1PERLUNICODE(1) Perl Programmers Reference Guide PERLUNICODE(1)
2
6 perlunicode - Unicode support in Perl
7
9 If you haven't already, before reading this document, you should become
10 familiar with both perlunitut and perluniintro.
11 Unicode aims to UNI-fy the en-CODE-ings of all the world's character
13 sets into a single Standard. For quite a few of the various coding
14 standards that existed when Unicode was first created, converting from
15 each to Unicode essentially meant adding a constant to each code point
16 in the original standard, and converting back meant just subtracting
17 that same constant. For ASCII and ISO-8859-1, the constant is 0. For
18 ISO-8859-5, (Cyrillic) the constant is 864; for Hebrew (ISO-8859-8),
19 it's 1488; Thai (ISO-8859-11), 3424; and so forth. This made it easy
20 to do the conversions, and facilitated the adoption of Unicode.
21 And it worked; nowadays, those legacy standards are rarely used. Most
23 everyone uses Unicode.
24 Unicode is a comprehensive standard. It specifies many things outside
26 the scope of Perl, such as how to display sequences of characters. For
27 a full discussion of all aspects of Unicode, see
28 <https://www.unicode.org>.
29 Important Caveats
31 Even though some of this section may not be understandable to you on
32 first reading, we think it's important enough to highlight some of the
33 gotchas before delving further, so here goes:
34 Unicode support is an extensive requirement. While Perl does not
36 implement the Unicode standard or the accompanying technical reports
37 from cover to cover, Perl does support many Unicode features.
38 Also, the use of Unicode may present security issues that aren't
40 obvious, see "Security Implications of Unicode" below.
41 Safest if you "use feature 'unicode_strings'"
43 In order to preserve backward compatibility, Perl does not turn on
44 full internal Unicode support unless the pragma
45 "use feature 'unicode_strings'" is specified. (This is
46 automatically selected if you "use v5.12" or higher.) Failure to
47 do this can trigger unexpected surprises. See "The "Unicode Bug""
48 below.
49 This pragma doesn't affect I/O. Nor does it change the internal
51 representation of strings, only their interpretation. There are
52 still several places where Unicode isn't fully supported, such as
53 in filenames.
54 Input and Output Layers
56 Use the :encoding(...) layer to read from and write to filehandles
57 using the specified encoding. (See open.)
58 You must convert your non-ASCII, non-UTF-8 Perl scripts to be UTF-8.
60 The encoding module has been deprecated since perl 5.18 and the
61 perl internals it requires have been removed with perl 5.26.
62 "use utf8" still needed to enable UTF-8 in scripts
64 If your Perl script is itself encoded in UTF-8, the "use utf8"
65 pragma must be explicitly included to enable recognition of that
66 (in string or regular expression literals, or in identifier names).
67 This is the only time when an explicit "use utf8" is needed. (See
68 utf8).
69 If a Perl script begins with the bytes that form the UTF-8 encoding
71 of the Unicode BYTE ORDER MARK ("BOM", see "Unicode Encodings"),
72 those bytes are completely ignored.
73 UTF-16 scripts autodetected
75 If a Perl script begins with the Unicode "BOM" (UTF-16LE,
76 UTF16-BE), or if the script looks like non-"BOM"-marked UTF-16 of
77 either endianness, Perl will correctly read in the script as the
78 appropriate Unicode encoding.
79 Byte and Character Semantics
81 Before Unicode, most encodings used 8 bits (a single byte) to encode
82 each character. Thus a character was a byte, and a byte was a
83 character, and there could be only 256 or fewer possible characters.
84 "Byte Semantics" in the title of this section refers to this behavior.
85 There was no need to distinguish between "Byte" and "Character".
86 Then along comes Unicode which has room for over a million characters
88 (and Perl allows for even more). This means that a character may
89 require more than a single byte to represent it, and so the two terms
90 are no longer equivalent. What matter are the characters as whole
91 entities, and not usually the bytes that comprise them. That's what
92 the term "Character Semantics" in the title of this section refers to.
93 Perl had to change internally to decouple "bytes" from "characters".
95 It is important that you too change your ideas, if you haven't already,
96 so that "byte" and "character" no longer mean the same thing in your
97 mind.
98 The basic building block of Perl strings has always been a "character".
100 The changes basically come down to that the implementation no longer
101 thinks that a character is always just a single byte.
102 There are various things to note:
104 • String handling functions, for the most part, continue to operate
106 in terms of characters. length(), for example, returns the number
107 of characters in a string, just as before. But that number no
108 longer is necessarily the same as the number of bytes in the string
109 (there may be more bytes than characters). The other such
110 functions include chop(), chomp(), substr(), pos(), index(),
111 rindex(), sort(), sprintf(), and write().
112 The exceptions are:
114 • the bit-oriented "vec"
116 • the byte-oriented "pack"/"unpack" "C" format
120 However, the "W" specifier does operate on whole characters, as
122 does the "U" specifier.
123 • some operators that interact with the platform's operating
125 system
126 Operators dealing with filenames are examples.
128 • when the functions are called from within the scope of the
130 "use bytes" pragma
131 Likely, you should use this only for debugging anyway.
133 • Strings--including hash keys--and regular expression patterns may
135 contain characters that have ordinal values larger than 255.
136 If you use a Unicode editor to edit your program, Unicode
138 characters may occur directly within the literal strings in UTF-8
139 encoding, or UTF-16. (The former requires a "use utf8", the latter
140 may require a "BOM".)
141 "Creating Unicode" in perluniintro gives other ways to place non-
143 ASCII characters in your strings.
144 • The chr() and ord() functions work on whole characters.
146 • Regular expressions match whole characters. For example, "."
148 matches a whole character instead of only a single byte.
149 • The "tr///" operator translates whole characters. (Note that the
151 "tr///CU" functionality has been removed. For similar
152 functionality to that, see "pack('U0', ...)" and
153 "pack('C0', ...)").
154 • "scalar reverse()" reverses by character rather than by byte.
156 • The bit string operators, "& | ^ ~" and (starting in v5.22) "&. |.
158 ^. ~." can operate on bit strings encoded in UTF-8, but this can
159 give unexpected results if any of the strings contain code points
160 above 0xFF. Starting in v5.28, it is a fatal error to have such an
161 operand. Otherwise, the operation is performed on a non-UTF-8 copy
162 of the operand. If you're not sure about the encoding of a string,
163 downgrade it before using any of these operators; you can use
164 utf8::utf8_downgrade().
165 The bottom line is that Perl has always practiced "Character
167 Semantics", but with the advent of Unicode, that is now different than
168 "Byte Semantics".
169 ASCII Rules versus Unicode Rules
171 Before Unicode, when a character was a byte was a character, Perl knew
172 only about the 128 characters defined by ASCII, code points 0 through
173 127 (except for under "use locale"). That left the code points 128 to
174 255 as unassigned, and available for whatever use a program might want.
175 The only semantics they have is their ordinal numbers, and that they
176 are members of none of the non-negative character classes. None are
177 considered to match "\w" for example, but all match "\W".
178 Unicode, of course, assigns each of those code points a particular
180 meaning (along with ones above 255). To preserve backward
181 compatibility, Perl only uses the Unicode meanings when there is some
182 indication that Unicode is what is intended; otherwise the non-ASCII
183 code points remain treated as if they are unassigned.
184 Here are the ways that Perl knows that a string should be treated as
186 Unicode:
187 • Within the scope of "use utf8"
189 If the whole program is Unicode (signified by using 8-bit Unicode
191 Transformation Format), then all literal strings within it must be
192 Unicode.
193 • Within the scope of "use feature 'unicode_strings'"
195 This pragma was created so you can explicitly tell Perl that
197 operations executed within its scope are to use Unicode rules.
198 More operations are affected with newer perls. See "The "Unicode
199 Bug"".
200 • Within the scope of "use v5.12" or higher
202 This implicitly turns on "use feature 'unicode_strings'".
204 • Within the scope of "use locale 'not_characters'", or "use locale"
206 and the current locale is a UTF-8 locale.
207 The former is defined to imply Unicode handling; and the latter
209 indicates a Unicode locale, hence a Unicode interpretation of all
210 strings within it.
211 • When the string contains a Unicode-only code point
213 Perl has never accepted code points above 255 without them being
215 Unicode, so their use implies Unicode for the whole string.
216 • When the string contains a Unicode named code point "\N{...}"
218 The "\N{...}" construct explicitly refers to a Unicode code point,
220 even if it is one that is also in ASCII. Therefore the string
221 containing it must be Unicode.
222 • When the string has come from an external source marked as Unicode
224 The "-C" command line option can specify that certain inputs to the
226 program are Unicode, and the values of this can be read by your
227 Perl code, see "${^UNICODE}" in perlvar.
228 • When the string has been upgraded to UTF-8
230 The function utf8::utf8_upgrade() can be explicitly used to
232 permanently (unless a subsequent utf8::utf8_downgrade() is called)
233 cause a string to be treated as Unicode.
234 • There are additional methods for regular expression patterns
236 A pattern that is compiled with the "/u" or "/a" modifiers is
238 treated as Unicode (though there are some restrictions with "/a").
239 Under the "/d" and "/l" modifiers, there are several other
240 indications for Unicode; see "Character set modifiers" in perlre.
241 Note that all of the above are overridden within the scope of "use
243 bytes"; but you should be using this pragma only for debugging.
244 Note also that some interactions with the platform's operating system
246 never use Unicode rules.
247 When Unicode rules are in effect:
249 • Case translation operators use the Unicode case translation tables.
251 Note that uc(), or "\U" in interpolated strings, translates to
253 uppercase, while "ucfirst", or "\u" in interpolated strings,
254 translates to titlecase in languages that make the distinction
255 (which is equivalent to uppercase in languages without the
256 distinction).
257 There is a CPAN module, "Unicode::Casing", which allows you to
259 define your own mappings to be used in lc(), lcfirst(), uc(),
260 ucfirst(), and "fc" (or their double-quoted string inlined versions
261 such as "\U"). (Prior to Perl 5.16, this functionality was
262 partially provided in the Perl core, but suffered from a number of
263 insurmountable drawbacks, so the CPAN module was written instead.)
264 • Character classes in regular expressions match based on the
266 character properties specified in the Unicode properties database.
267 "\w" can be used to match a Japanese ideograph, for instance; and
269 "[[:digit:]]" a Bengali number.
270 • Named Unicode properties, scripts, and block ranges may be used
272 (like bracketed character classes) by using the "\p{}" "matches
273 property" construct and the "\P{}" negation, "doesn't match
274 property".
275 See "Unicode Character Properties" for more details.
277 You can define your own character properties and use them in the
279 regular expression with the "\p{}" or "\P{}" construct. See "User-
280 Defined Character Properties" for more details.
281 Extended Grapheme Clusters (Logical characters)
283 Consider a character, say "H". It could appear with various marks
284 around it, such as an acute accent, or a circumflex, or various hooks,
285 circles, arrows, etc., above, below, to one side or the other, etc.
