3 perlunicode - Unicode support in Perl
7 =head2 Important Caveats
9 Unicode support is an extensive requirement. While Perl does not
10 implement the Unicode standard or the accompanying technical reports
11 from cover to cover, Perl does support many Unicode features.
13 People who want to learn to use Unicode in Perl, should probably read
14 the L<Perl Unicode tutorial, perlunitut|perlunitut>, before reading
15 this reference document.
17 Also, the use of Unicode may present security issues that aren't obvious.
18 Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>.
22 =item Input and Output Layers
24 Perl knows when a filehandle uses Perl's internal Unicode encodings
25 (UTF-8, or UTF-EBCDIC if in EBCDIC) if the filehandle is opened with
26 the ":utf8" layer. Other encodings can be converted to Perl's
27 encoding on input or from Perl's encoding on output by use of the
28 ":encoding(...)" layer. See L<open>.
30 To indicate that Perl source itself is in UTF-8, use C<use utf8;>.
32 =item Regular Expressions
34 The regular expression compiler produces polymorphic opcodes. That is,
35 the pattern adapts to the data and automatically switches to the Unicode
36 character scheme when presented with data that is internally encoded in
37 UTF-8, or instead uses a traditional byte scheme when presented with
40 =item C<use utf8> still needed to enable UTF-8/UTF-EBCDIC in scripts
42 As a compatibility measure, the C<use utf8> pragma must be explicitly
43 included to enable recognition of UTF-8 in the Perl scripts themselves
44 (in string or regular expression literals, or in identifier names) on
45 ASCII-based machines or to recognize UTF-EBCDIC on EBCDIC-based
46 machines. B<These are the only times when an explicit C<use utf8>
47 is needed.> See L<utf8>.
49 =item BOM-marked scripts and UTF-16 scripts autodetected
51 If a Perl script begins marked with the Unicode BOM (UTF-16LE, UTF16-BE,
52 or UTF-8), or if the script looks like non-BOM-marked UTF-16 of either
53 endianness, Perl will correctly read in the script as Unicode.
54 (BOMless UTF-8 cannot be effectively recognized or differentiated from
55 ISO 8859-1 or other eight-bit encodings.)
57 =item C<use encoding> needed to upgrade non-Latin-1 byte strings
59 By default, there is a fundamental asymmetry in Perl's Unicode model:
60 implicit upgrading from byte strings to Unicode strings assumes that
61 they were encoded in I<ISO 8859-1 (Latin-1)>, but Unicode strings are
62 downgraded with UTF-8 encoding. This happens because the first 256
63 codepoints in Unicode happens to agree with Latin-1.
65 See L</"Byte and Character Semantics"> for more details.
69 =head2 Byte and Character Semantics
71 Beginning with version 5.6, Perl uses logically-wide characters to
72 represent strings internally.
74 In future, Perl-level operations will be expected to work with
75 characters rather than bytes.
77 However, as an interim compatibility measure, Perl aims to
78 provide a safe migration path from byte semantics to character
79 semantics for programs. For operations where Perl can unambiguously
80 decide that the input data are characters, Perl switches to
81 character semantics. For operations where this determination cannot
82 be made without additional information from the user, Perl decides in
83 favor of compatibility and chooses to use byte semantics.
85 Under byte semantics, when C<use locale> is in effect, Perl uses the
86 semantics associated with the current locale. Absent a C<use locale>, and
87 absent a C<use feature 'unicode_strings'> pragma, Perl currently uses US-ASCII
88 (or Basic Latin in Unicode terminology) byte semantics, meaning that characters
89 whose ordinal numbers are in the range 128 - 255 are undefined except for their
90 ordinal numbers. This means that none have case (upper and lower), nor are any
91 a member of character classes, like C<[:alpha:]> or C<\w>. (But all do belong
92 to the C<\W> class or the Perl regular expression extension C<[:^alpha:]>.)
94 This behavior preserves compatibility with earlier versions of Perl,
95 which allowed byte semantics in Perl operations only if
96 none of the program's inputs were marked as being a source of Unicode
97 character data. Such data may come from filehandles, from calls to
98 external programs, from information provided by the system (such as %ENV),
99 or from literals and constants in the source text.
101 The C<bytes> pragma will always, regardless of platform, force byte
102 semantics in a particular lexical scope. See L<bytes>.
104 The C<use feature 'unicode_strings'> pragma is intended to always, regardless
105 of platform, force character (Unicode) semantics in a particular lexical scope.
106 In release 5.12, it is partially implemented, applying only to case changes.
107 See L</The "Unicode Bug"> below.
109 The C<utf8> pragma is primarily a compatibility device that enables
110 recognition of UTF-(8|EBCDIC) in literals encountered by the parser.
111 Note that this pragma is only required while Perl defaults to byte
112 semantics; when character semantics become the default, this pragma
113 may become a no-op. See L<utf8>.
115 Unless explicitly stated, Perl operators use character semantics
116 for Unicode data and byte semantics for non-Unicode data.
117 The decision to use character semantics is made transparently. If
118 input data comes from a Unicode source--for example, if a character
119 encoding layer is added to a filehandle or a literal Unicode
120 string constant appears in a program--character semantics apply.
121 Otherwise, byte semantics are in effect. The C<bytes> pragma should
122 be used to force byte semantics on Unicode data, and the C<use feature
123 'unicode_strings'> pragma to force Unicode semantics on byte data (though in
124 5.12 it isn't fully implemented).
126 If strings operating under byte semantics and strings with Unicode
127 character data are concatenated, the new string will have
128 character semantics. This can cause surprises: See L</BUGS>, below.
129 You can choose to be warned when this happens. See L<encoding::warnings>.
131 Under character semantics, many operations that formerly operated on
132 bytes now operate on characters. A character in Perl is
133 logically just a number ranging from 0 to 2**31 or so. Larger
134 characters may encode into longer sequences of bytes internally, but
135 this internal detail is mostly hidden for Perl code.
136 See L<perluniintro> for more.
138 =head2 Effects of Character Semantics
140 Character semantics have the following effects:
146 Strings--including hash keys--and regular expression patterns may
147 contain characters that have an ordinal value larger than 255.
149 If you use a Unicode editor to edit your program, Unicode characters may
150 occur directly within the literal strings in UTF-8 encoding, or UTF-16.
151 (The former requires a BOM or C<use utf8>, the latter requires a BOM.)
153 Unicode characters can also be added to a string by using the C<\N{U+...}>
154 notation. The Unicode code for the desired character, in hexadecimal,
155 should be placed in the braces, after the C<U>. For instance, a smiley face is
158 Alternatively, you can use the C<\x{...}> notation for characters 0x100 and
159 above. For characters below 0x100 you may get byte semantics instead of
160 character semantics; see L</The "Unicode Bug">. On EBCDIC machines there is
161 the additional problem that the value for such characters gives the EBCDIC
162 character rather than the Unicode one.
166 use charnames ':full';
168 you can use the C<\N{...}> notation and put the official Unicode
169 character name within the braces, such as C<\N{WHITE SMILING FACE}>.
174 If an appropriate L<encoding> is specified, identifiers within the
175 Perl script may contain Unicode alphanumeric characters, including
176 ideographs. Perl does not currently attempt to canonicalize variable
181 Regular expressions match characters instead of bytes. "." matches
182 a character instead of a byte.
186 Bracketed character classes in regular expressions match characters instead of
187 bytes and match against the character properties specified in the
188 Unicode properties database. C<\w> can be used to match a Japanese
189 ideograph, for instance.
193 Named Unicode properties, scripts, and block ranges may be used (like bracketed
194 character classes) by using the C<\p{}> "matches property" construct and
195 the C<\P{}> negation, "doesn't match property".
196 See L</"Unicode Character Properties"> for more details.
198 You can define your own character properties and use them
199 in the regular expression with the C<\p{}> or C<\P{}> construct.
