This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perlre.pod: Update for 5.14
[perl5.git] / pod / perlrecharclass.pod
CommitLineData
8a118206 1=head1 NAME
ea449505 2X<character class>
8a118206
RGS
3
4perlrecharclass - Perl Regular Expression Character Classes
5
6=head1 DESCRIPTION
7
8The top level documentation about Perl regular expressions
9is found in L<perlre>.
10
11This manual page discusses the syntax and use of character
6b83a163 12classes in Perl regular expressions.
8a118206 13
6b83a163 14A character class is a way of denoting a set of characters
8a118206 15in such a way that one character of the set is matched.
6b83a163 16It's important to remember that: matching a character class
8a118206
RGS
17consumes exactly one character in the source string. (The source
18string is the string the regular expression is matched against.)
19
20There are three types of character classes in Perl regular
6b83a163 21expressions: the dot, backslash sequences, and the form enclosed in square
ea449505 22brackets. Keep in mind, though, that often the term "character class" is used
6b83a163 23to mean just the bracketed form. Certainly, most Perl documentation does that.
8a118206
RGS
24
25=head2 The dot
26
27The dot (or period), C<.> is probably the most used, and certainly
28the most well-known character class. By default, a dot matches any
29character, except for the newline. The default can be changed to
6b83a163
KW
30add matching the newline by using the I<single line> modifier: either
31for the entire regular expression with the C</s> modifier, or
32locally with C<(?s)>. (The experimental C<\N> backslash sequence, described
33below, matches any character except newline without regard to the
34I<single line> modifier.)
8a118206
RGS
35
36Here are some examples:
37
38 "a" =~ /./ # Match
39 "." =~ /./ # Match
40 "" =~ /./ # No match (dot has to match a character)
41 "\n" =~ /./ # No match (dot does not match a newline)
42 "\n" =~ /./s # Match (global 'single line' modifier)
43 "\n" =~ /(?s:.)/ # Match (local 'single line' modifier)
44 "ab" =~ /^.$/ # No match (dot matches one character)
45
6b83a163 46=head2 Backslash sequences
ea449505
KW
47X<\w> X<\W> X<\s> X<\S> X<\d> X<\D> X<\p> X<\P>
48X<\N> X<\v> X<\V> X<\h> X<\H>
49X<word> X<whitespace>
8a118206 50
6b83a163
KW
51A backslash sequence is a sequence of characters, the first one of which is a
52backslash. Perl ascribes special meaning to many such sequences, and some of
53these are character classes. That is, they match a single character each,
54provided that the character belongs to the specific set of characters defined
55by the sequence.
8a118206 56
6b83a163
KW
57Here's a list of the backslash sequences that are character classes. They
58are discussed in more detail below. (For the backslash sequences that aren't
59character classes, see L<perlrebackslash>.)
8a118206 60
6b83a163
KW
61 \d Match a decimal digit character.
62 \D Match a non-decimal-digit character.
8a118206
RGS
63 \w Match a "word" character.
64 \W Match a non-"word" character.
ea449505
KW
65 \s Match a whitespace character.
66 \S Match a non-whitespace character.
67 \h Match a horizontal whitespace character.
68 \H Match a character that isn't horizontal whitespace.
ea449505
KW
69 \v Match a vertical whitespace character.
70 \V Match a character that isn't vertical whitespace.
6b83a163
KW
71 \N Match a character that isn't a newline. Experimental.
72 \pP, \p{Prop} Match a character that has the given Unicode property.
6c5a041f 73 \PP, \P{Prop} Match a character that doesn't have the Unicode property
8a118206
RGS
74
75=head3 Digits
76
b6538e4f 77C<\d> matches a single character considered to be a decimal I<digit>.
17657a39
KW
78What is considered a decimal digit depends on several factors, detailed
79below in L</Locale, EBCDIC, Unicode and UTF-8>. If those factors
80indicate a Unicode interpretation, C<\d> not only matches the digits
b6538e4f
TC
81'0' - '9', but also Arabic, Devanagari, and digits from other languages.
82Otherwise, if a locale is in effect, it matches whatever characters that
83locale considers decimal digits. Only when neither a Unicode interpretation
84nor locale prevails does C<\d> match only the digits '0' to '9' alone.
6b83a163
KW
85
86Unicode digits may cause some confusion, and some security issues. In UTF-8
409a7f61 87strings, unless the C</a> regular expression modifier is specified,
f7d1198f 88C<\d> matches the same characters matched by
6b83a163
KW
89C<\p{General_Category=Decimal_Number}>, or synonymously,
90C<\p{General_Category=Digit}>. Starting with Unicode version 4.1, this is the
91same set of characters matched by C<\p{Numeric_Type=Decimal}>.
92
93But Unicode also has a different property with a similar name,
94C<\p{Numeric_Type=Digit}>, which matches a completely different set of
95characters. These characters are things such as subscripts.