286 There are many possibilities among the world's languages. The number
287 of combinations is astronomical, and if there were a character for each
288 combination, it would soon exhaust Unicode's more than a million
289 possible characters. So Unicode took a different approach: there is a
290 character for the base "H", and a character for each of the possible
291 marks, and these can be variously combined to get a final logical
292 character. So a logical character--what appears to be a single
293 character--can be a sequence of more than one individual characters.
294 The Unicode standard calls these "extended grapheme clusters" (which is
295 an improved version of the no-longer much used "grapheme cluster");
296 Perl furnishes the "\X" regular expression construct to match such
297 sequences in their entirety.
298 But Unicode's intent is to unify the existing character set standards
300 and practices, and several pre-existing standards have single
301 characters that mean the same thing as some of these combinations, like
302 ISO-8859-1, which has quite a few of them. For example, "LATIN CAPITAL
303 LETTER E WITH ACUTE" was already in this standard when Unicode came
304 along. Unicode therefore added it to its repertoire as that single
305 character. But this character is considered by Unicode to be
306 equivalent to the sequence consisting of the character "LATIN CAPITAL
307 LETTER E" followed by the character "COMBINING ACUTE ACCENT".
308 "LATIN CAPITAL LETTER E WITH ACUTE" is called a "pre-composed"
310 character, and its equivalence with the "E" and the "COMBINING ACCENT"
311 sequence is called canonical equivalence. All pre-composed characters
312 are said to have a decomposition (into the equivalent sequence), and
313 the decomposition type is also called canonical. A string may be
314 comprised as much as possible of precomposed characters, or it may be
315 comprised of entirely decomposed characters. Unicode calls these
316 respectively, "Normalization Form Composed" (NFC) and "Normalization
317 Form Decomposed". The "Unicode::Normalize" module contains functions
318 that convert between the two. A string may also have both composed
319 characters and decomposed characters; this module can be used to make
320 it all one or the other.
321 You may be presented with strings in any of these equivalent forms.
323 There is currently nothing in Perl 5 that ignores the differences. So
324 you'll have to specially handle it. The usual advice is to convert
325 your inputs to "NFD" before processing further.
326 For more detailed information, see <http://unicode.org/reports/tr15/>.
328 Unicode Character Properties
330 (The only time that Perl considers a sequence of individual code points
331 as a single logical character is in the "\X" construct, already
332 mentioned above. Therefore "character" in this discussion means a
333 single Unicode code point.)
334 Very nearly all Unicode character properties are accessible through
336 regular expressions by using the "\p{}" "matches property" construct
337 and the "\P{}" "doesn't match property" for its negation.
338 For instance, "\p{Uppercase}" matches any single character with the
340 Unicode "Uppercase" property, while "\p{L}" matches any character with
341 a "General_Category" of "L" (letter) property (see "General_Category"
342 below). Brackets are not required for single letter property names, so
343 "\p{L}" is equivalent to "\pL".
344 More formally, "\p{Uppercase}" matches any single character whose
346 Unicode "Uppercase" property value is "True", and "\P{Uppercase}"
347 matches any character whose "Uppercase" property value is "False", and
348 they could have been written as "\p{Uppercase=True}" and
349 "\p{Uppercase=False}", respectively.
350 This formality is needed when properties are not binary; that is, if
352 they can take on more values than just "True" and "False". For
353 example, the "Bidi_Class" property (see "Bidirectional Character Types"
354 below), can take on several different values, such as "Left", "Right",
355 "Whitespace", and others. To match these, one needs to specify both
356 the property name ("Bidi_Class"), AND the value being matched against
357 ("Left", "Right", etc.). This is done, as in the examples above, by
358 having the two components separated by an equal sign (or
359 interchangeably, a colon), like "\p{Bidi_Class: Left}".
360 All Unicode-defined character properties may be written in these
362 compound forms of "\p{property=value}" or "\p{property:value}", but
363 Perl provides some additional properties that are written only in the
364 single form, as well as single-form short-cuts for all binary
365 properties and certain others described below, in which you may omit
366 the property name and the equals or colon separator.
367 Most Unicode character properties have at least two synonyms (or
369 aliases if you prefer): a short one that is easier to type and a longer
370 one that is more descriptive and hence easier to understand. Thus the
371 "L" and "Letter" properties above are equivalent and can be used
372 interchangeably. Likewise, "Upper" is a synonym for "Uppercase", and
373 we could have written "\p{Uppercase}" equivalently as "\p{Upper}".
374 Also, there are typically various synonyms for the values the property
375 can be. For binary properties, "True" has 3 synonyms: "T", "Yes", and
376 "Y"; and "False" has correspondingly "F", "No", and "N". But be
377 careful. A short form of a value for one property may not mean the
378 same thing as the short form spelled the same for another. Thus, for
379 the "General_Category" property, "L" means "Letter", but for the
380 "Bidi_Class" property, "L" means "Left". A complete list of properties
381 and synonyms is in perluniprops.
382 Upper/lower case differences in property names and values are
384 irrelevant; thus "\p{Upper}" means the same thing as "\p{upper}" or
385 even "\p{UpPeR}". Similarly, you can add or subtract underscores
386 anywhere in the middle of a word, so that these are also equivalent to
387 "\p{U_p_p_e_r}". And white space is generally irrelevant adjacent to
388 non-word characters, such as the braces and the equals or colon
389 separators, so "\p{ Upper }" and "\p{ Upper_case : Y }" are
390 equivalent to these as well. In fact, white space and even hyphens can
391 usually be added or deleted anywhere. So even "\p{ Up-per case = Yes}"
392 is equivalent. All this is called "loose-matching" by Unicode. The
393 "name" property has some restrictions on this due to a few outlier
394 names. Full details are given in
395 <https://www.unicode.org/reports/tr44/tr44-24.html#UAX44-LM2>.
396 The few places where stricter matching is used is in the middle of
398 numbers, the "name" property, and in the Perl extension properties that
399 begin or end with an underscore. Stricter matching cares about white
400 space (except adjacent to non-word characters), hyphens, and non-
401 interior underscores.
402 You can also use negation in both "\p{}" and "\P{}" by introducing a
404 caret ("^") between the first brace and the property name: "\p{^Tamil}"
405 is equal to "\P{Tamil}".
406 Almost all properties are immune to case-insensitive matching. That
408 is, adding a "/i" regular expression modifier does not change what they
409 match. There are two sets that are affected. The first set is
410 "Uppercase_Letter", "Lowercase_Letter", and "Titlecase_Letter", all of
411 which match "Cased_Letter" under "/i" matching. And the second set is
412 "Uppercase", "Lowercase", and "Titlecase", all of which match "Cased"
413 under "/i" matching. This set also includes its subsets "PosixUpper"
414 and "PosixLower" both of which under "/i" match "PosixAlpha". (The
415 difference between these sets is that some things, such as Roman
416 numerals, come in both upper and lower case so they are "Cased", but
417 aren't considered letters, so they aren't "Cased_Letter"'s.)
418 See "Beyond Unicode code points" for special considerations when
420 matching Unicode properties against non-Unicode code points.
421 General_Category
423 Every Unicode character is assigned a general category, which is the
425 "most usual categorization of a character" (from
426 <https://www.unicode.org/reports/tr44>).
427 The compound way of writing these is like "\p{General_Category=Number}"
429 (short: "\p{gc:n}"). But Perl furnishes shortcuts in which everything
430 up through the equal or colon separator is omitted. So you can instead
431 just write "\pN".
432 Here are the short and long forms of the values the "General Category"
434 property can have:
435 Short Long
437 L Letter
439 LC, L& Cased_Letter (that is: [\p{Ll}\p{Lu}\p{Lt}])
440 Lu Uppercase_Letter
441 Ll Lowercase_Letter
442 Lt Titlecase_Letter
443 Lm Modifier_Letter
444 Lo Other_Letter
445 M Mark
447 Mn Nonspacing_Mark
448 Mc Spacing_Mark
449 Me Enclosing_Mark
450 N Number
452 Nd Decimal_Number (also Digit)
453 Nl Letter_Number
454 No Other_Number
455 P Punctuation (also Punct)
457 Pc Connector_Punctuation
458 Pd Dash_Punctuation
459 Ps Open_Punctuation
460 Pe Close_Punctuation
461 Pi Initial_Punctuation
462 (may behave like Ps or Pe depending on usage)
463 Pf Final_Punctuation
464 (may behave like Ps or Pe depending on usage)
465 Po Other_Punctuation
466 S Symbol
468 Sm Math_Symbol
469 Sc Currency_Symbol
470 Sk Modifier_Symbol
471 So Other_Symbol
472 Z Separator
474 Zs Space_Separator
475 Zl Line_Separator
476 Zp Paragraph_Separator
477 C Other
479 Cc Control (also Cntrl)
480 Cf Format
481 Cs Surrogate
482 Co Private_Use
483 Cn Unassigned
484 Single-letter properties match all characters in any of the two-letter
486 sub-properties starting with the same letter. "LC" and "L&" are
487 special: both are aliases for the set consisting of everything matched
488 by "Ll", "Lu", and "Lt".
489 Bidirectional Character Types
491 Because scripts differ in their directionality (Hebrew and Arabic are
493 written right to left, for example) Unicode supplies a "Bidi_Class"
494 property. Some of the values this property can have are:
495 Value Meaning
497 L Left-to-Right
499 LRE Left-to-Right Embedding
500 LRO Left-to-Right Override
501 R Right-to-Left
502 AL Arabic Letter
503 RLE Right-to-Left Embedding
504 RLO Right-to-Left Override
505 PDF Pop Directional Format
506 EN European Number
507 ES European Separator
508 ET European Terminator
509 AN Arabic Number
510 CS Common Separator
511 NSM Non-Spacing Mark
512 BN Boundary Neutral
513 B Paragraph Separator
514 S Segment Separator
515 WS Whitespace
516 ON Other Neutrals
517 This property is always written in the compound form. For example,
519 "\p{Bidi_Class:R}" matches characters that are normally written right
520 to left. Unlike the "General_Category" property, this property can
521 have more values added in a future Unicode release. Those listed above
522 comprised the complete set for many Unicode releases, but others were
523 added in Unicode 6.3; you can always find what the current ones are in
524 perluniprops. And <https://www.unicode.org/reports/tr9/> describes how
525 to use them.
526 Scripts
528 The world's languages are written in many different scripts. This
530 sentence (unless you're reading it in translation) is written in Latin,
531 while Russian is written in Cyrillic, and Greek is written in, well,
532 Greek; Japanese mainly in Hiragana or Katakana. There are many more.
533 The Unicode "Script" and "Script_Extensions" properties give what
535 script a given character is in. The "Script_Extensions" property is an
536 improved version of "Script", as demonstrated below. Either property
537 can be specified with the compound form like "\p{Script=Hebrew}"
538 (short: "\p{sc=hebr}"), or "\p{Script_Extensions=Javanese}" (short:
539 "\p{scx=java}"). In addition, Perl furnishes shortcuts for all
540 "Script_Extensions" property names. You can omit everything up through
541 the equals (or colon), and simply write "\p{Latin}" or "\P{Cyrillic}".