200 See L</"User-Defined Character Properties"> for more details.
204 The special pattern C<\X> matches a logical character, an "extended grapheme
205 cluster" in Standardese. In Unicode what appears to the user to be a single
206 character, for example an accented C<G>, may in fact be composed of a sequence
207 of characters, in this case a C<G> followed by an accent character. C<\X>
208 will match the entire sequence.
212 The C<tr///> operator translates characters instead of bytes. Note
213 that the C<tr///CU> functionality has been removed. For similar
214 functionality see pack('U0', ...) and pack('C0', ...).
218 Case translation operators use the Unicode case translation tables
219 when character input is provided. Note that C<uc()>, or C<\U> in
220 interpolated strings, translates to uppercase, while C<ucfirst>,
221 or C<\u> in interpolated strings, translates to titlecase in languages
222 that make the distinction (which is equivalent to uppercase in languages
223 without the distinction).
227 Most operators that deal with positions or lengths in a string will
228 automatically switch to using character positions, including
229 C<chop()>, C<chomp()>, C<substr()>, C<pos()>, C<index()>, C<rindex()>,
230 C<sprintf()>, C<write()>, and C<length()>. An operator that
231 specifically does not switch is C<vec()>. Operators that really don't
232 care include operators that treat strings as a bucket of bits such as
233 C<sort()>, and operators dealing with filenames.
237 The C<pack()>/C<unpack()> letter C<C> does I<not> change, since it is often
238 used for byte-oriented formats. Again, think C<char> in the C language.
240 There is a new C<U> specifier that converts between Unicode characters
241 and code points. There is also a C<W> specifier that is the equivalent of
242 C<chr>/C<ord> and properly handles character values even if they are above 255.
246 The C<chr()> and C<ord()> functions work on characters, similar to
247 C<pack("W")> and C<unpack("W")>, I<not> C<pack("C")> and
248 C<unpack("C")>. C<pack("C")> and C<unpack("C")> are methods for
249 emulating byte-oriented C<chr()> and C<ord()> on Unicode strings.
250 While these methods reveal the internal encoding of Unicode strings,
251 that is not something one normally needs to care about at all.
255 The bit string operators, C<& | ^ ~>, can operate on character data.
256 However, for backward compatibility, such as when using bit string
257 operations when characters are all less than 256 in ordinal value, one
258 should not use C<~> (the bit complement) with characters of both
259 values less than 256 and values greater than 256. Most importantly,
260 DeMorgan's laws (C<~($x|$y) eq ~$x&~$y> and C<~($x&$y) eq ~$x|~$y>)
261 will not hold. The reason for this mathematical I<faux pas> is that
262 the complement cannot return B<both> the 8-bit (byte-wide) bit
263 complement B<and> the full character-wide bit complement.
267 You can define your own mappings to be used in C<lc()>,
268 C<lcfirst()>, C<uc()>, and C<ucfirst()> (or their double-quoted string inlined
269 versions such as C<\U>).
270 See L</"User-Defined Case Mappings"> for more details.
278 And finally, C<scalar reverse()> reverses by character rather than by byte.
282 =head2 Unicode Character Properties
284 Most Unicode character properties are accessible by using regular expressions.
285 They are used (like bracketed character classes) by using the C<\p{}> "matches
286 property" construct and the C<\P{}> negation, "doesn't match property".
288 Note that the only time that Perl considers a sequence of individual code
289 points as a single logical character is in the C<\X> construct, already
290 mentioned above. Therefore "character" in this discussion means a single
293 For instance, C<\p{Uppercase}> matches any single character with the Unicode
294 "Uppercase" property, while C<\p{L}> matches any character with a
295 General_Category of "L" (letter) property. Brackets are not
296 required for single letter property names, so C<\p{L}> is equivalent to C<\pL>.
298 More formally, C<\p{Uppercase}> matches any single character whose Unicode
299 Uppercase property value is True, and C<\P{Uppercase}> matches any character
300 whose Uppercase property value is False, and they could have been written as
301 C<\p{Uppercase=True}> and C<\p{Uppercase=False}>, respectively.
303 This formality is needed when properties are not binary, that is if they can
304 take on more values than just True and False. For example, the Bidi_Class (see
305 L</"Bidirectional Character Types"> below), can take on a number of different
306 values, such as Left, Right, Whitespace, and others. To match these, one needs
307 to specify the property name (Bidi_Class), and the value being matched against
308 (Left, Right, etc.). This is done, as in the examples above, by having the
309 two components separated by an equal sign (or interchangeably, a colon), like
310 C<\p{Bidi_Class: Left}>.
312 All Unicode-defined character properties may be written in these compound forms
313 of C<\p{property=value}> or C<\p{property:value}>, but Perl provides some
314 additional properties that are written only in the single form, as well as
315 single-form short-cuts for all binary properties and certain others described
316 below, in which you may omit the property name and the equals or colon
319 Most Unicode character properties have at least two synonyms (or aliases if you
320 prefer), a short one that is easier to type, and a longer one which is more
321 descriptive and hence it is easier to understand what it means. Thus the "L"
322 and "Letter" above are equivalent and can be used interchangeably. Likewise,
323 "Upper" is a synonym for "Uppercase", and we could have written
324 C<\p{Uppercase}> equivalently as C<\p{Upper}>. Also, there are typically
325 various synonyms for the values the property can be. For binary properties,
326 "True" has 3 synonyms: "T", "Yes", and "Y"; and "False has correspondingly "F",
327 "No", and "N". But be careful. A short form of a value for one property may
328 not mean the same thing as the same short form for another. Thus, for the
329 General_Category property, "L" means "Letter", but for the Bidi_Class property,
330 "L" means "Left". A complete list of properties and synonyms is in
333 Upper/lower case differences in the property names and values are irrelevant,
334 thus C<\p{Upper}> means the same thing as C<\p{upper}> or even C<\p{UpPeR}>.
335 Similarly, you can add or subtract underscores anywhere in the middle of a
336 word, so that these are also equivalent to C<\p{U_p_p_e_r}>. And white space
337 is irrelevant adjacent to non-word characters, such as the braces and the equals
338 or colon separators so C<\p{ Upper }> and C<\p{ Upper_case : Y }> are
339 equivalent to these as well. In fact, in most cases, white space and even
340 hyphens can be added or deleted anywhere. So even C<\p{ Up-per case = Yes}> is
341 equivalent. All this is called "loose-matching" by Unicode. The few places
342 where stricter matching is employed is in the middle of numbers, and the Perl
343 extension properties that begin or end with an underscore. Stricter matching
344 cares about white space (except adjacent to the non-word characters) and
345 hyphens, and non-interior underscores.
347 You can also use negation in both C<\p{}> and C<\P{}> by introducing a caret
348 (^) between the first brace and the property name: C<\p{^Tamil}> is
349 equal to C<\P{Tamil}>.
351 =head3 B<General_Category>
353 Every Unicode character is assigned a general category, which is the "most
354 usual categorization of a character" (from
355 L<http://www.unicode.org/reports/tr44>).
357 The compound way of writing these is like C<\p{General_Category=Number}>
358 (short, C<\p{gc:n}>). But Perl furnishes shortcuts in which everything up
359 through the equal or colon separator is omitted. So you can instead just write
362 Here are the short and long forms of the General Category properties:
367 LC, L& Cased_Letter (that is: [\p{Ll}\p{Lu}\p{Lt}])
380 Nd Decimal_Number (also Digit)
384 P Punctuation (also Punct)
385 Pc Connector_Punctuation
389 Pi Initial_Punctuation
390 (may behave like Ps or Pe depending on usage)
392 (may behave like Ps or Pe depending on usage)
404 Zp Paragraph_Separator
407 Cc Control (also Cntrl)
409 Cs Surrogate (not usable)
413 Single-letter properties match all characters in any of the
414 two-letter sub-properties starting with the same letter.
415 C<LC> and C<L&> are special cases, which are both aliases for the set consisting of everything matched by C<Ll>, C<Lu>, and C<Lt>.
