = "I have 2 numbers: 53147";
    if ( /(.*)(\d*)/ ) {                                # Wrong!
        print "Beginning is <$1>, number is <$2>.\n";
    }

That won't work at all, because C<.*> was greedy and gobbled up the
whole string. As C<\d*> can match on an empty string the complete
regular expression matched successfully.

    Beginning is <I have 2 numbers: 53147>, number is <>.

Here are some variants, most of which don't work:

    $_ = "I have 2 numbers: 53147";
    @pats = qw{
        (.*)(\d*)
        (.*)(\d+)
        (.*?)(\d*)
        (.*?)(\d+)
        (.*)(\d+)$
        (.*?)(\d+)$
        (.*)\b(\d+)$
        (.*\D)(\d+)$
    };

    for $pat (@pats) {
        printf "%-12s ", $pat;
        if ( /$pat/ ) {
            print "<$1> <$2>\n";
        } else {
            print "FAIL\n";
        }
    }

That will print out:

    (.*)(\d*)    <I have 2 numbers: 53147> <>
    (.*)(\d+)    <I have 2 numbers: 5314> <7>
    (.*?)(\d*)   <> <>
    (.*?)(\d+)   <I have > <2>
    (.*)(\d+)$   <I have 2 numbers: 5314> <7>
    (.*?)(\d+)$  <I have 2 numbers: > <53147>
    (.*)\b(\d+)$ <I have 2 numbers: > <53147>
    (.*\D)(\d+)$ <I have 2 numbers: > <53147>

As you see, this can be a bit tricky.  It's important to realize that a
regular expression is merely a set of assertions that gives a definition
of success.  There may be 0, 1, or several different ways that the
definition might succeed against a particular string.  And if there are
multiple ways it might succeed, you need to understand backtracking to
know which variety of success you will achieve.

When using look-ahead assertions and negations, this can all get even
trickier.  Imagine you'd like to find a sequence of non-digits not
followed by "123".  You might try to write that as

    $_ = "ABC123";
    if ( /^\D*(?!123)/ ) {                # Wrong!
        print "Yup, no 123 in $_\n";
    }

But that isn't going to match; at least, not the way you're hoping.  It
claims that there is no 123 in the string.  Here's a clearer picture of
why that pattern matches, contrary to popular expectations:

    $x = 'ABC123';
    $y = 'ABC445';

    print "1: got $1\n" if $x =~ /^(ABC)(?!123)/;
    print "2: got $1\n" if $y =~ /^(ABC)(?!123)/;

    print "3: got $1\n" if $x =~ /^(\D*)(?!123)/;
    print "4: got $1\n" if $y =~ /^(\D*)(?!123)/;

This prints

    2: got ABC
    3: got AB
    4: got ABC

You might have expected test 3 to fail because it seems to a more
general purpose version of test 1.  The important difference between
them is that test 3 contains a quantifier (C<\D*>) and so can use
backtracking, whereas test 1 will not.  What's happening is
that you've asked "Is it true that at the start of $x, following 0 or more
non-digits, you have something that's not 123?"  If the pattern matcher had
let C<\D*> expand to "ABC", this would have caused the whole pattern to
fail.

The search engine will initially match C<\D*> with "ABC".  Then it will
try to match C<(?!123)> with "123", which fails.  But because
a quantifier (C<\D*>) has been used in the regular expression, the
search engine can backtrack and retry the match differently
in the hope of matching the complete regular expression.

The pattern really, I<really> wants to succeed, so it uses the
standard pattern back-off-and-retry and lets C<\D*> expand to just "AB" this
time.  Now there's indeed something following "AB" that is not
"123".  It's "C123", which suffices.

We can deal with this by using both an assertion and a negation.
We'll say that the first part in $1 must be followed both by a digit
and by something that's not "123".  Remember that the look-aheads
are zero-width expressions--they only look, but don't consume any
of the string in their match.  So rewriting this way produces what
you'd expect; that is, case 5 will fail, but case 6 succeeds:

    print "5: got $1\n" if $x =~ /^(\D*)(?=\d)(?!123)/;
    print "6: got $1\n" if $y =~ /^(\D*)(?=\d)(?!123)/;

    6: got ABC

In other words, the two zero-width assertions next to each other work as though
they're ANDed together, just as you'd use any built-in assertions:  C</^$/>
matches only if you're at the beginning of the line AND the end of the
line simultaneously.  The deeper underlying truth is that juxtaposition in
regular expressions always means AND, except when you write an explicit OR
using the vertical bar.  C</ab/> means match "a" AND (then) match "b",
although the attempted matches are made at different positions because "a"
is not a zero-width assertion, but a one-width assertion.