96
b6538e4f 97The design intent is for C<\d> to match all digits (and no other characters)
6b83a163
KW
98that can be used with "normal" big-endian positional decimal syntax, whereby a
99sequence of such digits {N0, N1, N2, ...Nn} has the numeric value (...(N0 * 10
100+ N1) * 10 + N2) * 10 ... + Nn). In Unicode 5.2, the Tamil digits (U+0BE6 -
101U+0BEF) can also legally be used in old-style Tamil numbers in which they would
102appear no more than one in a row, separated by characters that mean "times 10",
103"times 100", etc. (See L<http://www.unicode.org/notes/tn21>.)
104
b6538e4f 105Some non-European digits that C<\d> matches look like European ones, but
6671dd37
KW
106have different values. For example, BENGALI DIGIT FOUR (U+09EA) looks
107very much like an ASCII DIGIT EIGHT (U+0038).
6b83a163
KW
108
109It may be useful for security purposes for an application to require that all
bf7786d4
KW
110digits in a row be from the same script. This can be checked by using
111L<Unicode::UCD/num()>.
8a118206 112
b6538e4f 113Any character not matched by C<\d> is matched by C<\D>.
8a118206
RGS
114
115=head3 Word characters
116
ea449505 117A C<\w> matches a single alphanumeric character (an alphabetic character, or a
d35dd6c6
KW
118decimal digit) or a connecting punctuation character, such as an
119underscore ("_"). It does not match a whole word. To match a whole
6b83a163 120word, use C<\w+>. This isn't the same thing as matching an English word, but
765fa144 121in the ASCII range it is the same as a string of Perl-identifier
d35dd6c6 122characters. What is considered a
17657a39
KW
123word character depends on several factors, detailed below in L</Locale,
124EBCDIC, Unicode and UTF-8>. If those factors indicate a Unicode
b6538e4f 125interpretation, C<\w> matches the characters considered word
ea449505 126characters in the Unicode database. That is, it not only matches ASCII letters,
d35dd6c6
KW
127but also Thai letters, Greek letters, etc. This includes connector
128punctuation (like the underscore) which connect two words together, or
b6538e4f
TC
129diacritics, such as a C<COMBINING TILDE> and the modifier letters, which
130are generally used to add auxiliary markings to letters. If a Unicode
131interpretation is not indicated, C<\w> matches those characters considered
17657a39 132word characters by the current locale or EBCDIC code page. Without a
b6538e4f
TC
133locale or EBCDIC code page, C<\w> matches the underscore and ASCII letters
134and digits.
8a118206 135
6b83a163
KW
136There are a number of security issues with the full Unicode list of word
137characters. See L<http://unicode.org/reports/tr36>.
138
139Also, for a somewhat finer-grained set of characters that are in programming
140language identifiers beyond the ASCII range, you may wish to instead use the
141more customized Unicode properties, "ID_Start", ID_Continue", "XID_Start", and
142"XID_Continue". See L<http://unicode.org/reports/tr31>.
143
b6538e4f 144Any character not matched by C<\w> is matched by C<\W>.
8a118206 145
ea449505
KW
146=head3 Whitespace
147
b6538e4f 148C<\s> matches any single character considered whitespace. The exact
17657a39
KW
149set of characters matched by C<\s> depends on several factors, detailed
150below in L</Locale, EBCDIC, Unicode and UTF-8>. If those factors
151indicate a Unicode interpretation, C<\s> matches what is considered
152whitespace in the Unicode database; the complete list is in the table
b6538e4f 153below. Otherwise, if a locale or EBCDIC code page is in effect,
17657a39
KW
154C<\s> matches whatever is considered whitespace by the current locale or
155EBCDIC code page. Without a locale or EBCDIC code page, C<\s> matches
156the horizontal tab (C<\t>), the newline (C<\n>), the form feed (C<\f>),
157the carriage return (C<\r>), and the space. (Note that it doesn't match
158the vertical tab, C<\cK>.) Perhaps the most notable possible surprise
b6538e4f 159is that C<\s> matches a non-breaking space B<only> if a Unicode
17657a39 160interpretation is indicated, or the locale or EBCDIC code page that is
b6538e4f 161in effect happens to have that character.
8a118206 162
b6538e4f 163Any character not matched by C<\s> is matched by C<\S>.
8a118206 164
b6538e4f
TC
165C<\h> matches any character considered horizontal whitespace;
166this includes the space and tab characters and several others
167listed in the table below. C<\H> matches any character
168not considered horizontal whitespace.
ea449505 169
b6538e4f
TC
170C<\v> matches any character considered vertical whitespace;
171this includes the carriage return and line feed characters (newline)
172plus several other characters, all listed in the table below.
173C<\V> matches any character not considered vertical whitespace.
8a118206
RGS
174
175C<\R> matches anything that can be considered a newline under Unicode
176rules. It's not a character class, as it can match a multi-character
177sequence. Therefore, it cannot be used inside a bracketed character
ea449505
KW
178class; use C<\v> instead (vertical whitespace).
179Details are discussed in L<perlrebackslash>.
8a118206
RGS
180
181Note that unlike C<\s>, C<\d> and C<\w>, C<\h> and C<\v> always match
b6538e4f
TC
182the same characters, without regard to other factors, such as whether the
183source string is in UTF-8 format.