417 Because Perl hides the need for the user to understand the internal
418 representation of Unicode characters, there is no need to implement
419 the somewhat messy concept of surrogates. C<Cs> is therefore not
422 =head3 B<Bidirectional Character Types>
424 Because scripts differ in their directionality (Hebrew is
425 written right to left, for example) Unicode supplies these properties in
426 the Bidi_Class class:
431 LRE Left-to-Right Embedding
432 LRO Left-to-Right Override
435 RLE Right-to-Left Embedding
436 RLO Right-to-Left Override
437 PDF Pop Directional Format
439 ES European Separator
440 ET European Terminator
445 B Paragraph Separator
450 This property is always written in the compound form.
451 For example, C<\p{Bidi_Class:R}> matches characters that are normally
452 written right to left.
456 The world's languages are written in a number of scripts. This sentence
457 (unless you're reading it in translation) is written in Latin, while Russian is
458 written in Cyrllic, and Greek is written in, well, Greek; Japanese mainly in
459 Hiragana or Katakana. There are many more.
461 The Unicode Script property gives what script a given character is in,
462 and the property can be specified with the compound form like
463 C<\p{Script=Hebrew}> (short: C<\p{sc=hebr}>). Perl furnishes shortcuts for all
464 script names. You can omit everything up through the equals (or colon), and
465 simply write C<\p{Latin}> or C<\P{Cyrillic}>.
467 A complete list of scripts and their shortcuts is in L<perluniprops>.
469 =head3 B<Use of "Is" Prefix>
471 For backward compatibility (with Perl 5.6), all properties mentioned
472 so far may have C<Is> or C<Is_> prepended to their name, so C<\P{Is_Lu}>, for
473 example, is equal to C<\P{Lu}>, and C<\p{IsScript:Arabic}> is equal to
478 In addition to B<scripts>, Unicode also defines B<blocks> of
479 characters. The difference between scripts and blocks is that the
480 concept of scripts is closer to natural languages, while the concept
481 of blocks is more of an artificial grouping based on groups of Unicode
482 characters with consecutive ordinal values. For example, the "Basic Latin"
483 block is all characters whose ordinals are between 0 and 127, inclusive, in
484 other words, the ASCII characters. The "Latin" script contains some letters
485 from this block as well as several more, like "Latin-1 Supplement",
486 "Latin Extended-A", etc., but it does not contain all the characters from
487 those blocks. It does not, for example, contain digits, because digits are
488 shared across many scripts. Digits and similar groups, like punctuation, are in
489 the script called C<Common>. There is also a script called C<Inherited> for
490 characters that modify other characters, and inherit the script value of the
491 controlling character.
493 For more about scripts versus blocks, see UAX#24 "Unicode Script Property":
494 L<http://www.unicode.org/reports/tr24>
496 The Script property is likely to be the one you want to use when processing
497 natural language; the Block property may be useful in working with the nuts and
500 Block names are matched in the compound form, like C<\p{Block: Arrows}> or
501 C<\p{Blk=Hebrew}>. Unlike most other properties only a few block names have a
502 Unicode-defined short name. But Perl does provide a (slight) shortcut: You
503 can say, for example C<\p{In_Arrows}> or C<\p{In_Hebrew}>. For backwards
504 compatibility, the C<In> prefix may be omitted if there is no naming conflict
505 with a script or any other property, and you can even use an C<Is> prefix
506 instead in those cases. But it is not a good idea to do this, for a couple
513 It is confusing. There are many naming conflicts, and you may forget some.
514 For example, C<\p{Hebrew}> means the I<script> Hebrew, and NOT the I<block>
515 Hebrew. But would you remember that 6 months from now?
519 It is unstable. A new version of Unicode may pre-empt the current meaning by
520 creating a property with the same name. There was a time in very early Unicode
521 releases when C<\p{Hebrew}> would have matched the I<block> Hebrew; now it
526 Some people just prefer to always use C<\p{Block: foo}> and C<\p{Script: bar}>
527 instead of the shortcuts, for clarity, and because they can't remember the
528 difference between 'In' and 'Is' anyway (or aren't confident that those who
529 eventually will read their code will know).
531 A complete list of blocks and their shortcuts is in L<perluniprops>.
533 =head3 B<Other Properties>
535 There are many more properties than the very basic ones described here.
536 A complete list is in L<perluniprops>.
538 Unicode defines all its properties in the compound form, so all single-form
539 properties are Perl extensions. A number of these are just synonyms for the
540 Unicode ones, but some are genunine extensions, including a couple that are in
541 the compound form. And quite a few of these are actually recommended by Unicode
542 (in L<http://www.unicode.org/reports/tr18>).
544 This section gives some details on all the extensions that aren't synonyms for
545 compound-form Unicode properties (for those, you'll have to refer to the
546 L<Unicode Standard|http://www.unicode.org/reports/tr44>.
552 This matches any of the 1_114_112 Unicode code points. It is a synonym for
555 =item B<C<\p{Alnum}>>
557 This matches any C<\p{Alphabetic}> or C<\p{Decimal_Number}> character.
561 This matches any of the 1_114_112 Unicode code points. It is a synonym for
564 =item B<C<\p{Assigned}>>
566 This matches any assigned code point; that is, any code point whose general
567 category is not Unassigned (or equivalently, not Cn).
569 =item B<C<\p{Blank}>>
571 This is the same as C<\h> and C<\p{HorizSpace}>: A character that changes the
572 spacing horizontally.
574 =item B<C<\p{Decomposition_Type: Non_Canonical}>> (Short: C<\p{Dt=NonCanon}>)
576 Matches a character that has a non-canonical decomposition.
578 To understand the use of this rarely used property=value combination, it is
579 necessary to know some basics about decomposition.
580 Consider a character, say H. It could appear with various marks around it,
581 such as an acute accent, or a circumflex, or various hooks, circles, arrows,
582 I<etc.>, above, below, to one side and/or the other, etc. There are many
583 possibilities among the world's languages. The number of combinations is
584 astronomical, and if there were a character for each combination, it would
585 soon exhaust Unicode's more than a million possible characters. So Unicode
586 took a different approach: there is a character for the base H, and a
587 character for each of the possible marks, and they can be combined variously
588 to get a final logical character. So a logical character--what appears to be a
589 single character--can be a sequence of more than one individual characters.
590 This is called an "extended grapheme cluster". (Perl furnishes the C<\X>
591 construct to match such sequences.)
593 But Unicode's intent is to unify the existing character set standards and
594 practices, and a number of pre-existing standards have single characters that
595 mean the same thing as some of these combinations. An example is ISO-8859-1,
596 which has quite a few of these in the Latin-1 range, an example being "LATIN
597 CAPITAL LETTER E WITH ACUTE". Because this character was in this pre-existing
598 standard, Unicode added it to its repertoire. But this character is considered
599 by Unicode to be equivalent to the sequence consisting of first the character
600 "LATIN CAPITAL LETTER E", then the character "COMBINING ACUTE ACCENT".
602 "LATIN CAPITAL LETTER E WITH ACUTE" is called a "pre-composed" character, and
603 the equivalence with the sequence is called canonical equivalence. All
604 pre-composed characters are said to have a decomposition (into the equivalent
605 sequence) and the decomposition type is also called canonical.
607 However, many more characters have a different type of decomposition, a
608 "compatible" or "non-canonical" decomposition. The sequences that form these
609 decompositions are not considered canonically equivalent to the pre-composed
610 character. An example, again in the Latin-1 range, is the "SUPERSCRIPT ONE".
611 It is kind of like a regular digit 1, but not exactly; its decomposition
612 into the digit 1 is called a "compatible" decomposition, specifically a
613 "super" decomposition. There are several such compatibility
614 decompositions (see L<http://www.unicode.org/reports/tr44>), including one
615 called "compat" which means some miscellaneous type of decomposition
616 that doesn't fit into the decomposition categories that Unicode has chosen.
618 Note that most Unicode characters don't have a decomposition, so their
619 decomposition type is "None".