B<WARNING>: Particularly complicated regular expressions can take
exponential time to solve because of the immense number of possible
ways they can use backtracking to try for a match.  For example, without
internal optimizations done by the regular expression engine, this will
take a painfully long time to run:

    'aaaaaaaaaaaa' =~ /((a{0,5}){0,5})*[c]/

And if you used C<*>'s in the internal groups instead of limiting them
to 0 through 5 matches, then it would take forever--or until you ran
out of stack space.  Moreover, these internal optimizations are not
always applicable.  For example, if you put C<{0,5}> instead of C<*>
on the external group, no current optimization is applicable, and the
match takes a long time to finish.

A powerful tool for optimizing such beasts is what is known as an
"independent group",
which does not backtrack (see L</C<< (?>pattern) >>>).  Note also that
zero-length look-ahead/look-behind assertions will not backtrack to make
the tail match, since they are in "logical" context: only
whether they match is considered relevant.  For an example
where side-effects of look-ahead I<might> have influenced the
following match, see L</C<< (?>pattern) >>>.

=head2 Version 8 Regular Expressions
X<regular expression, version 8> X<regex, version 8> X<regexp, version 8>

In case you're not familiar with the "regular" Version 8 regex
routines, here are the pattern-matching rules not described above.

Any single character matches itself, unless it is a I<metacharacter>
with a special meaning described here or above.  You can cause
characters that normally function as metacharacters to be interpreted
literally by prefixing them with a "\" (e.g., "\." matches a ".", not any
character; "\\" matches a "\"). This escape mechanism is also required
for the character used as the pattern delimiter.

A series of characters matches that series of characters in the target
string, so the pattern C<blurfl> would match "blurfl" in the target
string.

You can specify a character class, by enclosing a list of characters
in C<[]>, which will match any character from the list.  If the
first character after the "[" is "^", the class matches any character not
in the list.  Within a list, the "-" character specifies a
range, so that C<a-z> represents all characters between "a" and "z",
inclusive.  If you want either "-" or "]" itself to be a member of a
class, put it at the start of the list (possibly after a "^"), or
escape it with a backslash.  "-" is also taken literally when it is
at the end of the list, just before the closing "]".  (The
following all specify the same class of three characters: C<[-az]>,
C<[az-]>, and C<[a\-z]>.  All are different from C<[a-z]>, which
specifies a class containing twenty-six characters, even on EBCDIC-based
character sets.)  Also, if you try to use the character
classes C<\w>, C<\W>, C<\s>, C<\S>, C<\d>, or C<\D> as endpoints of
a range, the "-" is understood literally.

Note also that the whole range idea is rather unportable between
character sets--and even within character sets they may cause results
you probably didn't expect.  A sound principle is to use only ranges
that begin from and end at either alphabetics of equal case ([a-e],
[A-E]), or digits ([0-9]).  Anything else is unsafe.  If in doubt,
spell out the character sets in full.

Characters may be specified using a metacharacter syntax much like that
used in C: "\n" matches a newline, "\t" a tab, "\r" a carriage return,
"\f" a form feed, etc.  More generally, \I<nnn>, where I<nnn> is a string
of three octal digits, matches the character whose coded character set value
is I<nnn>.  Similarly, \xI<nn>, where I<nn> are hexadecimal digits,
matches the character whose ordinal is I<nn>. The expression \cI<x>
matches the character control-I<x>.  Finally, the "." metacharacter
matches any character except "\n" (unless you use C</s>).

You can specify a series of alternatives for a pattern using "|" to
separate them, so that C<fee|fie|foe> will match any of "fee", "fie",
or "foe" in the target string (as would C<f(e|i|o)e>).  The
first alternative includes everything from the last pattern delimiter
("(", "(?:", etc. or the beginning of the pattern) up to the first "|", and
the last alternative contains everything from the last "|" to the next
closing pattern delimiter.  That's why it's common practice to include
alternatives in parentheses: to minimize confusion about where they
start and end.