542 (This is not true for "Script", which is required to be written in the
543 compound form. Prior to Perl v5.26, the single form returned the plain
544 old "Script" version, but was changed because "Script_Extensions" gives
545 better results.)
546 The difference between these two properties involves characters that
548 are used in multiple scripts. For example the digits '0' through '9'
549 are used in many parts of the world. These are placed in a script
550 named "Common". Other characters are used in just a few scripts. For
551 example, the "KATAKANA-HIRAGANA DOUBLE HYPHEN" is used in both Japanese
552 scripts, Katakana and Hiragana, but nowhere else. The "Script"
553 property places all characters that are used in multiple scripts in the
554 "Common" script, while the "Script_Extensions" property places those
555 that are used in only a few scripts into each of those scripts; while
556 still using "Common" for those used in many scripts. Thus both these
557 match:
558 "0" =~ /\p{sc=Common}/ # Matches
560 "0" =~ /\p{scx=Common}/ # Matches
561 and only the first of these match:
563 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{sc=Common} # Matches
565 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{scx=Common} # No match
566 And only the last two of these match:
568 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{sc=Hiragana} # No match
570 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{sc=Katakana} # No match
571 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{scx=Hiragana} # Matches
572 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{scx=Katakana} # Matches
573 "Script_Extensions" is thus an improved "Script", in which there are
575 fewer characters in the "Common" script, and correspondingly more in
576 other scripts. It is new in Unicode version 6.0, and its data are
577 likely to change significantly in later releases, as things get sorted
578 out. New code should probably be using "Script_Extensions" and not
579 plain "Script". If you compile perl with a Unicode release that
580 doesn't have "Script_Extensions", the single form Perl extensions will
581 instead refer to the plain "Script" property. If you compile with a
582 version of Unicode that doesn't have the "Script" property, these
583 extensions will not be defined at all.
584 (Actually, besides "Common", the "Inherited" script, contains
586 characters that are used in multiple scripts. These are modifier
587 characters which inherit the script value of the controlling character.
588 Some of these are used in many scripts, and so go into "Inherited" in
589 both "Script" and "Script_Extensions". Others are used in just a few
590 scripts, so are in "Inherited" in "Script", but not in
591 "Script_Extensions".)
592 It is worth stressing that there are several different sets of digits
594 in Unicode that are equivalent to 0-9 and are matchable by "\d" in a
595 regular expression. If they are used in a single language only, they
596 are in that language's "Script" and "Script_Extensions". If they are
597 used in more than one script, they will be in "sc=Common", but only if
598 they are used in many scripts should they be in "scx=Common".
599 The explanation above has omitted some detail; refer to UAX#24 "Unicode
601 Script Property": <https://www.unicode.org/reports/tr24>.
602 A complete list of scripts and their shortcuts is in perluniprops.
604 Use of the "Is" Prefix
606 For backward compatibility (with ancient Perl 5.6), all properties
608 writable without using the compound form mentioned so far may have "Is"
609 or "Is_" prepended to their name, so "\P{Is_Lu}", for example, is equal
610 to "\P{Lu}", and "\p{IsScript:Arabic}" is equal to "\p{Arabic}".
611 Blocks
613 In addition to scripts, Unicode also defines blocks of characters. The
615 difference between scripts and blocks is that the concept of scripts is
616 closer to natural languages, while the concept of blocks is more of an
617 artificial grouping based on groups of Unicode characters with
618 consecutive ordinal values. For example, the "Basic Latin" block is all
619 the characters whose ordinals are between 0 and 127, inclusive; in
620 other words, the ASCII characters. The "Latin" script contains some
621 letters from this as well as several other blocks, like "Latin-1
622 Supplement", "Latin Extended-A", etc., but it does not contain all the
623 characters from those blocks. It does not, for example, contain the
624 digits 0-9, because those digits are shared across many scripts, and
625 hence are in the "Common" script.
626 For more about scripts versus blocks, see UAX#24 "Unicode Script
628 Property": <https://www.unicode.org/reports/tr24>
629 The "Script_Extensions" or "Script" properties are likely to be the
631 ones you want to use when processing natural language; the "Block"
632 property may occasionally be useful in working with the nuts and bolts
633 of Unicode.
634 Block names are matched in the compound form, like "\p{Block: Arrows}"
636 or "\p{Blk=Hebrew}". Unlike most other properties, only a few block
637 names have a Unicode-defined short name.
638 Perl also defines single form synonyms for the block property in cases
640 where these do not conflict with something else. But don't use any of
641 these, because they are unstable. Since these are Perl extensions,
642 they are subordinate to official Unicode property names; Unicode
643 doesn't know nor care about Perl's extensions. It may happen that a
644 name that currently means the Perl extension will later be changed
645 without warning to mean a different Unicode property in a future
646 version of the perl interpreter that uses a later Unicode release, and
647 your code would no longer work. The extensions are mentioned here for
648 completeness: Take the block name and prefix it with one of: "In" (for
649 example "\p{Blk=Arrows}" can currently be written as "\p{In_Arrows}");
650 or sometimes "Is" (like "\p{Is_Arrows}"); or sometimes no prefix at all
651 ("\p{Arrows}"). As of this writing (Unicode 9.0) there are no
652 conflicts with using the "In_" prefix, but there are plenty with the
653 other two forms. For example, "\p{Is_Hebrew}" and "\p{Hebrew}" mean
654 "\p{Script_Extensions=Hebrew}" which is NOT the same thing as
655 "\p{Blk=Hebrew}". Our advice used to be to use the "In_" prefix as a
656 single form way of specifying a block. But Unicode 8.0 added
657 properties whose names begin with "In", and it's now clear that it's
658 only luck that's so far prevented a conflict. Using "In" is only
659 marginally less typing than "Blk:", and the latter's meaning is clearer
660 anyway, and guaranteed to never conflict. So don't take chances. Use
661 "\p{Blk=foo}" for new code. And be sure that block is what you really
662 really want to do. In most cases scripts are what you want instead.
663 A complete list of blocks is in perluniprops.
665 Other Properties
667 There are many more properties than the very basic ones described here.
669 A complete list is in perluniprops.
670 Unicode defines all its properties in the compound form, so all single-
672 form properties are Perl extensions. Most of these are just synonyms
673 for the Unicode ones, but some are genuine extensions, including
674 several that are in the compound form. And quite a few of these are
675 actually recommended by Unicode (in
676 <https://www.unicode.org/reports/tr18>).
677 This section gives some details on all extensions that aren't just
679 synonyms for compound-form Unicode properties (for those properties,
680 you'll have to refer to the Unicode Standard
681 <https://www.unicode.org/reports/tr44>.
682 "\p{All}"
684 This matches every possible code point. It is equivalent to
685 "qr/./s". Unlike all the other non-user-defined "\p{}" property
686 matches, no warning is ever generated if this is property is
687 matched against a non-Unicode code point (see "Beyond Unicode code
688 points" below).
689 "\p{Alnum}"
691 This matches any "\p{Alphabetic}" or "\p{Decimal_Number}"
692 character.
693 "\p{Any}"
695 This matches any of the 1_114_112 Unicode code points. It is a
696 synonym for "\p{Unicode}".
697 "\p{ASCII}"
699 This matches any of the 128 characters in the US-ASCII character
700 set, which is a subset of Unicode.
701 "\p{Assigned}"
703 This matches any assigned code point; that is, any code point whose
704 general category is not "Unassigned" (or equivalently, not "Cn").
705 "\p{Blank}"
707 This is the same as "\h" and "\p{HorizSpace}": A character that
708 changes the spacing horizontally.
709 "\p{Decomposition_Type: Non_Canonical}" (Short: "\p{Dt=NonCanon}")
711 Matches a character that has any of the non-canonical decomposition
712 types. Canonical decompositions are introduced in the "Extended
713 Grapheme Clusters (Logical characters)" section above. However,
714 many more characters have a different type of decomposition,
715 generically called "compatible" decompositions, or "non-canonical".
716 The sequences that form these decompositions are not considered
717 canonically equivalent to the pre-composed character. An example
718 is the "SUPERSCRIPT ONE". It is somewhat like a regular digit 1,
719 but not exactly; its decomposition into the digit 1 is called a
720 "compatible" decomposition, specifically a "super" (for
721 "superscript") decomposition. There are several such compatibility
722 decompositions (see <https://www.unicode.org/reports/tr44>).
723 "\p{Dt: Non_Canon}" is a Perl extension that uses just one name to
724 refer to the union of all of them.
725 Most Unicode characters don't have a decomposition, so their
727 decomposition type is "None". Hence, "Non_Canonical" is equivalent
728 to
729 qr/(?[ \P{DT=Canonical} - \p{DT=None} ])/
731 (Note that one of the non-canonical decompositions is named
733 "compat", which could perhaps have been better named
734 "miscellaneous". It includes just the things that Unicode couldn't
735 figure out a better generic name for.)
736 "\p{Graph}"
738 Matches any character that is graphic. Theoretically, this means a
739 character that on a printer would cause ink to be used.
740 "\p{HorizSpace}"
742 This is the same as "\h" and "\p{Blank}": a character that changes
743 the spacing horizontally.
744 "\p{In=*}"
746 This is a synonym for "\p{Present_In=*}"
747 "\p{PerlSpace}"
749 This is the same as "\s", restricted to ASCII, namely "[ \f\n\r\t]"
750 and starting in Perl v5.18, a vertical tab.
751 Mnemonic: Perl's (original) space
753 "\p{PerlWord}"
755 This is the same as "\w", restricted to ASCII, namely
756 "[A-Za-z0-9_]"
757 Mnemonic: Perl's (original) word.
759 "\p{Posix...}"
761 There are several of these, which are equivalents, using the "\p{}"
762 notation, for Posix classes and are described in "POSIX Character
763 Classes" in perlrecharclass.
764 "\p{Present_In: *}" (Short: "\p{In=*}")
766 This property is used when you need to know in what Unicode
767 version(s) a character is.
768 The "*" above stands for some Unicode version number, such as 1.1
770 or 12.0; or the "*" can also be "Unassigned". This property will
771 match the code points whose final disposition has been settled as
772 of the Unicode release given by the version number; "\p{Present_In:
773 Unassigned}" will match those code points whose meaning has yet to
774 be assigned.
775 For example, "U+0041" "LATIN CAPITAL LETTER A" was present in the
777 very first Unicode release available, which is 1.1, so this
778 property is true for all valid "*" versions. On the other hand,
779 "U+1EFF" was not assigned until version 5.1 when it became "LATIN
780 SMALL LETTER Y WITH LOOP", so the only "*" that would match it are
781 5.1, 5.2, and later.