621 Perl has added the C<Non_Canonical> type, for your convenience, to mean any of
622 the compatibility decompositions.
624 =item B<C<\p{Graph}>>
626 Matches any character that is graphic. Theoretically, this means a character
627 that on a printer would cause ink to be used.
629 =item B<C<\p{HorizSpace}>>
631 This is the same as C<\h> and C<\p{Blank}>: A character that changes the
632 spacing horizontally.
636 This is a synonym for C<\p{Present_In=*}>
638 =item B<C<\p{PerlSpace}>>
640 This is the same as C<\s>, restricted to ASCII, namely C<S<[ \f\n\r\t]>>.
642 Mnemonic: Perl's (original) space
644 =item B<C<\p{PerlWord}>>
646 This is the same as C<\w>, restricted to ASCII, namely C<[A-Za-z0-9_]>
648 Mnemonic: Perl's (original) word.
650 =item B<C<\p{PosixAlnum}>>
652 This matches any alphanumeric character in the ASCII range, namely
655 =item B<C<\p{PosixAlpha}>>
657 This matches any alphabetic character in the ASCII range, namely C<[A-Za-z]>.
659 =item B<C<\p{PosixBlank}>>
661 This matches any blank character in the ASCII range, namely C<S<[ \t]>>.
663 =item B<C<\p{PosixCntrl}>>
665 This matches any control character in the ASCII range, namely C<[\x00-\x1F\x7F]>
667 =item B<C<\p{PosixDigit}>>
669 This matches any digit character in the ASCII range, namely C<[0-9]>.
671 =item B<C<\p{PosixGraph}>>
673 This matches any graphical character in the ASCII range, namely C<[\x21-\x7E]>.
675 =item B<C<\p{PosixLower}>>
677 This matches any lowercase character in the ASCII range, namely C<[a-z]>.
679 =item B<C<\p{PosixPrint}>>
681 This matches any printable character in the ASCII range, namely C<[\x20-\x7E]>.
682 These are the graphical characters plus SPACE.
684 =item B<C<\p{PosixPunct}>>
686 This matches any punctuation character in the ASCII range, namely
687 C<[\x21-\x2F\x3A-\x40\x5B-\x60\x7B-\x7E]>. These are the
688 graphical characters that aren't word characters. Note that the Posix standard
689 includes in its definition of punctuation, those characters that Unicode calls
692 =item B<C<\p{PosixSpace}>>
694 This matches any space character in the ASCII range, namely
695 C<S<[ \f\n\r\t\x0B]>> (the last being a vertical tab).
697 =item B<C<\p{PosixUpper}>>
699 This matches any uppercase character in the ASCII range, namely C<[A-Z]>.
701 =item B<C<\p{Present_In: *}>> (Short: C<\p{In=*}>)
703 This property is used when you need to know in what Unicode version(s) a
706 The "*" above stands for some two digit Unicode version number, such as
707 C<1.1> or C<4.0>; or the "*" can also be C<Unassigned>. This property will
708 match the code points whose final disposition has been settled as of the
709 Unicode release given by the version number; C<\p{Present_In: Unassigned}>
710 will match those code points whose meaning has yet to be assigned.
712 For example, C<U+0041> "LATIN CAPITAL LETTER A" was present in the very first
713 Unicode release available, which is C<1.1>, so this property is true for all
714 valid "*" versions. On the other hand, C<U+1EFF> was not assigned until version
715 5.1 when it became "LATIN SMALL LETTER Y WITH LOOP", so the only "*" that
716 would match it are 5.1, 5.2, and later.
718 Unicode furnishes the C<Age> property from which this is derived. The problem
719 with Age is that a strict interpretation of it (which Perl takes) has it
720 matching the precise release a code point's meaning is introduced in. Thus
721 C<U+0041> would match only 1.1; and C<U+1EFF> only 5.1. This is not usually what
724 Some non-Perl implementations of the Age property may change its meaning to be
725 the same as the Perl Present_In property; just be aware of that.
727 Another confusion with both these properties is that the definition is not
728 that the code point has been assigned, but that the meaning of the code point
729 has been determined. This is because 66 code points will always be
730 unassigned, and, so the Age for them is the Unicode version the decision to
731 make them so was made in. For example, C<U+FDD0> is to be permanently
732 unassigned to a character, and the decision to do that was made in version 3.1,
733 so C<\p{Age=3.1}> matches this character and C<\p{Present_In: 3.1}> and up
736 =item B<C<\p{Print}>>
738 This matches any character that is graphical or blank, except controls.
740 =item B<C<\p{SpacePerl}>>
742 This is the same as C<\s>, including beyond ASCII.
744 Mnemonic: Space, as modified by Perl. (It doesn't include the vertical tab
745 which both the Posix standard and Unicode consider to be space.)
747 =item B<C<\p{VertSpace}>>
749 This is the same as C<\v>: A character that changes the spacing vertically.
753 This is the same as C<\w>, including beyond ASCII.
757 =head2 User-Defined Character Properties
759 You can define your own binary character properties by defining subroutines
760 whose names begin with "In" or "Is". The subroutines can be defined in any
761 package. The user-defined properties can be used in the regular expression
762 C<\p> and C<\P> constructs; if you are using a user-defined property from a
763 package other than the one you are in, you must specify its package in the
764 C<\p> or C<\P> construct.
766 # assuming property Is_Foreign defined in Lang::
767 package main; # property package name required
768 if ($txt =~ /\p{Lang::IsForeign}+/) { ... }
770 package Lang; # property package name not required
771 if ($txt =~ /\p{IsForeign}+/) { ... }
774 Note that the effect is compile-time and immutable once defined.
776 The subroutines must return a specially-formatted string, with one
777 or more newline-separated lines. Each line must be one of the following:
783 A single hexadecimal number denoting a Unicode code point to include.
787 Two hexadecimal numbers separated by horizontal whitespace (space or
788 tabular characters) denoting a range of Unicode code points to include.
792 Something to include, prefixed by "+": a built-in character
793 property (prefixed by "utf8::") or a user-defined character property,
794 to represent all the characters in that property; two hexadecimal code
795 points for a range; or a single hexadecimal code point.
799 Something to exclude, prefixed by "-": an existing character
800 property (prefixed by "utf8::") or a user-defined character property,
801 to represent all the characters in that property; two hexadecimal code
802 points for a range; or a single hexadecimal code point.
806 Something to negate, prefixed "!": an existing character
807 property (prefixed by "utf8::") or a user-defined character property,
808 to represent all the characters in that property; two hexadecimal code
809 points for a range; or a single hexadecimal code point.
813 Something to intersect with, prefixed by "&": an existing character
814 property (prefixed by "utf8::") or a user-defined character property,
815 for all the characters except the characters in the property; two
816 hexadecimal code points for a range; or a single hexadecimal code point.
820 For example, to define a property that covers both the Japanese
821 syllabaries (hiragana and katakana), you can define
830 Imagine that the here-doc end marker is at the beginning of the line.
831 Now you can use C<\p{InKana}> and C<\P{InKana}>.
833 You could also have used the existing block property names:
842 Suppose you wanted to match only the allocated characters,
843 not the raw block ranges: in other words, you want to remove
854 The negation is useful for defining (surprise!) negated classes.
864 Intersection is useful for getting the common characters matched by
865 two (or more) classes.
874 It's important to remember not to use "&" for the first set; that
875 would be intersecting with nothing (resulting in an empty set).
877 =head2 User-Defined Case Mappings
879 You can also define your own mappings to be used in the lc(),
880 lcfirst(), uc(), and ucfirst() (or their string-inlined versions).
881 The principle is similar to that of user-defined character
882 properties: to define subroutines
883 with names like C<ToLower> (for lc() and lcfirst()), C<ToTitle> (for
884 the first character in ucfirst()), and C<ToUpper> (for uc(), and the
885 rest of the characters in ucfirst()).