8a118206 184
ea449505
KW
185One might think that C<\s> is equivalent to C<[\h\v]>. This is not true. The
186vertical tab (C<"\x0b">) is not matched by C<\s>, it is however considered
187vertical whitespace. Furthermore, if the source string is not in UTF-8 format,
188and any locale or EBCDIC code page that is in effect doesn't include them, the
6b83a163
KW
189next line (ASCII-platform C<"\x85">) and the no-break space (ASCII-platform
190C<"\xA0">) characters are not matched by C<\s>, but are by C<\v> and C<\h>
409a7f61 191respectively. If the C</a> modifier is not in effect and the source
b6538e4f
TC
192string is in UTF-8 format, both the next line and the no-break space
193are matched by C<\s>.
8a118206
RGS
194
195The following table is a complete listing of characters matched by
ea449505 196C<\s>, C<\h> and C<\v> as of Unicode 5.2.
8a118206
RGS
197
198The first column gives the code point of the character (in hex format),
199the second column gives the (Unicode) name. The third column indicates
ea449505
KW
200by which class(es) the character is matched (assuming no locale or EBCDIC code
201page is in effect that changes the C<\s> matching).
8a118206
RGS
202
203 0x00009 CHARACTER TABULATION h s
204 0x0000a LINE FEED (LF) vs
205 0x0000b LINE TABULATION v
206 0x0000c FORM FEED (FF) vs
207 0x0000d CARRIAGE RETURN (CR) vs
208 0x00020 SPACE h s
209 0x00085 NEXT LINE (NEL) vs [1]
210 0x000a0 NO-BREAK SPACE h s [1]
211 0x01680 OGHAM SPACE MARK h s
212 0x0180e MONGOLIAN VOWEL SEPARATOR h s
213 0x02000 EN QUAD h s
214 0x02001 EM QUAD h s
215 0x02002 EN SPACE h s
216 0x02003 EM SPACE h s
217 0x02004 THREE-PER-EM SPACE h s
218 0x02005 FOUR-PER-EM SPACE h s
219 0x02006 SIX-PER-EM SPACE h s
220 0x02007 FIGURE SPACE h s
221 0x02008 PUNCTUATION SPACE h s
222 0x02009 THIN SPACE h s
223 0x0200a HAIR SPACE h s
224 0x02028 LINE SEPARATOR vs
225 0x02029 PARAGRAPH SEPARATOR vs
226 0x0202f NARROW NO-BREAK SPACE h s
227 0x0205f MEDIUM MATHEMATICAL SPACE h s
228 0x03000 IDEOGRAPHIC SPACE h s
229
230=over 4
231
232=item [1]
233
234NEXT LINE and NO-BREAK SPACE only match C<\s> if the source string is in
409a7f61 235UTF-8 format and the C</a> modifier is not in effect, or if the locale
b6538e4f 236or EBCDIC code page in effect includes them.
8a118206
RGS
237
238=back
239
240It is worth noting that C<\d>, C<\w>, etc, match single characters, not
e486b3cc 241complete numbers or words. To match a number (that consists of digits),
8a118206
RGS
242use C<\d+>; to match a word, use C<\w+>.
243
6b83a163
KW
244=head3 \N
245
b6538e4f 246C<\N> is new in 5.12, and is experimental. It, like the dot, matches any
6b83a163
KW
247character that is not a newline. The difference is that C<\N> is not influenced
248by the I<single line> regular expression modifier (see L</The dot> above). Note
249that the form C<\N{...}> may mean something completely different. When the
250C<{...}> is a L<quantifier|perlre/Quantifiers>, it means to match a non-newline
251character that many times. For example, C<\N{3}> means to match 3
252non-newlines; C<\N{5,}> means to match 5 or more non-newlines. But if C<{...}>
253is not a legal quantifier, it is presumed to be a named character. See
254L<charnames> for those. For example, none of C<\N{COLON}>, C<\N{4F}>, and
255C<\N{F4}> contain legal quantifiers, so Perl will try to find characters whose
b6538e4f 256names are respectively C<COLON>, C<4F>, and C<F4>.
8a118206
RGS
257
258=head3 Unicode Properties
259
c1c4ae3a
KW
260C<\pP> and C<\p{Prop}> are character classes to match characters that fit given
261Unicode properties. One letter property names can be used in the C<\pP> form,
262with the property name following the C<\p>, otherwise, braces are required.
263When using braces, there is a single form, which is just the property name
264enclosed in the braces, and a compound form which looks like C<\p{name=value}>,
b6538e4f 265which means to match if the property "name" for the character has that particular
c1c4ae3a 266"value".
e1b711da
KW
267For instance, a match for a number can be written as C</\pN/> or as
268C</\p{Number}/>, or as C</\p{Number=True}/>.
269Lowercase letters are matched by the property I<Lowercase_Letter> which
270has as short form I<Ll>. They need the braces, so are written as C</\p{Ll}/> or
271C</\p{Lowercase_Letter}/>, or C</\p{General_Category=Lowercase_Letter}/>
272(the underscores are optional).
273C</\pLl/> is valid, but means something different.
8a118206
RGS
274It matches a two character string: a letter (Unicode property C<\pL>),
275followed by a lowercase C<l>.
276
56ca34ca
KW
277Note that almost all properties are immune to case-insensitive matching.