782 Unicode furnishes the "Age" property from which this is derived.
784 The problem with Age is that a strict interpretation of it (which
785 Perl takes) has it matching the precise release a code point's
786 meaning is introduced in. Thus "U+0041" would match only 1.1; and
787 "U+1EFF" only 5.1. This is not usually what you want.
788 Some non-Perl implementations of the Age property may change its
790 meaning to be the same as the Perl "Present_In" property; just be
791 aware of that.
792 Another confusion with both these properties is that the definition
794 is not that the code point has been assigned, but that the meaning
795 of the code point has been determined. This is because 66 code
796 points will always be unassigned, and so the "Age" for them is the
797 Unicode version in which the decision to make them so was made.
798 For example, "U+FDD0" is to be permanently unassigned to a
799 character, and the decision to do that was made in version 3.1, so
800 "\p{Age=3.1}" matches this character, as also does "\p{Present_In:
801 3.1}" and up.
802 "\p{Print}"
804 This matches any character that is graphical or blank, except
805 controls.
806 "\p{SpacePerl}"
808 This is the same as "\s", including beyond ASCII.
809 Mnemonic: Space, as modified by Perl. (It doesn't include the
811 vertical tab until v5.18, which both the Posix standard and Unicode
812 consider white space.)
813 "\p{Title}" and "\p{Titlecase}"
815 Under case-sensitive matching, these both match the same code
816 points as "\p{General Category=Titlecase_Letter}" ("\p{gc=lt}").
817 The difference is that under "/i" caseless matching, these match
818 the same as "\p{Cased}", whereas "\p{gc=lt}" matches
819 "\p{Cased_Letter").
820 "\p{Unicode}"
822 This matches any of the 1_114_112 Unicode code points. "\p{Any}".
823 "\p{VertSpace}"
825 This is the same as "\v": A character that changes the spacing
826 vertically.
827 "\p{Word}"
829 This is the same as "\w", including over 100_000 characters beyond
830 ASCII.
831 "\p{XPosix...}"
833 There are several of these, which are the standard Posix classes
834 extended to the full Unicode range. They are described in "POSIX
835 Character Classes" in perlrecharclass.
836 Comparison of "\N{...}" and "\p{name=...}"
838 Starting in Perl 5.32, you can specify a character by its name in
839 regular expression patterns using "\p{name=...}". This is in addition
840 to the longstanding method of using "\N{...}". The following
841 summarizes the differences between these two:
842 \N{...} \p{Name=...}
844 can interpolate only with eval yes [1]
845 custom names yes no [2]
846 name aliases yes yes [3]
847 named sequences yes yes [4]
848 name value parsing exact Unicode loose [5]
849 [1] The ability to interpolate means you can do something like
851 qr/\p{na=latin capital letter $which}/
853 and specify $which elsewhere.
855 [2] You can create your own names for characters, and override official
857 ones when using "\N{...}". See "CUSTOM ALIASES" in charnames.
858 [3] Some characters have multiple names (synonyms).
860 [4] Some particular sequences of characters are given a single name, in
862 addition to their individual ones.
863 [5] Exact name value matching means you have to specify case, hyphens,
865 underscores, and spaces precisely in the name you want. Loose
866 matching follows the Unicode rules
867 <https://www.unicode.org/reports/tr44/tr44-24.html#UAX44-LM2>,
868 where these are mostly irrelevant. Except for a few outlier
869 character names, these are the same rules as are already used for
870 any other "\p{...}" property.
871 Wildcards in Property Values
873 Starting in Perl 5.30, it is possible to do something like this:
874 qr!\p{numeric_value=/\A[0-5]\z/}!
876 or, by abbreviating and adding "/x",
878 qr! \p{nv= /(?x) \A [0-5] \z / }!
880 This matches all code points whose numeric value is one of 0, 1, 2, 3,
882 4, or 5. This particular example could instead have been written as
883 qr! \A [ \p{nv=0}\p{nv=1}\p{nv=2}\p{nv=3}\p{nv=4}\p{nv=5} ] \z !xx
885 in earlier perls, so in this case this feature just makes things easier
887 and shorter to write. If we hadn't included the "\A" and "\z", these
888 would have matched things like "1/2" because that contains a 1 (as well
889 as a 2). As written, it matches things like subscripts that have these
890 numeric values. If we only wanted the decimal digits with those
891 numeric values, we could say,
892 qr! (?[ \d & \p{nv=/[0-5]/ ]) }!x
894 The "\d" gets rid of needing to anchor the pattern, since it forces the
896 result to only match "[0-9]", and the "[0-5]" further restricts it.
897 The text in the above examples enclosed between the "/" characters can
899 be just about any regular expression. It is independent of the main
900 pattern, so doesn't share any capturing groups, etc. The delimiters
901 for it must be ASCII punctuation, but it may NOT be delimited by "{",
902 nor "}" nor contain a literal "}", as that delimits the end of the
903 enclosing "\p{}". Like any pattern, certain other delimiters are
904 terminated by their mirror images. These are "(", ""["", and "<". If
905 the delimiter is any of "-", "_", "+", or "\", or is the same delimiter
906 as is used for the enclosing pattern, it must be preceded by a
907 backslash escape, both fore and aft.
908 Beware of using "$" to indicate to match the end of the string. It can
910 too easily be interpreted as being a punctuation variable, like $/.
911 No modifiers may follow the final delimiter. Instead, use
913 "(?adlupimnsx-imnsx)" in perlre and/or "(?adluimnsx-imnsx:pattern)" in
914 perlre to specify modifiers. However, certain modifiers are illegal in
915 your wildcard subpattern. The only character set modifier specifiable
916 is "/aa"; any other character set, and "-m", and "p", and "s" are all
917 illegal. Specifying modifiers like "qr/.../gc" that aren't legal in
918 the "(?...)" notation normally raise a warning, but with wildcard
919 subpatterns, their use is an error. The "m" modifier is ineffective;
920 everything that matches will be a single line.
921 By default, your pattern is matched case-insensitively, as if "/i" had
923 been specified. You can change this by saying "(?-i)" in your pattern.
924 There are also certain operations that are illegal. You can't nest
926 "\p{...}" and "\P{...}" calls within a wildcard subpattern, and "\G"
927 doesn't make sense, so is also prohibited.
928 And the "*" quantifier (or its equivalent "(0,}") is illegal.
930 This feature is not available when the left-hand side is prefixed by
932 "Is_", nor for any form that is marked as "Discouraged" in
933 "Discouraged" in perluniprops.
934 This experimental feature has been added to begin to implement
936 <https://www.unicode.org/reports/tr18/#Wildcard_Properties>. Using it
937 will raise a (default-on) warning in the
938 "experimental::uniprop_wildcards" category. We reserve the right to
939 change its operation as we gain experience.
940 Your subpattern can be just about anything, but for it to have some
942 utility, it should match when called with either or both of a) the full
943 name of the property value with underscores (and/or spaces in the Block
944 property) and some things uppercase; or b) the property value in all
945 lowercase with spaces and underscores squeezed out. For example,
946 qr!\p{Blk=/Old I.*/}!
948 qr!\p{Blk=/oldi.*/}!
949 would match the same things.
951 Another example that shows that within "\p{...}", "/x" isn't needed to
953 have spaces:
954 qr!\p{scx= /Hebrew|Greek/ }!
956 To be safe, we should have anchored the above example, to prevent
958 matches for something like "Hebrew_Braille", but there aren't any
959 script names like that, so far. A warning is issued if none of the
960 legal values for a property are matched by your pattern. It's likely
961 that a future release will raise a warning if your pattern ends up
962 causing every possible code point to match.
963 Starting in 5.32, the Name, Name Aliases, and Named Sequences
965 properties are allowed to be matched. They are considered to be a
966 single combination property, just as has long been the case for "\N{}".
967 Loose matching doesn't work in exactly the same way for these as it
968 does for the values of other properties. The rules are given in
969 <https://www.unicode.org/reports/tr44/tr44-24.html#UAX44-LM2>. As a
970 result, Perl doesn't try loose matching for you, like it does in other
971 properties. All letters in names are uppercase, but you can add "(?i)"
972 to your subpattern to ignore case. If you're uncertain where a blank
973 is, you can use " ?" in your subpattern. No character name contains an
974 underscore, so don't bother trying to match one. The use of hyphens is
975 particularly problematic; refer to the above link. But note that, as
976 of Unicode 13.0, the only script in modern usage which has weirdnesses
977 with these is Tibetan; also the two Korean characters U+116C HANGUL
978 JUNGSEONG OE and U+1180 HANGUL JUNGSEONG O-E. Unicode makes no
979 promises to not add hyphen-problematic names in the future.
980 Using wildcards on these is resource intensive, given the hundreds of
982 thousands of legal names that must be checked against.
983 An example of using Name property wildcards is
985 qr!\p{name=/(SMILING|GRINNING) FACE/}!
987 Another is
989 qr/(?[ \p{name=\/CJK\/} - \p{ideographic} ])/
991 which is the 200-ish (as of Unicode 13.0) CJK characters that aren't
993 ideographs.
994 There are certain properties that wildcard subpatterns don't currently
996 work with. These are:
997 Bidi Mirroring Glyph
999 Bidi Paired Bracket
1000 Case Folding
1001 Decomposition Mapping
1002 Equivalent Unified Ideograph
1003 Lowercase Mapping
1004 NFKC Case Fold
1005 Titlecase Mapping
1006 Uppercase Mapping
1007 Nor is the "@unicode_property@" form implemented.
1009 Here's a complete example of matching IPV4 internet protocol addresses
1011 in any (single) script
1012 no warnings 'experimental::uniprop_wildcards';
1014 # Can match a substring, so this intermediate regex needs to have
1016 # context or anchoring in its final use. Using nt=de yields decimal
1017 # digits. When specifying a subset of these, we must include \d to
1018 # prevent things like U+00B2 SUPERSCRIPT TWO from matching
1019 my $zero_through_255 =
1020 qr/ \b (*sr: # All from same sript
1021 (?[ \p{nv=0} & \d ])* # Optional leading zeros
1022 ( # Then one of:
1023 \d{1,2} # 0 - 99
1024 | (?[ \p{nv=1} & \d ]) \d{2} # 100 - 199
1025 | (?[ \p{nv=2} & \d ])
1026 ( (?[ \p{nv=:[0-4]:} & \d ]) \d # 200 - 249
1027 | (?[ \p{nv=5} & \d ])
1028 (?[ \p{nv=:[0-5]:} & \d ]) # 250 - 255
1029 )
1030 )
1031 )
1032 \b
1033 /x;
1034 my $ipv4 = qr/ \A (*sr: $zero_through_255
1036 (?: [.] $zero_through_255 ) {3}
1037 )
1038 \z
1039 /x;
1040 User-Defined Character Properties
1042 You can define your own binary character properties by defining
1043 subroutines whose names begin with "In" or "Is". (The regex sets
1044 feature "(?[ ])" in perlre provides an alternative which allows more
1045 complex definitions.) The subroutines can be defined in any package.