887 The string returned by the subroutines needs to be two hexadecimal numbers
888 separated by two tabulators: the two numbers being, respectively, the source
889 code point and the destination code point. For example:
897 defines an uc() mapping that causes only the character "a"
898 to be mapped to "A"; all other characters will remain unchanged.
900 (For serious hackers only) The above means you have to furnish a complete
901 mapping; you can't just override a couple of characters and leave the rest
902 unchanged. You can find all the mappings in the directory
903 C<$Config{privlib}>/F<unicore/To/>. The mapping data is returned as the
904 here-document, and the C<utf8::ToSpecFoo> are special exception mappings
905 derived from <$Config{privlib}>/F<unicore/SpecialCasing.txt>. The "Digit" and
906 "Fold" mappings that one can see in the directory are not directly
907 user-accessible, one can use either the C<Unicode::UCD> module, or just match
908 case-insensitively (that's when the "Fold" mapping is used).
910 The mappings will only take effect on scalars that have been marked as having
911 Unicode characters, for example by using C<utf8::upgrade()>.
912 Old byte-style strings are not affected.
914 The mappings are in effect for the package they are defined in.
916 =head2 Character Encodings for Input and Output
920 =head2 Unicode Regular Expression Support Level
922 The following list of Unicode support for regular expressions describes
923 all the features currently supported. The references to "Level N"
924 and the section numbers refer to the Unicode Technical Standard #18,
925 "Unicode Regular Expressions", version 11, in May 2005.
931 Level 1 - Basic Unicode Support
933 RL1.1 Hex Notation - done [1]
934 RL1.2 Properties - done [2][3]
935 RL1.2a Compatibility Properties - done [4]
936 RL1.3 Subtraction and Intersection - MISSING [5]
937 RL1.4 Simple Word Boundaries - done [6]
938 RL1.5 Simple Loose Matches - done [7]
939 RL1.6 Line Boundaries - MISSING [8]
940 RL1.7 Supplementary Code Points - done [9]
944 [3] supports not only minimal list, but all Unicode character
945 properties (see L</Unicode Character Properties>)
946 [4] \d \D \s \S \w \W \X [:prop:] [:^prop:]
947 [5] can use regular expression look-ahead [a] or
948 user-defined character properties [b] to emulate set operations
950 [7] note that Perl does Full case-folding in matching (but with bugs),
951 not Simple: for example U+1F88 is equivalent to U+1F00 U+03B9,
952 not with 1F80. This difference matters mainly for certain Greek
953 capital letters with certain modifiers: the Full case-folding
954 decomposes the letter, while the Simple case-folding would map
955 it to a single character.
956 [8] should do ^ and $ also on U+000B (\v in C), FF (\f), CR (\r),
957 CRLF (\r\n), NEL (U+0085), LS (U+2028), and PS (U+2029);
958 should also affect <>, $., and script line numbers;
959 should not split lines within CRLF [c] (i.e. there is no empty
960 line between \r and \n)
961 [9] UTF-8/UTF-EBDDIC used in perl allows not only U+10000 to U+10FFFF
962 but also beyond U+10FFFF [d]
964 [a] You can mimic class subtraction using lookahead.
965 For example, what UTS#18 might write as
967 [{Greek}-[{UNASSIGNED}]]
969 in Perl can be written as:
971 (?!\p{Unassigned})\p{InGreekAndCoptic}
972 (?=\p{Assigned})\p{InGreekAndCoptic}
974 But in this particular example, you probably really want
978 which will match assigned characters known to be part of the Greek script.
980 Also see the Unicode::Regex::Set module, it does implement the full
981 UTS#18 grouping, intersection, union, and removal (subtraction) syntax.
983 [b] '+' for union, '-' for removal (set-difference), '&' for intersection
984 (see L</"User-Defined Character Properties">)
986 [c] Try the C<:crlf> layer (see L<PerlIO>).
988 [d] U+FFFF will currently generate a warning message if 'utf8' warnings are
993 Level 2 - Extended Unicode Support
995 RL2.1 Canonical Equivalents - MISSING [10][11]
996 RL2.2 Default Grapheme Clusters - MISSING [12]
997 RL2.3 Default Word Boundaries - MISSING [14]
998 RL2.4 Default Loose Matches - MISSING [15]
999 RL2.5 Name Properties - MISSING [16]
1000 RL2.6 Wildcard Properties - MISSING
1002 [10] see UAX#15 "Unicode Normalization Forms"
1003 [11] have Unicode::Normalize but not integrated to regexes
1004 [12] have \X but we don't have a "Grapheme Cluster Mode"
1005 [14] see UAX#29, Word Boundaries
1006 [15] see UAX#21 "Case Mappings"
1007 [16] have \N{...} but neither compute names of CJK Ideographs
1008 and Hangul Syllables nor use a loose match [e]
1010 [e] C<\N{...}> allows namespaces (see L<charnames>).
1014 Level 3 - Tailored Support
1016 RL3.1 Tailored Punctuation - MISSING
1017 RL3.2 Tailored Grapheme Clusters - MISSING [17][18]
1018 RL3.3 Tailored Word Boundaries - MISSING
1019 RL3.4 Tailored Loose Matches - MISSING
1020 RL3.5 Tailored Ranges - MISSING
1021 RL3.6 Context Matching - MISSING [19]
1022 RL3.7 Incremental Matches - MISSING
1023 ( RL3.8 Unicode Set Sharing )
1024 RL3.9 Possible Match Sets - MISSING
1025 RL3.10 Folded Matching - MISSING [20]
1026 RL3.11 Submatchers - MISSING
1028 [17] see UAX#10 "Unicode Collation Algorithms"
1029 [18] have Unicode::Collate but not integrated to regexes
1030 [19] have (?<=x) and (?=x), but look-aheads or look-behinds should see
1031 outside of the target substring
1032 [20] need insensitive matching for linguistic features other than case;
1033 for example, hiragana to katakana, wide and narrow, simplified Han
1034 to traditional Han (see UTR#30 "Character Foldings")
1038 =head2 Unicode Encodings
1040 Unicode characters are assigned to I<code points>, which are abstract
1041 numbers. To use these numbers, various encodings are needed.
1049 UTF-8 is a variable-length (1 to 6 bytes, current character allocations
1050 require 4 bytes), byte-order independent encoding. For ASCII (and we
1051 really do mean 7-bit ASCII, not another 8-bit encoding), UTF-8 is
1054 The following table is from Unicode 3.2.
1056 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1058 U+0000..U+007F 00..7F
1059 U+0080..U+07FF * C2..DF 80..BF
1060 U+0800..U+0FFF E0 * A0..BF 80..BF
1061 U+1000..U+CFFF E1..EC 80..BF 80..BF
1062 U+D000..U+D7FF ED 80..9F 80..BF
1063 U+D800..U+DFFF +++++++ utf16 surrogates, not legal utf8 +++++++
1064 U+E000..U+FFFF EE..EF 80..BF 80..BF
1065 U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF
1066 U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
1067 U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
1069 Note the gaps before several of the byte entries above marked by '*'. These are
1070 caused by legal UTF-8 avoiding non-shortest encodings: it is technically
1071 possible to UTF-8-encode a single code point in different ways, but that is
1072 explicitly forbidden, and the shortest possible encoding should always be used
1073 (and that is what Perl does).
1075 Another way to look at it is via bits:
1077 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1080 00000bbbbbaaaaaa 110bbbbb 10aaaaaa
1081 ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
1082 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
1084 As you can see, the continuation bytes all begin with "10", and the
1085 leading bits of the start byte tell how many bytes there are in the
1092 Like UTF-8 but EBCDIC-safe, in the way that UTF-8 is ASCII-safe.
1096 UTF-16, UTF-16BE, UTF-16LE, Surrogates, and BOMs (Byte Order Marks)
1098 The followings items are mostly for reference and general Unicode
1099 knowledge, Perl doesn't use these constructs internally.