278That is, adding a C</i> regular expression modifier does not change what
b6538e4f 279they match. There are two sets affected. The first set is
56ca34ca
KW
280C<Uppercase_Letter>,
281C<Lowercase_Letter>,
282and C<Titlecase_Letter>,
283all of which match C<Cased_Letter> under C</i> matching.
b6538e4f 284The second set is
56ca34ca
KW
285C<Uppercase>,
286C<Lowercase>,
287and C<Titlecase>,
288all of which match C<Cased> under C</i> matching.
289(The difference between these sets is that some things, such as Roman
b6538e4f
TC
290Numerals, come in both upper and lower case so they are C<Cased>, but
291aren't considered to be letters, so they aren't C<Cased_Letter>s. They're
292actually C<Letter_Number>s.)
293This set also includes its subsets C<PosixUpper> and C<PosixLower>, both
56ca34ca
KW
294of which under C</i> matching match C<PosixAlpha>.
295
296For more details on Unicode properties, see L<perlunicode/Unicode
297Character Properties>; for a
e1b711da 298complete list of possible properties, see
56ca34ca
KW
299L<perluniprops/Properties accessible through \p{} and \P{}>,
300which notes all forms that have C</i> differences.
e1b711da 301It is also possible to define your own properties. This is discussed in
8a118206
RGS
302L<perlunicode/User-Defined Character Properties>.
303
8a118206
RGS
304=head4 Examples
305
306 "a" =~ /\w/ # Match, "a" is a 'word' character.
307 "7" =~ /\w/ # Match, "7" is a 'word' character as well.
308 "a" =~ /\d/ # No match, "a" isn't a digit.
309 "7" =~ /\d/ # Match, "7" is a digit.
ea449505 310 " " =~ /\s/ # Match, a space is whitespace.
8a118206
RGS
311 "a" =~ /\D/ # Match, "a" is a non-digit.
312 "7" =~ /\D/ # No match, "7" is not a non-digit.
ea449505 313 " " =~ /\S/ # No match, a space is not non-whitespace.
8a118206 314
ea449505
KW
315 " " =~ /\h/ # Match, space is horizontal whitespace.
316 " " =~ /\v/ # No match, space is not vertical whitespace.
317 "\r" =~ /\v/ # Match, a return is vertical whitespace.
8a118206
RGS
318
319 "a" =~ /\pL/ # Match, "a" is a letter.
320 "a" =~ /\p{Lu}/ # No match, /\p{Lu}/ matches upper case letters.
321
322 "\x{0e0b}" =~ /\p{Thai}/ # Match, \x{0e0b} is the character
323 # 'THAI CHARACTER SO SO', and that's in
324 # Thai Unicode class.
ea449505 325 "a" =~ /\P{Lao}/ # Match, as "a" is not a Laotian character.
8a118206
RGS
326
327
328=head2 Bracketed Character Classes
329
330The third form of character class you can use in Perl regular expressions
6b83a163 331is the bracketed character class. In its simplest form, it lists the characters
c1c4ae3a 332that may be matched, surrounded by square brackets, like this: C<[aeiou]>.
ea449505 333This matches one of C<a>, C<e>, C<i>, C<o> or C<u>. Like the other
b6538e4f 334character classes, exactly one character is matched. To match
ea449505 335a longer string consisting of characters mentioned in the character
6b83a163 336class, follow the character class with a L<quantifier|perlre/Quantifiers>. For
b6538e4f 337instance, C<[aeiou]+> matches one or more lowercase English vowels.
8a118206
RGS
338
339Repeating a character in a character class has no
340effect; it's considered to be in the set only once.
341
342Examples:
343
344 "e" =~ /[aeiou]/ # Match, as "e" is listed in the class.
345 "p" =~ /[aeiou]/ # No match, "p" is not listed in the class.
346 "ae" =~ /^[aeiou]$/ # No match, a character class only matches
347 # a single character.
348 "ae" =~ /^[aeiou]+$/ # Match, due to the quantifier.
349
350=head3 Special Characters Inside a Bracketed Character Class
351
352Most characters that are meta characters in regular expressions (that
df225385 353is, characters that carry a special meaning like C<.>, C<*>, or C<(>) lose
8a118206
RGS
354their special meaning and can be used inside a character class without
355the need to escape them. For instance, C<[()]> matches either an opening
356parenthesis, or a closing parenthesis, and the parens inside the character
357class don't group or capture.
358
359Characters that may carry a special meaning inside a character class are:
360C<\>, C<^>, C<->, C<[> and C<]>, and are discussed below. They can be
361escaped with a backslash, although this is sometimes not needed, in which
362case the backslash may be omitted.
363
364The sequence C<\b> is special inside a bracketed character class. While
6b83a163 365outside the character class, C<\b> is an assertion indicating a point
8a118206
RGS
366that does not have either two word characters or two non-word characters
367on either side, inside a bracketed character class, C<\b> matches a
368backspace character.