1046 They override any Unicode properties expressed as the same names. The
1047 user-defined properties can be used in the regular expression "\p{}"
1048 and "\P{}" constructs; if you are using a user-defined property from a
1049 package other than the one you are in, you must specify its package in
1050 the "\p{}" or "\P{}" construct.
1051 # assuming property IsForeign defined in Lang::
1053 package main; # property package name required
1054 if ($txt =~ /\p{Lang::IsForeign}+/) { ... }
1055 package Lang; # property package name not required
1057 if ($txt =~ /\p{IsForeign}+/) { ... }
1058 The subroutines are passed a single parameter, which is 0 if case-
1060 sensitive matching is in effect and non-zero if caseless matching is in
1061 effect. The subroutine may return different values depending on the
1062 value of the flag. But the subroutine is never called more than once
1063 for each flag value (zero vs non-zero). The return value is saved and
1064 used instead of calling the sub ever again. If the sub is defined at
1065 the time the pattern is compiled, it will be called then; if not, it
1066 will be called the first time its value (for that flag) is needed
1067 during execution.
1068 Note that if the regular expression is tainted, then Perl will die
1070 rather than calling the subroutine when the name of the subroutine is
1071 determined by the tainted data.
1072 The subroutines must return a specially-formatted string, with one or
1074 more newline-separated lines. Each line must be one of the following:
1075 • A single hexadecimal number denoting a code point to include.
1077 • Two hexadecimal numbers separated by horizontal whitespace (space
1079 or tabular characters) denoting a range of code points to include.
1080 The second number must not be smaller than the first.
1081 • Something to include, prefixed by "+": a built-in character
1083 property (prefixed by "utf8::") or a fully qualified (including
1084 package name) user-defined character property, to represent all the
1085 characters in that property; two hexadecimal code points for a
1086 range; or a single hexadecimal code point.
1087 • Something to exclude, prefixed by "-": an existing character
1089 property (prefixed by "utf8::") or a fully qualified (including
1090 package name) user-defined character property, to represent all the
1091 characters in that property; two hexadecimal code points for a
1092 range; or a single hexadecimal code point.
1093 • Something to negate, prefixed "!": an existing character property
1095 (prefixed by "utf8::") or a fully qualified (including package
1096 name) user-defined character property, to represent all the
1097 characters in that property; two hexadecimal code points for a
1098 range; or a single hexadecimal code point.
1099 • Something to intersect with, prefixed by "&": an existing character
1101 property (prefixed by "utf8::") or a fully qualified (including
1102 package name) user-defined character property, for all the
1103 characters except the characters in the property; two hexadecimal
1104 code points for a range; or a single hexadecimal code point.
1105 For example, to define a property that covers both the Japanese
1107 syllabaries (hiragana and katakana), you can define
1108 sub InKana {
1110 return <<END;
1111 3040\t309F
1112 30A0\t30FF
1113 END
1114 }
1115 Imagine that the here-doc end marker is at the beginning of the line.
1117 Now you can use "\p{InKana}" and "\P{InKana}".
1118 You could also have used the existing block property names:
1120 sub InKana {
1122 return <<'END';
1123 +utf8::InHiragana
1124 +utf8::InKatakana
1125 END
1126 }
1127 Suppose you wanted to match only the allocated characters, not the raw
1129 block ranges: in other words, you want to remove the unassigned
1130 characters:
1131 sub InKana {
1133 return <<'END';
1134 +utf8::InHiragana
1135 +utf8::InKatakana
1136 -utf8::IsCn
1137 END
1138 }
1139 The negation is useful for defining (surprise!) negated classes.
1141 sub InNotKana {
1143 return <<'END';
1144 !utf8::InHiragana
1145 -utf8::InKatakana
1146 +utf8::IsCn
1147 END
1148 }
1149 This will match all non-Unicode code points, since every one of them is
1151 not in Kana. You can use intersection to exclude these, if desired, as
1152 this modified example shows:
1153 sub InNotKana {
1155 return <<'END';
1156 !utf8::InHiragana
1157 -utf8::InKatakana
1158 +utf8::IsCn
1159 &utf8::Any
1160 END
1161 }
1162 &utf8::Any must be the last line in the definition.
1164 Intersection is used generally for getting the common characters
1166 matched by two (or more) classes. It's important to remember not to
1167 use "&" for the first set; that would be intersecting with nothing,
1168 resulting in an empty set. (Similarly using "-" for the first set does
1169 nothing).
1170 Unlike non-user-defined "\p{}" property matches, no warning is ever
1172 generated if these properties are matched against a non-Unicode code
1173 point (see "Beyond Unicode code points" below).
1174 User-Defined Case Mappings (for serious hackers only)
1176 This feature has been removed as of Perl 5.16. The CPAN module
1177 "Unicode::Casing" provides better functionality without the drawbacks
1178 that this feature had. If you are using a Perl earlier than 5.16, this
1179 feature was most fully documented in the 5.14 version of this pod:
1180 <http://perldoc.perl.org/5.14.0/perlunicode.html#User-Defined-Case-Mappings-%28for-serious-hackers-only%29>
1181 Character Encodings for Input and Output
1183 See Encode.
1184 Unicode Regular Expression Support Level
1186 The following list of Unicode supported features for regular
1187 expressions describes all features currently directly supported by core
1188 Perl. The references to "Level N" and the section numbers refer to
1189 UTS#18 "Unicode Regular Expressions"
1190 <https://www.unicode.org/reports/tr18>, version 18, October 2016.
1191 Level 1 - Basic Unicode Support
1193 RL1.1 Hex Notation - Done [1]
1195 RL1.2 Properties - Done [2]
1196 RL1.2a Compatibility Properties - Done [3]
1197 RL1.3 Subtraction and Intersection - Done [4]
1198 RL1.4 Simple Word Boundaries - Done [5]
1199 RL1.5 Simple Loose Matches - Done [6]
1200 RL1.6 Line Boundaries - Partial [7]
1201 RL1.7 Supplementary Code Points - Done [8]
1202 [1] "\N{U+...}" and "\x{...}"
1204 [2] "\p{...}" "\P{...}". This requirement is for a minimal list of
1205 properties. Perl supports these. See R2.7 for other properties.
1206 [3] Perl has "\d" "\D" "\s" "\S" "\w" "\W" "\X" "[:prop:]" "[:^prop:]",
1207 plus all the properties specified by
1208 <https://www.unicode.org/reports/tr18/#Compatibility_Properties>.
1209 These are described above in "Other Properties"
1210 [4] The regex sets feature "(?[...])" starting in v5.18 accomplishes
1212 this. See "(?[ ])" in perlre.
1213 [5] "\b" "\B" meet most, but not all, the details of this requirement,
1215 but "\b{wb}" and "\B{wb}" do, as well as the stricter R2.3.
1216 [6] Note that Perl does Full case-folding in matching, not Simple:
1217 For example "U+1F88" is equivalent to "U+1F00 U+03B9", instead of
1219 just "U+1F80". This difference matters mainly for certain Greek
1220 capital letters with certain modifiers: the Full case-folding
1221 decomposes the letter, while the Simple case-folding would map it
1222 to a single character.
1223 [7] The reason this is considered to be only partially implemented is
1225 that Perl has "qr/\b{lb}/" and "Unicode::LineBreak" that are
1226 conformant with UAX#14 "Unicode Line Breaking Algorithm"
1227 <https://www.unicode.org/reports/tr14>. The regular expression
1228 construct provides default behavior, while the heavier-weight
1229 module provides customizable line breaking.
1230 But Perl treats "\n" as the start- and end-line delimiter, whereas
1232 Unicode specifies more characters that should be so-interpreted.
1233 These are:
1235 VT U+000B (\v in C)
1237 FF U+000C (\f)
1238 CR U+000D (\r)
1239 NEL U+0085
1240 LS U+2028
1241 PS U+2029
1242 "^" and "$" in regular expression patterns are supposed to match
1244 all these, but don't. These characters also don't, but should,
1245 affect "<>" $., and script line numbers.
1246 Also, lines should not be split within "CRLF" (i.e. there is no
1248 empty line between "\r" and "\n"). For "CRLF", try the ":crlf"
1249 layer (see PerlIO).
1250 [8] UTF-8/UTF-EBDDIC used in Perl allows not only "U+10000" to
1252 "U+10FFFF" but also beyond "U+10FFFF"
1253 Level 2 - Extended Unicode Support
1255 RL2.1 Canonical Equivalents - Retracted [9]
1257 by Unicode
1258 RL2.2 Extended Grapheme Clusters and - Partial [10]
1259 Character Classes with Strings
1260 RL2.3 Default Word Boundaries - Done [11]
1261 RL2.4 Default Case Conversion - Done
1262 RL2.5 Name Properties - Done
1263 RL2.6 Wildcards in Property Values - Partial [12]
1264 RL2.7 Full Properties - Partial [13]
1265 RL2.8 Optional Properties - Partial [14]
1266 [9] Unicode has rewritten this portion of UTS#18 to say that getting
1268 canonical equivalence (see UAX#15 "Unicode Normalization Forms"
1269 <https://www.unicode.org/reports/tr15>) is basically to be done at the
1270 programmer level. Use NFD to write both your regular expressions and
1271 text to match them against (you can use Unicode::Normalize).
1272 [10] Perl has "\X" and "\b{gcb}". Unicode has retracted their
1273 "Grapheme Cluster Mode", and recently added string properties, which
1274 Perl does not yet support.
1275 [11] see UAX#29 "Unicode Text Segmentation"
1276 <https://www.unicode.org/reports/tr29>,
1277 [12] see "Wildcards in Property Values" above.
1278 [13] Perl supports all the properties in the Unicode Character Database
1279 (UCD). It does not yet support the listed properties that come from
1280 other Unicode sources.
1281 [14] The only optional property that Perl supports is Named Sequence.
1282 None of these properties are in the UCD.
1283 Level 3 - Tailored Support
1285 This has been retracted by Unicode.
1287 Unicode Encodings
1289 Unicode characters are assigned to code points, which are abstract
1290 numbers. To use these numbers, various encodings are needed.
1291 • UTF-8
1293 UTF-8 is a variable-length (1 to 4 bytes), byte-order independent
1295 encoding. In most of Perl's documentation, including elsewhere in
1296 this document, the term "UTF-8" means also "UTF-EBCDIC". But in
1297 this section, "UTF-8" refers only to the encoding used on ASCII
1298 platforms. It is a superset of 7-bit US-ASCII, so anything encoded
1299 in ASCII has the identical representation when encoded in UTF-8.