1101 UTF-16 is a 2 or 4 byte encoding. The Unicode code points
1102 C<U+0000..U+FFFF> are stored in a single 16-bit unit, and the code
1103 points C<U+10000..U+10FFFF> in two 16-bit units. The latter case is
1104 using I<surrogates>, the first 16-bit unit being the I<high
1105 surrogate>, and the second being the I<low surrogate>.
1107 Surrogates are code points set aside to encode the C<U+10000..U+10FFFF>
1108 range of Unicode code points in pairs of 16-bit units. The I<high
1109 surrogates> are the range C<U+D800..U+DBFF> and the I<low surrogates>
1110 are the range C<U+DC00..U+DFFF>. The surrogate encoding is
1112 $hi = ($uni - 0x10000) / 0x400 + 0xD800;
1113 $lo = ($uni - 0x10000) % 0x400 + 0xDC00;
1117 $uni = 0x10000 + ($hi - 0xD800) * 0x400 + ($lo - 0xDC00);
1119 If you try to generate surrogates (for example by using chr()), you
1120 will get a warning, if warnings are turned on, because those code
1121 points are not valid for a Unicode character.
1123 Because of the 16-bitness, UTF-16 is byte-order dependent. UTF-16
1124 itself can be used for in-memory computations, but if storage or
1125 transfer is required either UTF-16BE (big-endian) or UTF-16LE
1126 (little-endian) encodings must be chosen.
1128 This introduces another problem: what if you just know that your data
1129 is UTF-16, but you don't know which endianness? Byte Order Marks, or
1130 BOMs, are a solution to this. A special character has been reserved
1131 in Unicode to function as a byte order marker: the character with the
1132 code point C<U+FEFF> is the BOM.
1134 The trick is that if you read a BOM, you will know the byte order,
1135 since if it was written on a big-endian platform, you will read the
1136 bytes C<0xFE 0xFF>, but if it was written on a little-endian platform,
1137 you will read the bytes C<0xFF 0xFE>. (And if the originating platform
1138 was writing in UTF-8, you will read the bytes C<0xEF 0xBB 0xBF>.)
1140 The way this trick works is that the character with the code point
1141 C<U+FFFE> is guaranteed not to be a valid Unicode character, so the
1142 sequence of bytes C<0xFF 0xFE> is unambiguously "BOM, represented in
1143 little-endian format" and cannot be C<U+FFFE>, represented in big-endian
1144 format". (Actually, C<U+FFFE> is legal for use by your program, even for
1145 input/output, but better not use it if you need a BOM. But it is "illegal for
1146 interchange", so that an unsuspecting program won't get confused.)
1150 UTF-32, UTF-32BE, UTF-32LE
1152 The UTF-32 family is pretty much like the UTF-16 family, expect that
1153 the units are 32-bit, and therefore the surrogate scheme is not
1154 needed. The BOM signatures will be C<0x00 0x00 0xFE 0xFF> for BE and
1155 C<0xFF 0xFE 0x00 0x00> for LE.
1161 Encodings defined by the ISO 10646 standard. UCS-2 is a 16-bit
1162 encoding. Unlike UTF-16, UCS-2 is not extensible beyond C<U+FFFF>,
1163 because it does not use surrogates. UCS-4 is a 32-bit encoding,
1164 functionally identical to UTF-32.
1170 A seven-bit safe (non-eight-bit) encoding, which is useful if the
1171 transport or storage is not eight-bit safe. Defined by RFC 2152.
1175 =head2 Security Implications of Unicode
1177 Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>.
1178 Also, note the following:
1186 Unfortunately, the specification of UTF-8 leaves some room for
1187 interpretation of how many bytes of encoded output one should generate
1188 from one input Unicode character. Strictly speaking, the shortest
1189 possible sequence of UTF-8 bytes should be generated,
1190 because otherwise there is potential for an input buffer overflow at
1191 the receiving end of a UTF-8 connection. Perl always generates the
1192 shortest length UTF-8, and with warnings on, Perl will warn about
1193 non-shortest length UTF-8 along with other malformations, such as the
1194 surrogates, which are not real Unicode code points.
1198 Regular expressions behave slightly differently between byte data and
1199 character (Unicode) data. For example, the "word character" character
1200 class C<\w> will work differently depending on if data is eight-bit bytes
1203 In the first case, the set of C<\w> characters is either small--the
1204 default set of alphabetic characters, digits, and the "_"--or, if you
1205 are using a locale (see L<perllocale>), the C<\w> might contain a few
1206 more letters according to your language and country.
1208 In the second case, the C<\w> set of characters is much, much larger.
1209 Most importantly, even in the set of the first 256 characters, it will
1210 probably match different characters: unlike most locales, which are
1211 specific to a language and country pair, Unicode classifies all the
1212 characters that are letters I<somewhere> as C<\w>. For example, your
1213 locale might not think that LATIN SMALL LETTER ETH is a letter (unless
1214 you happen to speak Icelandic), but Unicode does.
1216 As discussed elsewhere, Perl has one foot (two hooves?) planted in
1217 each of two worlds: the old world of bytes and the new world of
1218 characters, upgrading from bytes to characters when necessary.
1219 If your legacy code does not explicitly use Unicode, no automatic
1220 switch-over to characters should happen. Characters shouldn't get
1221 downgraded to bytes, either. It is possible to accidentally mix bytes
1222 and characters, however (see L<perluniintro>), in which case C<\w> in
1223 regular expressions might start behaving differently. Review your
1224 code. Use warnings and the C<strict> pragma.
1228 =head2 Unicode in Perl on EBCDIC
1230 The way Unicode is handled on EBCDIC platforms is still
1231 experimental. On such platforms, references to UTF-8 encoding in this
1232 document and elsewhere should be read as meaning the UTF-EBCDIC
1233 specified in Unicode Technical Report 16, unless ASCII vs. EBCDIC issues
1234 are specifically discussed. There is no C<utfebcdic> pragma or
1235 ":utfebcdic" layer; rather, "utf8" and ":utf8" are reused to mean
1236 the platform's "natural" 8-bit encoding of Unicode. See L<perlebcdic>
1237 for more discussion of the issues.
1241 Usually locale settings and Unicode do not affect each other, but
1242 there are a couple of exceptions:
1248 You can enable automatic UTF-8-ification of your standard file
1249 handles, default C<open()> layer, and C<@ARGV> by using either
1250 the C<-C> command line switch or the C<PERL_UNICODE> environment
1251 variable, see L<perlrun> for the documentation of the C<-C> switch.
1255 Perl tries really hard to work both with Unicode and the old
1256 byte-oriented world. Most often this is nice, but sometimes Perl's
1257 straddling of the proverbial fence causes problems.
1261 =head2 When Unicode Does Not Happen
1263 While Perl does have extensive ways to input and output in Unicode,
1264 and few other 'entry points' like the @ARGV which can be interpreted
1265 as Unicode (UTF-8), there still are many places where Unicode (in some
1266 encoding or another) could be given as arguments or received as
1267 results, or both, but it is not.
1269 The following are such interfaces. Also, see L</The "Unicode Bug">.
1270 For all of these interfaces Perl
1271 currently (as of 5.8.3) simply assumes byte strings both as arguments
1272 and results, or UTF-8 strings if the C<encoding> pragma has been used.
1274 One reason why Perl does not attempt to resolve the role of Unicode in
1275 these cases is that the answers are highly dependent on the operating
1276 system and the file system(s). For example, whether filenames can be
1277 in Unicode, and in exactly what kind of encoding, is not exactly a
1278 portable concept. Similarly for the qx and system: how well will the
1279 'command line interface' (and which of them?) handle Unicode?
1285 chdir, chmod, chown, chroot, exec, link, lstat, mkdir,
1286 rename, rmdir, stat, symlink, truncate, unlink, utime, -X
1298 open, opendir, sysopen
1302 qx (aka the backtick operator), system
1310 =head2 The "Unicode Bug"
1312 The term, the "Unicode bug" has been applied to an inconsistency with the
1313 Unicode characters whose ordinals are in the Latin-1 Supplement block, that
1314 is, between 128 and 255. Without a locale specified, unlike all other
1315 characters or code points, these characters have very different semantics in
1316 byte semantics versus character semantics.