369
df225385
KW
370The sequences
371C<\a>,
372C<\c>,
373C<\e>,
374C<\f>,
375C<\n>,
e526e8bb 376C<\N{I<NAME>}>,
765fa144 377C<\N{U+I<hex char>}>,
df225385
KW
378C<\r>,
379C<\t>,
380and
381C<\x>
06ee63cd
KW
382are also special and have the same meanings as they do outside a
383bracketed character class. (However, inside a bracketed character
384class, if C<\N{I<NAME>}> expands to a sequence of characters, only the first
385one in the sequence is used, with a warning.)
df225385 386
ea449505
KW
387Also, a backslash followed by two or three octal digits is considered an octal
388number.
df225385 389
6b83a163
KW
390A C<[> is not special inside a character class, unless it's the start of a
391POSIX character class (see L</POSIX Character Classes> below). It normally does
392not need escaping.
8a118206 393
6b83a163
KW
394A C<]> is normally either the end of a POSIX character class (see
395L</POSIX Character Classes> below), or it signals the end of the bracketed
396character class. If you want to include a C<]> in the set of characters, you
397must generally escape it.
b6538e4f 398
8a118206
RGS
399However, if the C<]> is the I<first> (or the second if the first
400character is a caret) character of a bracketed character class, it
401does not denote the end of the class (as you cannot have an empty class)
402and is considered part of the set of characters that can be matched without
403escaping.
404
405Examples:
406
407 "+" =~ /[+?*]/ # Match, "+" in a character class is not special.
408 "\cH" =~ /[\b]/ # Match, \b inside in a character class
c1c4ae3a 409 # is equivalent to a backspace.
8a118206
RGS
410 "]" =~ /[][]/ # Match, as the character class contains.
411 # both [ and ].
412 "[]" =~ /[[]]/ # Match, the pattern contains a character class
413 # containing just ], and the character class is
414 # followed by a ].
415
416=head3 Character Ranges
417
418It is not uncommon to want to match a range of characters. Luckily, instead
b6538e4f 419of listing all characters in the range, one may use the hyphen (C<->).
8a118206 420If inside a bracketed character class you have two characters separated
b6538e4f 421by a hyphen, it's treated as if all characters between the two were in
8a118206 422the class. For instance, C<[0-9]> matches any ASCII digit, and C<[a-m]>
b6538e4f 423matches any lowercase letter from the first half of the old ASCII alphabet.
8a118206
RGS
424
425Note that the two characters on either side of the hyphen are not
765fa144 426necessarily both letters or both digits. Any character is possible,
8a118206 427although not advisable. C<['-?]> contains a range of characters, but
b6538e4f 428most people will not know which characters that means. Furthermore,
8a118206
RGS
429such ranges may lead to portability problems if the code has to run on
430a platform that uses a different character set, such as EBCDIC.
431
ea449505
KW
432If a hyphen in a character class cannot syntactically be part of a range, for
433instance because it is the first or the last character of the character class,
b6538e4f
TC
434or if it immediately follows a range, the hyphen isn't special, and so is
435considered a character to be matched literally. If you want a hyphen in
436your set of characters to be matched and its position in the class is such
437that it could be considered part of a range, you must escape that hyphen
438with a backslash.
8a118206
RGS
439
440Examples:
441
442 [a-z] # Matches a character that is a lower case ASCII letter.
c1c4ae3a
KW
443 [a-fz] # Matches any letter between 'a' and 'f' (inclusive) or
444 # the letter 'z'.
8a118206
RGS
445 [-z] # Matches either a hyphen ('-') or the letter 'z'.
446 [a-f-m] # Matches any letter between 'a' and 'f' (inclusive), the
447 # hyphen ('-'), or the letter 'm'.
448 ['-?] # Matches any of the characters '()*+,-./0123456789:;<=>?
449 # (But not on an EBCDIC platform).
450
451
452=head3 Negation
453
454It is also possible to instead list the characters you do not want to
455match. You can do so by using a caret (C<^>) as the first character in the
b6538e4f
TC
456character class. For instance, C<[^a-z]> matches any character that is not a
457lowercase ASCII letter, which therefore includes almost a hundred thousand
458Unicode letters.
8a118206
RGS
459
460This syntax make the caret a special character inside a bracketed character
461class, but only if it is the first character of the class. So if you want
b6538e4f
TC
462the caret as one of the characters to match, either escape the caret or
463else not list it first.
8a118206
RGS
464
465Examples:
466
467 "e" =~ /[^aeiou]/ # No match, the 'e' is listed.
468 "x" =~ /[^aeiou]/ # Match, as 'x' isn't a lowercase vowel.
469 "^" =~ /[^^]/ # No match, matches anything that isn't a caret.
470 "^" =~ /[x^]/ # Match, caret is not special here.
471
472=head3 Backslash Sequences
473
ea449505 474You can put any backslash sequence character class (with the exception of
765fa144 475C<\N> and C<\R>) inside a bracketed character class, and it will act just
b6538e4f
TC
476as if you had put all characters matched by the backslash sequence inside the
477character class. For instance, C<[a-f\d]> matches any decimal digit, or any
6b83a163
KW
478of the lowercase letters between 'a' and 'f' inclusive.