1300 The following table is from Unicode 3.2.
1302 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1304 U+0000..U+007F 00..7F
1306 U+0080..U+07FF * C2..DF 80..BF
1307 U+0800..U+0FFF E0 * A0..BF 80..BF
1308 U+1000..U+CFFF E1..EC 80..BF 80..BF
1309 U+D000..U+D7FF ED 80..9F 80..BF
1310 U+D800..U+DFFF +++++ utf16 surrogates, not legal utf8 +++++
1311 U+E000..U+FFFF EE..EF 80..BF 80..BF
1312 U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF
1313 U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
1314 U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
1315 Note the gaps marked by "*" before several of the byte entries
1317 above. These are caused by legal UTF-8 avoiding non-shortest
1318 encodings: it is technically possible to UTF-8-encode a single code
1319 point in different ways, but that is explicitly forbidden, and the
1320 shortest possible encoding should always be used (and that is what
1321 Perl does).
1322 Another way to look at it is via bits:
1324 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1326 0aaaaaaa 0aaaaaaa
1328 00000bbbbbaaaaaa 110bbbbb 10aaaaaa
1329 ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
1330 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
1331 As you can see, the continuation bytes all begin with "10", and the
1333 leading bits of the start byte tell how many bytes there are in the
1334 encoded character.
1335 The original UTF-8 specification allowed up to 6 bytes, to allow
1337 encoding of numbers up to "0x7FFF_FFFF". Perl continues to allow
1338 those, and has extended that up to 13 bytes to encode code points
1339 up to what can fit in a 64-bit word. However, Perl will warn if
1340 you output any of these as being non-portable; and under strict
1341 UTF-8 input protocols, they are forbidden. In addition, it is now
1342 illegal to use a code point larger than what a signed integer
1343 variable on your system can hold. On 32-bit ASCII systems, this
1344 means "0x7FFF_FFFF" is the legal maximum (much higher on 64-bit
1345 systems).
1346 • UTF-EBCDIC
1348 Like UTF-8, but EBCDIC-safe, in the way that UTF-8 is ASCII-safe.
1350 This means that all the basic characters (which includes all those
1351 that have ASCII equivalents (like "A", "0", "%", etc.) are the
1352 same in both EBCDIC and UTF-EBCDIC.)
1353 UTF-EBCDIC is used on EBCDIC platforms. It generally requires more
1355 bytes to represent a given code point than UTF-8 does; the largest
1356 Unicode code points take 5 bytes to represent (instead of 4 in
1357 UTF-8), and, extended for 64-bit words, it uses 14 bytes instead of
1358 13 bytes in UTF-8.
1359 • UTF-16, UTF-16BE, UTF-16LE, Surrogates, and "BOM"'s (Byte Order
1361 Marks)
1362 The followings items are mostly for reference and general Unicode
1364 knowledge, Perl doesn't use these constructs internally.
1365 Like UTF-8, UTF-16 is a variable-width encoding, but where UTF-8
1367 uses 8-bit code units, UTF-16 uses 16-bit code units. All code
1368 points occupy either 2 or 4 bytes in UTF-16: code points
1369 "U+0000..U+FFFF" are stored in a single 16-bit unit, and code
1370 points "U+10000..U+10FFFF" in two 16-bit units. The latter case is
1371 using surrogates, the first 16-bit unit being the high surrogate,
1372 and the second being the low surrogate.
1373 Surrogates are code points set aside to encode the
1375 "U+10000..U+10FFFF" range of Unicode code points in pairs of 16-bit
1376 units. The high surrogates are the range "U+D800..U+DBFF" and the
1377 low surrogates are the range "U+DC00..U+DFFF". The surrogate
1378 encoding is
1379 $hi = ($uni - 0x10000) / 0x400 + 0xD800;
1381 $lo = ($uni - 0x10000) % 0x400 + 0xDC00;
1382 and the decoding is
1384 $uni = 0x10000 + ($hi - 0xD800) * 0x400 + ($lo - 0xDC00);
1386 Because of the 16-bitness, UTF-16 is byte-order dependent. UTF-16
1388 itself can be used for in-memory computations, but if storage or
1389 transfer is required either UTF-16BE (big-endian) or UTF-16LE
1390 (little-endian) encodings must be chosen.
1391 This introduces another problem: what if you just know that your
1393 data is UTF-16, but you don't know which endianness? Byte Order
1394 Marks, or "BOM"'s, are a solution to this. A special character has
1395 been reserved in Unicode to function as a byte order marker: the
1396 character with the code point "U+FEFF" is the "BOM".
1397 The trick is that if you read a "BOM", you will know the byte
1399 order, since if it was written on a big-endian platform, you will
1400 read the bytes "0xFE 0xFF", but if it was written on a little-
1401 endian platform, you will read the bytes "0xFF 0xFE". (And if the
1402 originating platform was writing in ASCII platform UTF-8, you will
1403 read the bytes "0xEF 0xBB 0xBF".)
1404 The way this trick works is that the character with the code point
1406 "U+FFFE" is not supposed to be in input streams, so the sequence of
1407 bytes "0xFF 0xFE" is unambiguously ""BOM", represented in little-
1408 endian format" and cannot be "U+FFFE", represented in big-endian
1409 format".
1410 Surrogates have no meaning in Unicode outside their use in pairs to
1412 represent other code points. However, Perl allows them to be
1413 represented individually internally, for example by saying
1414 chr(0xD801), so that all code points, not just those valid for open
1415 interchange, are representable. Unicode does define semantics for
1416 them, such as their "General_Category" is "Cs". But because their
1417 use is somewhat dangerous, Perl will warn (using the warning
1418 category "surrogate", which is a sub-category of "utf8") if an
1419 attempt is made to do things like take the lower case of one, or
1420 match case-insensitively, or to output them. (But don't try this
1421 on Perls before 5.14.)
1422 • UTF-32, UTF-32BE, UTF-32LE
1424 The UTF-32 family is pretty much like the UTF-16 family, except
1426 that the units are 32-bit, and therefore the surrogate scheme is
1427 not needed. UTF-32 is a fixed-width encoding. The "BOM"
1428 signatures are "0x00 0x00 0xFE 0xFF" for BE and "0xFF 0xFE 0x00
1429 0x00" for LE.
1430 • UCS-2, UCS-4
1432 Legacy, fixed-width encodings defined by the ISO 10646 standard.
1434 UCS-2 is a 16-bit encoding. Unlike UTF-16, UCS-2 is not extensible
1435 beyond "U+FFFF", because it does not use surrogates. UCS-4 is a
1436 32-bit encoding, functionally identical to UTF-32 (the difference
1437 being that UCS-4 forbids neither surrogates nor code points larger
1438 than "0x10_FFFF").
1439 • UTF-7
1441 A seven-bit safe (non-eight-bit) encoding, which is useful if the
1443 transport or storage is not eight-bit safe. Defined by RFC 2152.
1444 Noncharacter code points
1446 66 code points are set aside in Unicode as "noncharacter code points".
1447 These all have the "Unassigned" ("Cn") "General_Category", and no
1448 character will ever be assigned to any of them. They are the 32 code
1449 points between "U+FDD0" and "U+FDEF" inclusive, and the 34 code points:
1450 U+FFFE U+FFFF
1452 U+1FFFE U+1FFFF
1453 U+2FFFE U+2FFFF
1454 ...
1455 U+EFFFE U+EFFFF
1456 U+FFFFE U+FFFFF
1457 U+10FFFE U+10FFFF
1458 Until Unicode 7.0, the noncharacters were "forbidden for use in open
1460 interchange of Unicode text data", so that code that processed those
1461 streams could use these code points as sentinels that could be mixed in
1462 with character data, and would always be distinguishable from that
1463 data. (Emphasis above and in the next paragraph are added in this
1464 document.)
1465 Unicode 7.0 changed the wording so that they are "not recommended for
1467 use in open interchange of Unicode text data". The 7.0 Standard goes
1468 on to say:
1469 "If a noncharacter is received in open interchange, an application
1471 is not required to interpret it in any way. It is good practice,
1472 however, to recognize it as a noncharacter and to take appropriate
1473 action, such as replacing it with "U+FFFD" replacement character,
1474 to indicate the problem in the text. It is not recommended to
1475 simply delete noncharacter code points from such text, because of
1476 the potential security issues caused by deleting uninterpreted
1477 characters. (See conformance clause C7 in Section 3.2, Conformance
1478 Requirements, and Unicode Technical Report #36, "Unicode Security
1479 Considerations"
1480 <https://www.unicode.org/reports/tr36/#Substituting_for_Ill_Formed_Subsequences>)."
1481 This change was made because it was found that various commercial tools
1483 like editors, or for things like source code control, had been written
1484 so that they would not handle program files that used these code
1485 points, effectively precluding their use almost entirely! And that was
1486 never the intent. They've always been meant to be usable within an
1487 application, or cooperating set of applications, at will.
1488 If you're writing code, such as an editor, that is supposed to be able
1490 to handle any Unicode text data, then you shouldn't be using these code
1491 points yourself, and instead allow them in the input. If you need
1492 sentinels, they should instead be something that isn't legal Unicode.
1493 For UTF-8 data, you can use the bytes 0xC0 and 0xC1 as sentinels, as
1494 they never appear in well-formed UTF-8. (There are equivalents for
1495 UTF-EBCDIC). You can also store your Unicode code points in integer
1496 variables and use negative values as sentinels.
1497 If you're not writing such a tool, then whether you accept
1499 noncharacters as input is up to you (though the Standard recommends
1500 that you not). If you do strict input stream checking with Perl, these
1501 code points continue to be forbidden. This is to maintain backward
1502 compatibility (otherwise potential security holes could open up, as an
1503 unsuspecting application that was written assuming the noncharacters
1504 would be filtered out before getting to it, could now, without warning,
1505 start getting them). To do strict checking, you can use the layer
1506 :encoding('UTF-8').
1507 Perl continues to warn (using the warning category "nonchar", which is
1509 a sub-category of "utf8") if an attempt is made to output
1510 noncharacters.
1511 Beyond Unicode code points
1513 The maximum Unicode code point is "U+10FFFF", and Unicode only defines
1514 operations on code points up through that. But Perl works on code
1515 points up to the maximum permissible signed number available on the
1516 platform. However, Perl will not accept these from input streams
1517 unless lax rules are being used, and will warn (using the warning
1518 category "non_unicode", which is a sub-category of "utf8") if any are
1519 output.
1520 Since Unicode rules are not defined on these code points, if a Unicode-
1522 defined operation is done on them, Perl uses what we believe are
1523 sensible rules, while generally warning, using the "non_unicode"
1524 category. For example, uc("\x{11_0000}") will generate such a warning,
1525 returning the input parameter as its result, since Perl defines the
1526 uppercase of every non-Unicode code point to be the code point itself.
1527 (All the case changing operations, not just uppercasing, work this
1528 way.)