Alternatives are tried from left to right, so the first
alternative found for which the entire expression matches, is the one that
is chosen. This means that alternatives are not necessarily greedy. For
example: when matching C<foo|foot> against "barefoot", only the "foo"
part will match, as that is the first alternative tried, and it successfully
matches the target string. (This might not seem important, but it is
important when you are capturing matched text using parentheses.)

Also remember that "|" is interpreted as a literal within square brackets,
so if you write C<[fee|fie|foe]> you're really only matching C<[feio|]>.

Within a pattern, you may designate subpatterns for later reference
by enclosing them in parentheses, and you may refer back to the
I<n>th subpattern later in the pattern using the metacharacter
\I<n> or \gI<n>.  Subpatterns are numbered based on the left to right order
of their opening parenthesis.  A backreference matches whatever
actually matched the subpattern in the string being examined, not
the rules for that subpattern.  Therefore, C<(0|0x)\d*\s\g1\d*> will
match "0x1234 0x4321", but not "0x1234 01234", because subpattern
1 matched "0x", even though the rule C<0|0x> could potentially match
the leading 0 in the second number.

=head2 Warning on \1 Instead of $1

Some people get too used to writing things like:

    $pattern =~ s/(\W)/\\\1/g;

This is grandfathered (for \1 to \9) for the RHS of a substitute to avoid
shocking the
B<sed> addicts, but it's a dirty habit to get into.  That's because in
PerlThink, the righthand side of an C<s///> is a double-quoted string.  C<\1> in
the usual double-quoted string means a control-A.  The customary Unix
meaning of C<\1> is kludged in for C<s///>.  However, if you get into the habit
of doing that, you get yourself into trouble if you then add an C</e>
modifier.

    s/(\d+)/ \1 + 1 /eg;            # causes warning under -w

Or if you try to do

    s/(\d+)/\1000/;

You can't disambiguate that by saying C<\{1}000>, whereas you can fix it with
C<${1}000>.  The operation of interpolation should not be confused
with the operation of matching a backreference.  Certainly they mean two
different things on the I<left> side of the C<s///>.

=head2 Repeated Patterns Matching a Zero-length Substring

B<WARNING>: Difficult material (and prose) ahead.  This section needs a rewrite.

Regular expressions provide a terse and powerful programming language.  As
with most other power tools, power comes together with the ability
to wreak havoc.

A common abuse of this power stems from the ability to make infinite
loops using regular expressions, with something as innocuous as:

    'foo' =~ m{ ( o? )* }x;

The C<o?> matches at the beginning of C<'foo'>, and since the position
in the string is not moved by the match, C<o?> would match again and again
because of the C<*> quantifier.  Another common way to create a similar cycle
is with the looping modifier C<//g>:

    @matches = ( 'foo' =~ m{ o? }xg );

or

    print "match: <$&>\n" while 'foo' =~ m{ o? }xg;

or the loop implied by split().

However, long experience has shown that many programming tasks may
be significantly simplified by using repeated subexpressions that
may match zero-length substrings.  Here's a simple example being:

    @chars = split //, $string;                  # // is not magic in split
    ($whitewashed = $string) =~ s/()/ /g; # parens avoid magic s// /

Thus Perl allows such constructs, by I<forcefully breaking
the infinite loop>.  The rules for this are different for lower-level
loops given by the greedy quantifiers C<*+{}>, and for higher-level
ones like the C</g> modifier or split() operator.

The lower-level loops are I<interrupted> (that is, the loop is
broken) when Perl detects that a repeated expression matched a
zero-length substring.   Thus

   m{ (?: NON_ZERO_LENGTH | ZERO_LENGTH )* }x;

is made equivalent to

   m{ (?: NON_ZERO_LENGTH )* (?: ZERO_LENGTH )? }x;

For example, this program

   #!perl -l
   "aaaaab" =~ /
     (?:
        a                 # non-zero
        |                 # or
       (?{print "hello"}) # print hello whenever this
                          #    branch is tried
       (?=(b))            # zero-width assertion
     )*  # any number of times
    /x;
   print $&;
   print $1;

prints

   hello
   aaaaa
   b

Notice that "hello" is only printed once, as when Perl sees that the sixth
iteration of the outermost C<(?:)*> matches a zero-length string, it stops
the C<*>.