479
480C<\N> within a bracketed character class must be of the forms C<\N{I<name>}>
765fa144 481or C<\N{U+I<hex char>}>, and NOT be the form that matches non-newlines,
6b83a163
KW
482for the same reason that a dot C<.> inside a bracketed character class loses
483its special meaning: it matches nearly anything, which generally isn't what you
484want to happen.
df225385 485
8a118206
RGS
486
487Examples:
488
489 /[\p{Thai}\d]/ # Matches a character that is either a Thai
490 # character, or a digit.
491 /[^\p{Arabic}()]/ # Matches a character that is neither an Arabic
492 # character, nor a parenthesis.
493
494Backslash sequence character classes cannot form one of the endpoints
6b83a163
KW
495of a range. Thus, you can't say:
496
497 /[\p{Thai}-\d]/ # Wrong!
8a118206 498
6b83a163 499=head3 POSIX Character Classes
ea449505 500X<character class> X<\p> X<\p{}>
ea449505
KW
501X<alpha> X<alnum> X<ascii> X<blank> X<cntrl> X<digit> X<graph>
502X<lower> X<print> X<punct> X<space> X<upper> X<word> X<xdigit>
8a118206 503
6b83a163
KW
504POSIX character classes have the form C<[:class:]>, where I<class> is
505name, and the C<[:> and C<:]> delimiters. POSIX character classes only appear
8a118206 506I<inside> bracketed character classes, and are a convenient and descriptive
f7d1198f 507way of listing a group of characters, though they can suffer from
6b83a163
KW
508portability issues (see below and L<Locale, EBCDIC, Unicode and UTF-8>).
509
510Be careful about the syntax,
8a118206
RGS
511
512 # Correct:
513 $string =~ /[[:alpha:]]/
514
515 # Incorrect (will warn):
516 $string =~ /[:alpha:]/
517
518The latter pattern would be a character class consisting of a colon,
519and the letters C<a>, C<l>, C<p> and C<h>.
b6538e4f
TC
520POSIX character classes can be part of a larger bracketed character class.
521For example,
ea449505
KW
522
523 [01[:alpha:]%]
524
525is valid and matches '0', '1', any alphabetic character, and the percent sign.
8a118206
RGS
526
527Perl recognizes the following POSIX character classes:
528
ea449505 529 alpha Any alphabetical character ("[A-Za-z]").
b6538e4f 530 alnum Any alphanumeric character. ("[A-Za-z0-9]")
ea449505 531 ascii Any character in the ASCII character set.
ea8b8ad2 532 blank A GNU extension, equal to a space or a horizontal tab ("\t").
ea449505
KW
533 cntrl Any control character. See Note [2] below.
534 digit Any decimal digit ("[0-9]"), equivalent to "\d".
535 graph Any printable character, excluding a space. See Note [3] below.
536 lower Any lowercase character ("[a-z]").
537 print Any printable character, including a space. See Note [4] below.
c1c4ae3a 538 punct Any graphical character excluding "word" characters. Note [5].
ea449505
KW
539 space Any whitespace character. "\s" plus the vertical tab ("\cK").
540 upper Any uppercase character ("[A-Z]").
541 word A Perl extension ("[A-Za-z0-9_]"), equivalent to "\w".
542 xdigit Any hexadecimal digit ("[0-9a-fA-F]").
543
544Most POSIX character classes have two Unicode-style C<\p> property
545counterparts. (They are not official Unicode properties, but Perl extensions
546derived from official Unicode properties.) The table below shows the relation
547between POSIX character classes and these counterparts.
548
549One counterpart, in the column labelled "ASCII-range Unicode" in
b6538e4f 550the table, matches only characters in the ASCII character set.
ea449505
KW
551
552The other counterpart, in the column labelled "Full-range Unicode", matches any
553appropriate characters in the full Unicode character set. For example,
b6538e4f
TC
554C<\p{Alpha}> matches not just the ASCII alphabetic characters, but any
555character in the entire Unicode character set considered alphabetic.
556The column labelled "backslash sequence" is a (short) synonym for
cbc24f92 557the Full-range Unicode form.
ea449505
KW
558
559(Each of the counterparts has various synonyms as well.
b6538e4f
TC
560L<perluniprops/Properties accessible through \p{} and \P{}> lists all
561synonyms, plus all characters matched by each ASCII-range property.
562For example, C<\p{AHex}> is a synonym for C<\p{ASCII_Hex_Digit}>,
ea449505
KW
563and any C<\p> property name can be prefixed with "Is" such as C<\p{IsAlpha}>.)
564
b6538e4f 565Both the C<\p> forms are unaffected by any locale in effect, or whether
ea449505 566the string is in UTF-8 format or not, or whether the platform is EBCDIC or not.
f7d1198f 567In contrast, the POSIX character classes are affected, unless the
409a7f61 568regular expression is compiled with the C</a> modifier. If the C</a>
f7d1198f
KW
569modifier is not in effect, and the source string is in UTF-8 format, the
570POSIX classes behave like their "Full-range" Unicode counterparts. If
409a7f61 571C</a> modifier is in effect; or the source string is not in UTF-8
f7d1198f
KW
572format, and no locale is in effect, and the platform is not EBCDIC, all
573the POSIX classes behave like their ASCII-range counterparts.
574Otherwise, they behave based on the rules of the locale or EBCDIC code
575page.