1529 The situation with matching Unicode properties in regular expressions,
1531 the "\p{}" and "\P{}" constructs, against these code points is not as
1532 clear cut, and how these are handled has changed as we've gained
1533 experience.
1534 One possibility is to treat any match against these code points as
1536 undefined. But since Perl doesn't have the concept of a match being
1537 undefined, it converts this to failing or "FALSE". This is almost, but
1538 not quite, what Perl did from v5.14 (when use of these code points
1539 became generally reliable) through v5.18. The difference is that Perl
1540 treated all "\p{}" matches as failing, but all "\P{}" matches as
1541 succeeding.
1542 One problem with this is that it leads to unexpected, and confusing
1544 results in some cases:
1545 chr(0x110000) =~ \p{ASCII_Hex_Digit=True} # Failed on <= v5.18
1547 chr(0x110000) =~ \p{ASCII_Hex_Digit=False} # Failed! on <= v5.18
1548 That is, it treated both matches as undefined, and converted that to
1550 false (raising a warning on each). The first case is the expected
1551 result, but the second is likely counterintuitive: "How could both be
1552 false when they are complements?" Another problem was that the
1553 implementation optimized many Unicode property matches down to already
1554 existing simpler, faster operations, which don't raise the warning. We
1555 chose to not forgo those optimizations, which help the vast majority of
1556 matches, just to generate a warning for the unlikely event that an
1557 above-Unicode code point is being matched against.
1558 As a result of these problems, starting in v5.20, what Perl does is to
1560 treat non-Unicode code points as just typical unassigned Unicode
1561 characters, and matches accordingly. (Note: Unicode has atypical
1562 unassigned code points. For example, it has noncharacter code points,
1563 and ones that, when they do get assigned, are destined to be written
1564 Right-to-left, as Arabic and Hebrew are. Perl assumes that no non-
1565 Unicode code point has any atypical properties.)
1566 Perl, in most cases, will raise a warning when matching an above-
1568 Unicode code point against a Unicode property when the result is "TRUE"
1569 for "\p{}", and "FALSE" for "\P{}". For example:
1570 chr(0x110000) =~ \p{ASCII_Hex_Digit=True} # Fails, no warning
1572 chr(0x110000) =~ \p{ASCII_Hex_Digit=False} # Succeeds, with warning
1573 In both these examples, the character being matched is non-Unicode, so
1575 Unicode doesn't define how it should match. It clearly isn't an ASCII
1576 hex digit, so the first example clearly should fail, and so it does,
1577 with no warning. But it is arguable that the second example should
1578 have an undefined, hence "FALSE", result. So a warning is raised for
1579 it.
1580 Thus the warning is raised for many fewer cases than in earlier Perls,
1582 and only when what the result is could be arguable. It turns out that
1583 none of the optimizations made by Perl (or are ever likely to be made)
1584 cause the warning to be skipped, so it solves both problems of Perl's
1585 earlier approach. The most commonly used property that is affected by
1586 this change is "\p{Unassigned}" which is a short form for
1587 "\p{General_Category=Unassigned}". Starting in v5.20, all non-Unicode
1588 code points are considered "Unassigned". In earlier releases the
1589 matches failed because the result was considered undefined.
1590 The only place where the warning is not raised when it might ought to
1592 have been is if optimizations cause the whole pattern match to not even
1593 be attempted. For example, Perl may figure out that for a string to
1594 match a certain regular expression pattern, the string has to contain
1595 the substring "foobar". Before attempting the match, Perl may look for
1596 that substring, and if not found, immediately fail the match without
1597 actually trying it; so no warning gets generated even if the string
1598 contains an above-Unicode code point.
1599 This behavior is more "Do what I mean" than in earlier Perls for most
1601 applications. But it catches fewer issues for code that needs to be
1602 strictly Unicode compliant. Therefore there is an additional mode of
1603 operation available to accommodate such code. This mode is enabled if
1604 a regular expression pattern is compiled within the lexical scope where
1605 the "non_unicode" warning class has been made fatal, say by:
1606 use warnings FATAL => "non_unicode"
1608 (see warnings). In this mode of operation, Perl will raise the warning
1610 for all matches against a non-Unicode code point (not just the arguable
1611 ones), and it skips the optimizations that might cause the warning to
1612 not be output. (It currently still won't warn if the match isn't even
1613 attempted, like in the "foobar" example above.)
1614 In summary, Perl now normally treats non-Unicode code points as typical
1616 Unicode unassigned code points for regular expression matches, raising
1617 a warning only when it is arguable what the result should be. However,
1618 if this warning has been made fatal, it isn't skipped.
1619 There is one exception to all this. "\p{All}" looks like a Unicode
1621 property, but it is a Perl extension that is defined to be true for all
1622 possible code points, Unicode or not, so no warning is ever generated
1623 when matching this against a non-Unicode code point. (Prior to v5.20,
1624 it was an exact synonym for "\p{Any}", matching code points 0 through
1625 0x10FFFF.)
1626 Security Implications of Unicode
1628 First, read Unicode Security Considerations
1629 <https://www.unicode.org/reports/tr36>.
1630 Also, note the following:
1632 • Malformed UTF-8
1634 UTF-8 is very structured, so many combinations of bytes are
1636 invalid. In the past, Perl tried to soldier on and make some sense
1637 of invalid combinations, but this can lead to security holes, so
1638 now, if the Perl core needs to process an invalid combination, it
1639 will either raise a fatal error, or will replace those bytes by the
1640 sequence that forms the Unicode REPLACEMENT CHARACTER, for which
1641 purpose Unicode created it.
1642 Every code point can be represented by more than one possible
1644 syntactically valid UTF-8 sequence. Early on, both Unicode and
1645 Perl considered any of these to be valid, but now, all sequences
1646 longer than the shortest possible one are considered to be
1647 malformed.
1648 Unicode considers many code points to be illegal, or to be avoided.
1650 Perl generally accepts them, once they have passed through any
1651 input filters that may try to exclude them. These have been
1652 discussed above (see "Surrogates" under UTF-16 in "Unicode
1653 Encodings", "Noncharacter code points", and "Beyond Unicode code
1654 points").
1655 • Regular expression pattern matching may surprise you if you're not
1657 accustomed to Unicode. Starting in Perl 5.14, several pattern
1658 modifiers are available to control this, called the character set
1659 modifiers. Details are given in "Character set modifiers" in
1660 perlre.
1661 As discussed elsewhere, Perl has one foot (two hooves?) planted in each
1663 of two worlds: the old world of ASCII and single-byte locales, and the
1664 new world of Unicode, upgrading when necessary. If your legacy code
1665 does not explicitly use Unicode, no automatic switch-over to Unicode
1666 should happen.
1667 Unicode in Perl on EBCDIC
1669 Unicode is supported on EBCDIC platforms. See perlebcdic.
1670 Unless ASCII vs. EBCDIC issues are specifically being discussed,
1672 references to UTF-8 encoding in this document and elsewhere should be
1673 read as meaning UTF-EBCDIC on EBCDIC platforms. See "Unicode and UTF"
1674 in perlebcdic.
1675 Because UTF-EBCDIC is so similar to UTF-8, the differences are mostly
1677 hidden from you; "use utf8" (and NOT something like "use utfebcdic")
1678 declares the script is in the platform's "native" 8-bit encoding of
1679 Unicode. (Similarly for the ":utf8" layer.)
1680 Locales
1682 See "Unicode and UTF-8" in perllocale
1683 When Unicode Does Not Happen
1685 There are still many places where Unicode (in some encoding or another)
1686 could be given as arguments or received as results, or both in Perl,
1687 but it is not, in spite of Perl having extensive ways to input and
1688 output in Unicode, and a few other "entry points" like the @ARGV array
1689 (which can sometimes be interpreted as UTF-8).
1690 The following are such interfaces. Also, see "The "Unicode Bug"". For
1692 all of these interfaces Perl currently (as of v5.16.0) simply assumes
1693 byte strings both as arguments and results, or UTF-8 strings if the
1694 (deprecated) "encoding" pragma has been used.
1695 One reason that Perl does not attempt to resolve the role of Unicode in
1697 these situations is that the answers are highly dependent on the
1698 operating system and the file system(s). For example, whether
1699 filenames can be in Unicode and in exactly what kind of encoding, is
1700 not exactly a portable concept. Similarly for "qx" and "system": how
1701 well will the "command-line interface" (and which of them?) handle
1702 Unicode?
1703 • "chdir", "chmod", "chown", "chroot", "exec", "link", "lstat",
1705 "mkdir", "rename", "rmdir", "stat", "symlink", "truncate",
1706 "unlink", "utime", "-X"
1707 • %ENV
1709 • "glob" (aka the "<*>")
1711 • "open", "opendir", "sysopen"
1713 • "qx" (aka the backtick operator), "system"
1715 • "readdir", "readlink"
1717 The "Unicode Bug"
1719 The term, "Unicode bug" has been applied to an inconsistency with the
1720 code points in the "Latin-1 Supplement" block, that is, between 128 and
1721 255. Without a locale specified, unlike all other characters or code
1722 points, these characters can have very different semantics depending on
1723 the rules in effect. (Characters whose code points are above 255 force
1724 Unicode rules; whereas the rules for ASCII characters are the same
1725 under both ASCII and Unicode rules.)
1726 Under Unicode rules, these upper-Latin1 characters are interpreted as
1728 Unicode code points, which means they have the same semantics as
1729 Latin-1 (ISO-8859-1) and C1 controls.
1730 As explained in "ASCII Rules versus Unicode Rules", under ASCII rules,
1732 they are considered to be unassigned characters.
1733 This can lead to unexpected results. For example, a string's semantics
1735 can suddenly change if a code point above 255 is appended to it, which
1736 changes the rules from ASCII to Unicode. As an example, consider the
1737 following program and its output:
1738 $ perl -le'
1740 no feature "unicode_strings";
1741 $s1 = "\xC2";
1742 $s2 = "\x{2660}";
1743 for ($s1, $s2, $s1.$s2) {
1744 print /\w/ || 0;
1745 }
1746 '
1747 0
1748 0
1749 1
1750 If there's no "\w" in "s1" nor in "s2", why does their concatenation
1752 have one?
1753 This anomaly stems from Perl's attempt to not disturb older programs
1755 that didn't use Unicode, along with Perl's desire to add Unicode
1756 support seamlessly. But the result turned out to not be seamless. (By
1757 the way, you can choose to be warned when things like this happen. See
1758 "encoding::warnings".)
1759 "use feature 'unicode_strings'" was added, starting in Perl v5.12, to
1761 address this problem. It affects these things:
1762 • Changing the case of a scalar, that is, using uc(), ucfirst(),
1764 lc(), and lcfirst(), or "\L", "\U", "\u" and "\l" in double-quotish
1765 contexts, such as regular expression substitutions.
1766 Under "unicode_strings" starting in Perl 5.12.0, Unicode rules are
1768 generally used. See "lc" in perlfunc for details on how this works
1769 in combination with various other pragmas.