The higher-level loops preserve an additional state between iterations:
whether the last match was zero-length.  To break the loop, the following
match after a zero-length match is prohibited to have a length of zero.
This prohibition interacts with backtracking (see L<"Backtracking">),
and so the I<second best> match is chosen if the I<best> match is of
zero length.

For example:

    $_ = 'bar';
    s/\w??/<$&>/g;

results in C<< <><b><><a><><r><> >>.  At each position of the string the best
match given by non-greedy C<??> is the zero-length match, and the I<second
best> match is what is matched by C<\w>.  Thus zero-length matches
alternate with one-character-long matches.

Similarly, for repeated C<m/()/g> the second-best match is the match at the
position one notch further in the string.

The additional state of being I<matched with zero-length> is associated with
the matched string, and is reset by each assignment to pos().
Zero-length matches at the end of the previous match are ignored
during C<split>.

=head2 Combining RE Pieces

Each of the elementary pieces of regular expressions which were described
before (such as C<ab> or C<\Z>) could match at most one substring
at the given position of the input string.  However, in a typical regular
expression these elementary pieces are combined into more complicated
patterns using combining operators C<ST>, C<S|T>, C<S*> etc.
(in these examples C<S> and C<T> are regular subexpressions).

Such combinations can include alternatives, leading to a problem of choice:
if we match a regular expression C<a|ab> against C<"abc">, will it match
substring C<"a"> or C<"ab">?  One way to describe which substring is
actually matched is the concept of backtracking (see L<"Backtracking">).
However, this description is too low-level and makes you think
in terms of a particular implementation.

Another description starts with notions of "better"/"worse".  All the
substrings which may be matched by the given regular expression can be
sorted from the "best" match to the "worst" match, and it is the "best"
match which is chosen.  This substitutes the question of "what is chosen?"
by the question of "which matches are better, and which are worse?".

Again, for elementary pieces there is no such question, since at most
one match at a given position is possible.  This section describes the
notion of better/worse for combining operators.  In the description
below C<S> and C<T> are regular subexpressions.

=over 4

=item C<ST>

Consider two possible matches, C<AB> and C<A'B'>, C<A> and C<A'> are
substrings which can be matched by C<S>, C<B> and C<B'> are substrings
which can be matched by C<T>.

If C<A> is a better match for C<S> than C<A'>, C<AB> is a better
match than C<A'B'>.

If C<A> and C<A'> coincide: C<AB> is a better match than C<AB'> if
C<B> is a better match for C<T> than C<B'>.

=item C<S|T>

When C<S> can match, it is a better match than when only C<T> can match.

Ordering of two matches for C<S> is the same as for C<S>.  Similar for
two matches for C<T>.

=item C<S{REPEAT_COUNT}>

Matches as C<SSS...S> (repeated as many times as necessary).

=item C<S{min,max}>

Matches as C<S{max}|S{max-1}|...|S{min+1}|S{min}>.

=item C<S{min,max}?>

Matches as C<S{min}|S{min+1}|...|S{max-1}|S{max}>.

=item C<S?>, C<S*>, C<S+>

Same as C<S{0,1}>, C<S{0,BIG_NUMBER}>, C<S{1,BIG_NUMBER}> respectively.

=item C<S??>, C<S*?>, C<S+?>

Same as C<S{0,1}?>, C<S{0,BIG_NUMBER}?>, C<S{1,BIG_NUMBER}?> respectively.

=item C<< (?>S) >>

Matches the best match for C<S> and only that.

=item C<(?=S)>, C<(?<=S)>

Only the best match for C<S> is considered.  (This is important only if
C<S> has capturing parentheses, and backreferences are used somewhere
else in the whole regular expression.)

=item C<(?!S)>, C<(?<!S)>

For this grouping operator there is no need to describe the ordering, since
only whether or not C<S> can match is important.