6b83a163 576
ea449505 577It is proposed to change this behavior in a future release of Perl so that the
765fa144 578the UTF-8-ness of the source string will be irrelevant to the behavior of the
ea449505
KW
579POSIX character classes. This means they will always behave in strict
580accordance with the official POSIX standard. That is, if either locale or
581EBCDIC code page is present, they will behave in accordance with those; if
582absent, the classes will match only their ASCII-range counterparts. If you
765fa144 583wish to comment on this proposal, send email to C<perl5-porters@perl.org>.
ea449505 584
cbc24f92
KW
585 [[:...:]] ASCII-range Full-range backslash Note
586 Unicode Unicode sequence
ea449505 587 -----------------------------------------------------
cbc24f92
KW
588 alpha \p{PosixAlpha} \p{XPosixAlpha}
589 alnum \p{PosixAlnum} \p{XPosixAlnum}
ea449505 590 ascii \p{ASCII}
cbc24f92
KW
591 blank \p{PosixBlank} \p{XPosixBlank} \h [1]
592 or \p{HorizSpace} [1]
593 cntrl \p{PosixCntrl} \p{XPosixCntrl} [2]
594 digit \p{PosixDigit} \p{XPosixDigit} \d
595 graph \p{PosixGraph} \p{XPosixGraph} [3]
596 lower \p{PosixLower} \p{XPosixLower}
597 print \p{PosixPrint} \p{XPosixPrint} [4]
598 punct \p{PosixPunct} \p{XPosixPunct} [5]
599 \p{PerlSpace} \p{XPerlSpace} \s [6]
600 space \p{PosixSpace} \p{XPosixSpace} [6]
601 upper \p{PosixUpper} \p{XPosixUpper}
602 word \p{PosixWord} \p{XPosixWord} \w
603 xdigit \p{ASCII_Hex_Digit} \p{XPosixXDigit}
8a118206
RGS
604
605=over 4
606
ea449505
KW
607=item [1]
608
609C<\p{Blank}> and C<\p{HorizSpace}> are synonyms.
610
611=item [2]
8a118206 612
ea449505 613Control characters don't produce output as such, but instead usually control
b6538e4f 614the terminal somehow: for example, newline and backspace are control characters.
ea449505
KW
615In the ASCII range, characters whose ordinals are between 0 and 31 inclusive,
616plus 127 (C<DEL>) are control characters.
8a118206 617
c1c4ae3a
KW
618On EBCDIC platforms, it is likely that the code page will define C<[[:cntrl:]]>
619to be the EBCDIC equivalents of the ASCII controls, plus the controls
6b83a163 620that in Unicode have ordinals from 128 through 159.
ea449505
KW
621
622=item [3]
8a118206
RGS
623
624Any character that is I<graphical>, that is, visible. This class consists
b6538e4f 625of all alphanumeric characters and all punctuation characters.
8a118206 626
ea449505 627=item [4]
8a118206 628
b6538e4f
TC
629All printable characters, which is the set of all graphical characters
630plus those whitespace characters which are not also controls.
ea449505 631
b6dac59a 632=item [5]
ea449505 633
b6538e4f 634C<\p{PosixPunct}> and C<[[:punct:]]> in the ASCII range match all
ea449505
KW
635non-controls, non-alphanumeric, non-space characters:
636C<[-!"#$%&'()*+,./:;<=E<gt>?@[\\\]^_`{|}~]> (although if a locale is in effect,
637it could alter the behavior of C<[[:punct:]]>).
638
cbc24f92
KW
639The similarly named property, C<\p{Punct}>, matches a somewhat different
640set in the ASCII range, namely
6c5a041f
KW
641C<[-!"#%&'()*,./:;?@[\\\]_{}]>. That is, it is missing C<[$+E<lt>=E<gt>^`|~]>.
642This is because Unicode splits what POSIX considers to be punctuation into two
643categories, Punctuation and Symbols.
644
765fa144
KW
645C<\p{XPosixPunct}> and (in Unicode mode) C<[[:punct:]]>, match what
646C<\p{PosixPunct}> matches in the ASCII range, plus what C<\p{Punct}>
647matches. This is different than strictly matching according to
648C<\p{Punct}>. Another way to say it is that
b6538e4f
TC
649for a UTF-8 string, C<[[:punct:]]> matches all characters that Unicode
650considers punctuation, plus all ASCII-range characters that Unicode
651considers symbols.
8a118206 652
ea449505 653=item [6]
8a118206 654
ea449505
KW
655C<\p{SpacePerl}> and C<\p{Space}> differ only in that C<\p{Space}> additionally
656matches the vertical tab, C<\cK>. Same for the two ASCII-only range forms.
8a118206
RGS
657
658=back
659
cbc24f92 660There are various other synonyms that can be used for these besides
b6538e4f 661C<\p{HorizSpace}> and \C<\p{XPosixBlank}>. For example,
cbc24f92
KW
662C<\p{PosixAlpha}> can be written as C<\p{Alpha}>. All are listed
663in L<perluniprops/Properties accessible through \p{} and \P{}>.
664
8a118206 665=head4 Negation
ea449505 666X<character class, negation>
8a118206
RGS
667
668A Perl extension to the POSIX character class is the ability to
669negate it. This is done by prefixing the class name with a caret (C<^>).