1770 • Using caseless ("/i") regular expression matching.
1772 Starting in Perl 5.14.0, regular expressions compiled within the
1774 scope of "unicode_strings" use Unicode rules even when executed or
1775 compiled into larger regular expressions outside the scope.
1776 • Matching any of several properties in regular expressions.
1778 These properties are "\b" (without braces), "\B" (without braces),
1780 "\s", "\S", "\w", "\W", and all the Posix character classes except
1781 "[[:ascii:]]".
1782 Starting in Perl 5.14.0, regular expressions compiled within the
1784 scope of "unicode_strings" use Unicode rules even when executed or
1785 compiled into larger regular expressions outside the scope.
1786 • In "quotemeta" or its inline equivalent "\Q".
1788 Starting in Perl 5.16.0, consistent quoting rules are used within
1790 the scope of "unicode_strings", as described in "quotemeta" in
1791 perlfunc. Prior to that, or outside its scope, no code points
1792 above 127 are quoted in UTF-8 encoded strings, but in byte encoded
1793 strings, code points between 128-255 are always quoted.
1794 • In the ".." or range operator.
1796 Starting in Perl 5.26.0, the range operator on strings treats their
1798 lengths consistently within the scope of "unicode_strings". Prior
1799 to that, or outside its scope, it could produce strings whose
1800 length in characters exceeded that of the right-hand side, where
1801 the right-hand side took up more bytes than the correct range
1802 endpoint.
1803 • In "split"'s special-case whitespace splitting.
1805 Starting in Perl 5.28.0, the "split" function with a pattern
1807 specified as a string containing a single space handles whitespace
1808 characters consistently within the scope of "unicode_strings".
1809 Prior to that, or outside its scope, characters that are whitespace
1810 according to Unicode rules but not according to ASCII rules were
1811 treated as field contents rather than field separators when they
1812 appear in byte-encoded strings.
1813 You can see from the above that the effect of "unicode_strings"
1815 increased over several Perl releases. (And Perl's support for Unicode
1816 continues to improve; it's best to use the latest available release in
1817 order to get the most complete and accurate results possible.) Note
1818 that "unicode_strings" is automatically chosen if you "use v5.12" or
1819 higher.
1820 For Perls earlier than those described above, or when a string is
1822 passed to a function outside the scope of "unicode_strings", see the
1823 next section.
1824 Forcing Unicode in Perl (Or Unforcing Unicode in Perl)
1826 Sometimes (see "When Unicode Does Not Happen" or "The "Unicode Bug"")
1827 there are situations where you simply need to force a byte string into
1828 UTF-8, or vice versa. The standard module Encode can be used for this,
1829 or the low-level calls utf8::upgrade($bytestring) and
1830 "utf8::downgrade($utf8string[, FAIL_OK])".
1831 Note that utf8::downgrade() can fail if the string contains characters
1833 that don't fit into a byte.
1834 Calling either function on a string that already is in the desired
1836 state is a no-op.
1837 "ASCII Rules versus Unicode Rules" gives all the ways that a string is
1839 made to use Unicode rules.
1840 Using Unicode in XS
1842 See "Unicode Support" in perlguts for an introduction to Unicode at the
1843 XS level, and "Unicode Support" in perlapi for the API details.
1844 Hacking Perl to work on earlier Unicode versions (for very serious hackers
1846 only)
1847 Perl by default comes with the latest supported Unicode version built-
1848 in, but the goal is to allow you to change to use any earlier one. In
1849 Perls v5.20 and v5.22, however, the earliest usable version is Unicode
1850 5.1. Perl v5.18 and v5.24 are able to handle all earlier versions.
1851 Download the files in the desired version of Unicode from the Unicode
1853 web site <https://www.unicode.org>). These should replace the existing
1854 files in lib/unicore in the Perl source tree. Follow the instructions
1855 in README.perl in that directory to change some of their names, and
1856 then build perl (see INSTALL).
1857 Porting code from perl-5.6.X
1859 Perls starting in 5.8 have a different Unicode model from 5.6. In 5.6
1860 the programmer was required to use the "utf8" pragma to declare that a
1861 given scope expected to deal with Unicode data and had to make sure
1862 that only Unicode data were reaching that scope. If you have code that
1863 is working with 5.6, you will need some of the following adjustments to
1864 your code. The examples are written such that the code will continue to
1865 work under 5.6, so you should be safe to try them out.
1866 • A filehandle that should read or write UTF-8
1868 if ($] > 5.008) {
1870 binmode $fh, ":encoding(UTF-8)";
1871 }
1872 • A scalar that is going to be passed to some extension
1874 Be it "Compress::Zlib", "Apache::Request" or any extension that has
1876 no mention of Unicode in the manpage, you need to make sure that the
1877 UTF8 flag is stripped off. Note that at the time of this writing
1878 (January 2012) the mentioned modules are not UTF-8-aware. Please
1879 check the documentation to verify if this is still true.
1880 if ($] > 5.008) {
1882 require Encode;
1883 $val = Encode::encode("UTF-8", $val); # make octets
1884 }
1885 • A scalar we got back from an extension
1887 If you believe the scalar comes back as UTF-8, you will most likely
1889 want the UTF8 flag restored:
1890 if ($] > 5.008) {
1892 require Encode;
1893 $val = Encode::decode("UTF-8", $val);
1894 }
1895 • Same thing, if you are really sure it is UTF-8
1897 if ($] > 5.008) {
1899 require Encode;
1900 Encode::_utf8_on($val);
1901 }
1902 • A wrapper for DBI "fetchrow_array" and "fetchrow_hashref"
1904 When the database contains only UTF-8, a wrapper function or method
1906 is a convenient way to replace all your "fetchrow_array" and
1907 "fetchrow_hashref" calls. A wrapper function will also make it
1908 easier to adapt to future enhancements in your database driver. Note
1909 that at the time of this writing (January 2012), the DBI has no
1910 standardized way to deal with UTF-8 data. Please check the DBI
1911 documentation to verify if that is still true.
1912 sub fetchrow {
1914 # $what is one of fetchrow_{array,hashref}
1915 my($self, $sth, $what) = @_;
1916 if ($] < 5.008) {
1917 return $sth->$what;
1918 } else {
1919 require Encode;
1920 if (wantarray) {
1921 my @arr = $sth->$what;
1922 for (@arr) {
1923 defined && /[^\000-\177]/ && Encode::_utf8_on($_);
1924 }
1925 return @arr;
1926 } else {
1927 my $ret = $sth->$what;
1928 if (ref $ret) {
1929 for my $k (keys %$ret) {
1930 defined
1931 && /[^\000-\177]/
1932 && Encode::_utf8_on($_) for $ret->{$k};
1933 }
1934 return $ret;
1935 } else {
1936 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret;
1937 return $ret;
1938 }
1939 }
1940 }
1941 }
1942 • A large scalar that you know can only contain ASCII
1944 Scalars that contain only ASCII and are marked as UTF-8 are
1946 sometimes a drag to your program. If you recognize such a situation,
1947 just remove the UTF8 flag:
1948 utf8::downgrade($val) if $] > 5.008;
1950
1952 See also "The "Unicode Bug"" above.
1953 Interaction with Extensions
1955 When Perl exchanges data with an extension, the extension should be
1956 able to understand the UTF8 flag and act accordingly. If the extension
1957 doesn't recognize that flag, it's likely that the extension will return
1958 incorrectly-flagged data.
1959 So if you're working with Unicode data, consult the documentation of
1961 every module you're using if there are any issues with Unicode data
1962 exchange. If the documentation does not talk about Unicode at all,
1963 suspect the worst and probably look at the source to learn how the
1964 module is implemented. Modules written completely in Perl shouldn't
1965 cause problems. Modules that directly or indirectly access code written
1966 in other programming languages are at risk.
1967 For affected functions, the simple strategy to avoid data corruption is
1969 to always make the encoding of the exchanged data explicit. Choose an
1970 encoding that you know the extension can handle. Convert arguments
1971 passed to the extensions to that encoding and convert results back from
1972 that encoding. Write wrapper functions that do the conversions for you,
1973 so you can later change the functions when the extension catches up.
1974 To provide an example, let's say the popular "Foo::Bar::escape_html"
1976 function doesn't deal with Unicode data yet. The wrapper function would
1977 convert the argument to raw UTF-8 and convert the result back to Perl's
1978 internal representation like so:
1979 sub my_escape_html ($) {
1981 my($what) = shift;
1982 return unless defined $what;
1983 Encode::decode("UTF-8", Foo::Bar::escape_html(
1984 Encode::encode("UTF-8", $what)));
1985 }
1986 Sometimes, when the extension does not convert data but just stores and
1988 retrieves it, you will be able to use the otherwise dangerous
1989 Encode::_utf8_on() function. Let's say the popular "Foo::Bar"
1990 extension, written in C, provides a "param" method that lets you store
1991 and retrieve data according to these prototypes:
1992 $self->param($name, $value); # set a scalar
1994 $value = $self->param($name); # retrieve a scalar
1995 If it does not yet provide support for any encoding, one could write a
1997 derived class with such a "param" method:
1998 sub param {
2000 my($self,$name,$value) = @_;
2001 utf8::upgrade($name); # make sure it is UTF-8 encoded
2002 if (defined $value) {
2003 utf8::upgrade($value); # make sure it is UTF-8 encoded
2004 return $self->SUPER::param($name,$value);
2005 } else {
2006 my $ret = $self->SUPER::param($name);
2007 Encode::_utf8_on($ret); # we know, it is UTF-8 encoded
2008 return $ret;
2009 }
2010 }
2011 Some extensions provide filters on data entry/exit points, such as
2013 "DB_File::filter_store_key" and family. Look out for such filters in
2014 the documentation of your extensions; they can make the transition to
2015 Unicode data much easier.
2016 Speed
2018 Some functions are slower when working on UTF-8 encoded strings than on
2019 byte encoded strings. All functions that need to hop over characters
2020 such as length(), substr() or index(), or matching regular expressions
2021 can work much faster when the underlying data are byte-encoded.
2022 In Perl 5.8.0 the slowness was often quite spectacular; in Perl 5.8.1 a
2024 caching scheme was introduced which improved the situation. In
2025 general, operations with UTF-8 encoded strings are still slower. As an
2026 example, the Unicode properties (character classes) like "\p{Nd}" are
2027 known to be quite a bit slower (5-20 times) than their simpler
2028 counterparts like "[0-9]" (then again, there are hundreds of Unicode
2029 characters matching "Nd" compared with the 10 ASCII characters matching
2030 "[0-9]").
2031
2033 perlunitut, perluniintro, perluniprops, Encode, open, utf8, bytes,
2034 perlretut, "${^UNICODE}" in perlvar,
2035 <https://www.unicode.org/reports/tr44>).
2036perl v5.38.2 2023-11-30 PERLUNICODE(1)