1318 In character semantics they are interpreted as Unicode code points, which means
1319 they have the same semantics as Latin-1 (ISO-8859-1).
1321 In byte semantics, they are considered to be unassigned characters, meaning
1322 that the only semantics they have is their ordinal numbers, and that they are
1323 not members of various character classes. None are considered to match C<\w>
1324 for example, but all match C<\W>. (On EBCDIC platforms, the behavior may
1325 be different from this, depending on the underlying C language library
1328 The behavior is known to have effects on these areas:
1334 Changing the case of a scalar, that is, using C<uc()>, C<ucfirst()>, C<lc()>,
1335 and C<lcfirst()>, or C<\L>, C<\U>, C<\u> and C<\l> in regular expression
1340 Using caseless (C</i>) regular expression matching
1344 Matching a number of properties in regular expressions, such as C<\w>
1348 User-defined case change mappings. You can create a C<ToUpper()> function, for
1349 example, which overrides Perl's built-in case mappings. The scalar must be
1350 encoded in utf8 for your function to actually be invoked.
1354 This behavior can lead to unexpected results in which a string's semantics
1355 suddenly change if a code point above 255 is appended to or removed from it,
1356 which changes the string's semantics from byte to character or vice versa. As
1357 an example, consider the following program and its output:
1362 for ($s1, $s2, $s1.$s2) {
1370 If there's no C<\w> in C<s1> or in C<s2>, why does their concatenation have one?
1372 This anomaly stems from Perl's attempt to not disturb older programs that
1373 didn't use Unicode, and hence had no semantics for characters outside of the
1374 ASCII range (except in a locale), along with Perl's desire to add Unicode
1375 support seamlessly. The result wasn't seamless: these characters were
1378 Work is being done to correct this, but only some of it was complete in time
1379 for the 5.12 release. What has been finished is the important part of the case
1380 changing component. Due to concerns, and some evidence, that older code might
1381 have come to rely on the existing behavior, the new behavior must be explicitly
1382 enabled by the feature C<unicode_strings> in the L<feature> pragma, even though
1383 no new syntax is involved.
1385 See L<perlfunc/lc> for details on how this pragma works in combination with
1386 various others for casing. Even though the pragma only affects casing
1387 operations in the 5.12 release, it is planned to have it affect all the
1388 problematic behaviors in later releases: you can't have one without them all.
1390 In the meantime, a workaround is to always call utf8::upgrade($string), or to
1391 use the standard module L<Encode>. Also, a scalar that has any characters
1392 whose ordinal is above 0x100, or which were specified using either of the
1393 C<\N{...}> notations will automatically have character semantics.
1395 =head2 Forcing Unicode in Perl (Or Unforcing Unicode in Perl)
1397 Sometimes (see L</"When Unicode Does Not Happen"> or L</The "Unicode Bug">)
1398 there are situations where you simply need to force a byte
1399 string into UTF-8, or vice versa. The low-level calls
1400 utf8::upgrade($bytestring) and utf8::downgrade($utf8string[, FAIL_OK]) are
1403 Note that utf8::downgrade() can fail if the string contains characters
1404 that don't fit into a byte.
1406 Calling either function on a string that already is in the desired state is a
1409 =head2 Using Unicode in XS
1411 If you want to handle Perl Unicode in XS extensions, you may find the
1412 following C APIs useful. See also L<perlguts/"Unicode Support"> for an
1413 explanation about Unicode at the XS level, and L<perlapi> for the API
1420 C<DO_UTF8(sv)> returns true if the C<UTF8> flag is on and the bytes
1421 pragma is not in effect. C<SvUTF8(sv)> returns true if the C<UTF8>
1422 flag is on; the bytes pragma is ignored. The C<UTF8> flag being on
1423 does B<not> mean that there are any characters of code points greater
1424 than 255 (or 127) in the scalar or that there are even any characters
1425 in the scalar. What the C<UTF8> flag means is that the sequence of
1426 octets in the representation of the scalar is the sequence of UTF-8
1427 encoded code points of the characters of a string. The C<UTF8> flag
1428 being off means that each octet in this representation encodes a
1429 single character with code point 0..255 within the string. Perl's
1430 Unicode model is not to use UTF-8 until it is absolutely necessary.
1434 C<uvchr_to_utf8(buf, chr)> writes a Unicode character code point into
1435 a buffer encoding the code point as UTF-8, and returns a pointer
1436 pointing after the UTF-8 bytes. It works appropriately on EBCDIC machines.
1440 C<utf8_to_uvchr(buf, lenp)> reads UTF-8 encoded bytes from a buffer and
1441 returns the Unicode character code point and, optionally, the length of
1442 the UTF-8 byte sequence. It works appropriately on EBCDIC machines.
1446 C<utf8_length(start, end)> returns the length of the UTF-8 encoded buffer
1447 in characters. C<sv_len_utf8(sv)> returns the length of the UTF-8 encoded
1452 C<sv_utf8_upgrade(sv)> converts the string of the scalar to its UTF-8
1453 encoded form. C<sv_utf8_downgrade(sv)> does the opposite, if
1454 possible. C<sv_utf8_encode(sv)> is like sv_utf8_upgrade except that
1455 it does not set the C<UTF8> flag. C<sv_utf8_decode()> does the
1456 opposite of C<sv_utf8_encode()>. Note that none of these are to be
1457 used as general-purpose encoding or decoding interfaces: C<use Encode>
1458 for that. C<sv_utf8_upgrade()> is affected by the encoding pragma
1459 but C<sv_utf8_downgrade()> is not (since the encoding pragma is
1460 designed to be a one-way street).
1464 C<is_utf8_char(s)> returns true if the pointer points to a valid UTF-8
1469 C<is_utf8_string(buf, len)> returns true if C<len> bytes of the buffer
1474 C<UTF8SKIP(buf)> will return the number of bytes in the UTF-8 encoded
1475 character in the buffer. C<UNISKIP(chr)> will return the number of bytes
1476 required to UTF-8-encode the Unicode character code point. C<UTF8SKIP()>
1477 is useful for example for iterating over the characters of a UTF-8
1478 encoded buffer; C<UNISKIP()> is useful, for example, in computing
1479 the size required for a UTF-8 encoded buffer.
1483 C<utf8_distance(a, b)> will tell the distance in characters between the
1484 two pointers pointing to the same UTF-8 encoded buffer.
1488 C<utf8_hop(s, off)> will return a pointer to a UTF-8 encoded buffer
1489 that is C<off> (positive or negative) Unicode characters displaced
1490 from the UTF-8 buffer C<s>. Be careful not to overstep the buffer:
1491 C<utf8_hop()> will merrily run off the end or the beginning of the
1492 buffer if told to do so.
1496 C<pv_uni_display(dsv, spv, len, pvlim, flags)> and
1497 C<sv_uni_display(dsv, ssv, pvlim, flags)> are useful for debugging the
1498 output of Unicode strings and scalars. By default they are useful
1499 only for debugging--they display B<all> characters as hexadecimal code
1500 points--but with the flags C<UNI_DISPLAY_ISPRINT>,
1501 C<UNI_DISPLAY_BACKSLASH>, and C<UNI_DISPLAY_QQ> you can make the
1502 output more readable.
1506 C<ibcmp_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2)> can be used to
1507 compare two strings case-insensitively in Unicode. For case-sensitive
1508 comparisons you can just use C<memEQ()> and C<memNE()> as usual.
1512 For more information, see L<perlapi>, and F<utf8.c> and F<utf8.h>
1513 in the Perl source code distribution.
1515 =head2 Hacking Perl to work on earlier Unicode versions (for very serious hackers only)
1517 Perl by default comes with the latest supported Unicode version built in, but
1518 you can change to use any earlier one.