670Some examples:
671
ea449505
KW
672 POSIX ASCII-range Full-range backslash
673 Unicode Unicode sequence
674 -----------------------------------------------------
cbc24f92
KW
675 [[:^digit:]] \P{PosixDigit} \P{XPosixDigit} \D
676 [[:^space:]] \P{PosixSpace} \P{XPosixSpace}
677 \P{PerlSpace} \P{XPerlSpace} \S
678 [[:^word:]] \P{PerlWord} \P{XPosixWord} \W
679
765fa144
KW
680The backslash sequence can mean either ASCII- or Full-range Unicode,
681depending on various factors. See L</Locale, EBCDIC, Unicode and UTF-8>
682below.
8a118206
RGS
683
684=head4 [= =] and [. .]
685
b6538e4f
TC
686Perl recognizes the POSIX character classes C<[=class=]> and
687C<[.class.]>, but does not (yet?) support them. Any attempt to use
688either construct raises an exception.
8a118206
RGS
689
690=head4 Examples
691
692 /[[:digit:]]/ # Matches a character that is a digit.
693 /[01[:lower:]]/ # Matches a character that is either a
694 # lowercase letter, or '0' or '1'.
c1c4ae3a
KW
695 /[[:digit:][:^xdigit:]]/ # Matches a character that can be anything
696 # except the letters 'a' to 'f'. This is
697 # because the main character class is composed
698 # of two POSIX character classes that are ORed
699 # together, one that matches any digit, and
700 # the other that matches anything that isn't a
701 # hex digit. The result matches all
702 # characters except the letters 'a' to 'f' and
703 # 'A' to 'F'.
8a118206
RGS
704
705
ea449505 706=head2 Locale, EBCDIC, Unicode and UTF-8
8a118206 707
f7d1198f 708Some of the character classes have a somewhat different behaviour
b6538e4f
TC
709depending on the internal encoding of the source string, whether the regular
710expression is marked as having Unicode semantics, whatever locale is in
711effect, and whether the program is running on an EBCDIC platform.
f7d1198f
KW
712
713C<\w>, C<\d>, C<\s> and the POSIX character classes (and their
714negations, including C<\W>, C<\D>, C<\S>) have this behaviour. (Since
715the backslash sequences C<\b> and C<\B> are defined in terms of C<\w>
716and C<\W>, they also are affected.)
717
718Starting in Perl 5.14, if the regular expression is compiled with the
409a7f61 719C</a> modifier, the behavior doesn't differ regardless of any other
f7d1198f
KW
720factors. C<\d> matches the 10 digits 0-9; C<\D> any character but those
72110; C<\s>, exactly the five characters "[ \f\n\r\t]"; C<\w> only the 63
722characters "[A-Za-z0-9_]"; and the C<"[[:posix:]]"> classes only the
723appropriate ASCII characters, the same characters as are matched by the
724corresponding C<\p{}> property given in the "ASCII-range Unicode" column
725in the table above. (The behavior of all of their complements follows
726the same paradigm.)
727
728Otherwise, a regular expression is marked for Unicode semantics if it is
729encoded in utf8 (usually as a result of including a literal character
730whose code point is above 255), or if it contains a C<\N{U+...}> or
731C<\N{I<name>}> construct, or (starting in Perl 5.14) if it was compiled
732in the scope of a C<S<use feature "unicode_strings">> pragma and not in
409a7f61 733the scope of a C<S<use locale>> pragma, or has the C</u> regular
b6dac59a 734expression modifier.
17657a39 735
f7d1198f
KW
736Note that one can specify C<"use re '/l'"> for example, for any regular
737expression modifier, and this has precedence over either of the
738C<S<use feature "unicode_strings">> or C<S<use locale>> pragmas.
739
17657a39
KW
740The differences in behavior between locale and non-locale semantics
741can affect any character whose code point is 255 or less. The
742differences in behavior between Unicode and non-Unicode semantics
743affects only ASCII platforms, and only when matching against characters
744whose code points are between 128 and 255 inclusive. See
745L<perlunicode/The "Unicode Bug">.
8a118206 746
409a7f61 747For portability reasons, unless the C</a> modifier is specified,
f7d1198f
KW
748it may be better to not use C<\w>, C<\d>, C<\s> or the POSIX character
749classes and use the Unicode properties instead.
b6538e4f
TC
750
751That way you can control whether you want matching of characters in
752the ASCII character set alone, or whether to match Unicode characters.
753C<S<use feature "unicode_strings">> allows seamless Unicode behavior
754no matter the internal encodings, but won't allow restricting
755to ASCII characters only.
8a118206
RGS
756
757=head4 Examples
758
759 $str = "\xDF"; # $str is not in UTF-8 format.
760 $str =~ /^\w/; # No match, as $str isn't in UTF-8 format.
761 $str .= "\x{0e0b}"; # Now $str is in UTF-8 format.
762 $str =~ /^\w/; # Match! $str is now in UTF-8 format.
763 chop $str;
764 $str =~ /^\w/; # Still a match! $str remains in UTF-8 format.
765
766=cut