=item C<(??{ EXPR })>, C<(?PARNO)>

The ordering is the same as for the regular expression which is
the result of EXPR, or the pattern contained by capture group PARNO.

=item C<(?(condition)yes-pattern|no-pattern)>

Recall that which of C<yes-pattern> or C<no-pattern> actually matches is
already determined.  The ordering of the matches is the same as for the
chosen subexpression.

=back

The above recipes describe the ordering of matches I<at a given position>.
One more rule is needed to understand how a match is determined for the
whole regular expression: a match at an earlier position is always better
than a match at a later position.

=head2 Creating Custom RE Engines

As of Perl 5.10.0, one can create custom regular expression engines.  This
is not for the faint of heart, as they have to plug in at the C level.  See
L<perlreapi> for more details.

As an alternative, overloaded constants (see L<overload>) provide a simple
way to extend the functionality of the RE engine, by substituting one
pattern for another.

Suppose that we want to enable a new RE escape-sequence C<\Y|> which
matches at a boundary between whitespace characters and non-whitespace
characters.  Note that C<(?=\S)(?<!\S)|(?!\S)(?<=\S)> matches exactly
at these positions, so we want to have each C<\Y|> in the place of the
more complicated version.  We can create a module C<customre> to do
this:

    package customre;
    use overload;

    sub import {
      shift;
      die "No argument to customre::import allowed" if @_;
      overload::constant 'qr' => \&convert;
    }

    sub invalid { die "/$_[0]/: invalid escape '\\$_[1]'"}

    # We must also take care of not escaping the legitimate \\Y|
    # sequence, hence the presence of '\\' in the conversion rules.
    my %rules = ( '\\' => '\\\\',
                  'Y|' => qr/(?=\S)(?<!\S)|(?!\S)(?<=\S)/ );
    sub convert {
      my $re = shift;
      $re =~ s{
                \\ ( \\ | Y . )
              }
              { $rules{$1} or invalid($re,$1) }sgex;
      return $re;
    }

Now C<use customre> enables the new escape in constant regular
expressions, i.e., those without any runtime variable interpolations.
As documented in L<overload>, this conversion will work only over
literal parts of regular expressions.  For C<\Y|$re\Y|> the variable
part of this regular expression needs to be converted explicitly
(but only if the special meaning of C<\Y|> should be enabled inside $re):

    use customre;
    $re = <>;
    chomp $re;
    $re = customre::convert $re;
    /\Y|$re\Y|/;

=head2 PCRE/Python Support

As of Perl 5.10.0, Perl supports several Python/PCRE-specific extensions
to the regex syntax. While Perl programmers are encouraged to use the
Perl-specific syntax, the following are also accepted:

=over 4

=item C<< (?PE<lt>NAMEE<gt>pattern) >>

Define a named capture group. Equivalent to C<< (?<NAME>pattern) >>.

=item C<< (?P=NAME) >>

Backreference to a named capture group. Equivalent to C<< \g{NAME} >>.

=item C<< (?P>NAME) >>

Subroutine call to a named capture group. Equivalent to C<< (?&NAME) >>.

=back

=head1 BUGS

There are numerous problems with case-insensitive matching of characters
outside the ASCII range, especially with those whose folds are multiple
characters, such as ligatures like C<LATIN SMALL LIGATURE FF>.

In a bracketed character class with case-insensitive matching, ranges only work
for ASCII characters.  For example,
C<m/[\N{CYRILLIC CAPITAL LETTER A}-\N{CYRILLIC CAPITAL LETTER YA}]/i>
doesn't match all the Russian upper and lower case letters.

Many regular expression constructs don't work on EBCDIC platforms.

This document varies from difficult to understand to completely
and utterly opaque.  The wandering prose riddled with jargon is
hard to fathom in several places.

This document needs a rewrite that separates the tutorial content
from the reference content.

=head1 SEE ALSO

L<perlrequick>.

L<perlretut>.

L<perlop/"Regexp Quote-Like Operators">.

L<perlop/"Gory details of parsing quoted constructs">.

L<perlfaq6>.

L<perlfunc/pos>.

L<perllocale>.

L<perlebcdic>.

I<Mastering Regular Expressions> by Jeffrey Friedl, published
by O'Reilly and Associates.