1520 Download the files in the version of Unicode that you want from the Unicode web
1521 site L<http://www.unicode.org>). These should replace the existing files in
1522 C<\$Config{privlib}>/F<unicore>. (C<\%Config> is available from the Config
1523 module.) Follow the instructions in F<README.perl> in that directory to change
1524 some of their names, and then run F<make>.
1526 It is even possible to download them to a different directory, and then change
1527 F<utf8_heavy.pl> in the directory C<\$Config{privlib}> to point to the new
1528 directory, or maybe make a copy of that directory before making the change, and
1529 using C<@INC> or the C<-I> run-time flag to switch between versions at will
1530 (but because of caching, not in the middle of a process), but all this is
1531 beyond the scope of these instructions.
1535 =head2 Interaction with Locales
1537 Use of locales with Unicode data may lead to odd results. Currently,
1538 Perl attempts to attach 8-bit locale info to characters in the range
1539 0..255, but this technique is demonstrably incorrect for locales that
1540 use characters above that range when mapped into Unicode. Perl's
1541 Unicode support will also tend to run slower. Use of locales with
1542 Unicode is discouraged.
1544 =head2 Problems with characters in the Latin-1 Supplement range
1546 See L</The "Unicode Bug">
1548 =head2 Problems with case-insensitive regular expression matching
1550 There are problems with case-insensitive matches, including those involving
1551 character classes (enclosed in [square brackets]), characters whose fold
1552 is to multiple characters (such as the single character LATIN SMALL LIGATURE
1553 FFL matches case-insensitively with the 3-character string C<ffl>), and
1554 characters in the Latin-1 Supplement.
1556 =head2 Interaction with Extensions
1558 When Perl exchanges data with an extension, the extension should be
1559 able to understand the UTF8 flag and act accordingly. If the
1560 extension doesn't know about the flag, it's likely that the extension
1561 will return incorrectly-flagged data.
1563 So if you're working with Unicode data, consult the documentation of
1564 every module you're using if there are any issues with Unicode data
1565 exchange. If the documentation does not talk about Unicode at all,
1566 suspect the worst and probably look at the source to learn how the
1567 module is implemented. Modules written completely in Perl shouldn't
1568 cause problems. Modules that directly or indirectly access code written
1569 in other programming languages are at risk.
1571 For affected functions, the simple strategy to avoid data corruption is
1572 to always make the encoding of the exchanged data explicit. Choose an
1573 encoding that you know the extension can handle. Convert arguments passed
1574 to the extensions to that encoding and convert results back from that
1575 encoding. Write wrapper functions that do the conversions for you, so
1576 you can later change the functions when the extension catches up.
1578 To provide an example, let's say the popular Foo::Bar::escape_html
1579 function doesn't deal with Unicode data yet. The wrapper function
1580 would convert the argument to raw UTF-8 and convert the result back to
1581 Perl's internal representation like so:
1583 sub my_escape_html ($) {
1585 return unless defined $what;
1586 Encode::decode_utf8(Foo::Bar::escape_html(Encode::encode_utf8($what)));
1589 Sometimes, when the extension does not convert data but just stores
1590 and retrieves them, you will be in a position to use the otherwise
1591 dangerous Encode::_utf8_on() function. Let's say the popular
1592 C<Foo::Bar> extension, written in C, provides a C<param> method that
1593 lets you store and retrieve data according to these prototypes:
1595 $self->param($name, $value); # set a scalar
1596 $value = $self->param($name); # retrieve a scalar
1598 If it does not yet provide support for any encoding, one could write a
1599 derived class with such a C<param> method:
1602 my($self,$name,$value) = @_;
1603 utf8::upgrade($name); # make sure it is UTF-8 encoded
1604 if (defined $value) {
1605 utf8::upgrade($value); # make sure it is UTF-8 encoded
1606 return $self->SUPER::param($name,$value);
1608 my $ret = $self->SUPER::param($name);
1609 Encode::_utf8_on($ret); # we know, it is UTF-8 encoded
1614 Some extensions provide filters on data entry/exit points, such as
1615 DB_File::filter_store_key and family. Look out for such filters in
1616 the documentation of your extensions, they can make the transition to
1617 Unicode data much easier.
1621 Some functions are slower when working on UTF-8 encoded strings than
1622 on byte encoded strings. All functions that need to hop over
1623 characters such as length(), substr() or index(), or matching regular
1624 expressions can work B<much> faster when the underlying data are
1627 In Perl 5.8.0 the slowness was often quite spectacular; in Perl 5.8.1
1628 a caching scheme was introduced which will hopefully make the slowness
1629 somewhat less spectacular, at least for some operations. In general,
1630 operations with UTF-8 encoded strings are still slower. As an example,
1631 the Unicode properties (character classes) like C<\p{Nd}> are known to
1632 be quite a bit slower (5-20 times) than their simpler counterparts
1633 like C<\d> (then again, there 268 Unicode characters matching C<Nd>
1634 compared with the 10 ASCII characters matching C<d>).
1636 =head2 Problems on EBCDIC platforms
1638 There are a number of known problems with Perl on EBCDIC platforms. If you
1639 want to use Perl there, send email to perlbug@perl.org.
1641 In earlier versions, when byte and character data were concatenated,
1642 the new string was sometimes created by
1643 decoding the byte strings as I<ISO 8859-1 (Latin-1)>, even if the
1644 old Unicode string used EBCDIC.
1646 If you find any of these, please report them as bugs.
1648 =head2 Porting code from perl-5.6.X
1650 Perl 5.8 has a different Unicode model from 5.6. In 5.6 the programmer
1651 was required to use the C<utf8> pragma to declare that a given scope
1652 expected to deal with Unicode data and had to make sure that only
1653 Unicode data were reaching that scope. If you have code that is
1654 working with 5.6, you will need some of the following adjustments to
1655 your code. The examples are written such that the code will continue
1656 to work under 5.6, so you should be safe to try them out.
1662 A filehandle that should read or write UTF-8
1665 binmode $fh, ":encoding(utf8)";
1670 A scalar that is going to be passed to some extension
1672 Be it Compress::Zlib, Apache::Request or any extension that has no
1673 mention of Unicode in the manpage, you need to make sure that the
1674 UTF8 flag is stripped off. Note that at the time of this writing
1675 (October 2002) the mentioned modules are not UTF-8-aware. Please
1676 check the documentation to verify if this is still true.
1680 $val = Encode::encode_utf8($val); # make octets
1685 A scalar we got back from an extension
1687 If you believe the scalar comes back as UTF-8, you will most likely
1688 want the UTF8 flag restored:
1692 $val = Encode::decode_utf8($val);
1697 Same thing, if you are really sure it is UTF-8
1701 Encode::_utf8_on($val);
1706 A wrapper for fetchrow_array and fetchrow_hashref
1708 When the database contains only UTF-8, a wrapper function or method is
1709 a convenient way to replace all your fetchrow_array and
1710 fetchrow_hashref calls. A wrapper function will also make it easier to
1711 adapt to future enhancements in your database driver. Note that at the
1712 time of this writing (October 2002), the DBI has no standardized way
1713 to deal with UTF-8 data. Please check the documentation to verify if
1717 my($self, $sth, $what) = @_; # $what is one of fetchrow_{array,hashref}
1723 my @arr = $sth->$what;
1725 defined && /[^\000-\177]/ && Encode::_utf8_on($_);
1729 my $ret = $sth->$what;
1731 for my $k (keys %$ret) {
1732 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret->{$k};
1736 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret;
1746 A large scalar that you know can only contain ASCII
1748 Scalars that contain only ASCII and are marked as UTF-8 are sometimes
1749 a drag to your program. If you recognize such a situation, just remove
1752 utf8::downgrade($val) if $] > 5.007;
1758 L<perlunitut>, L<perluniintro>, L<perluniprops>, L<Encode>, L<open>, L<utf8>, L<bytes>,
1759 L<perlretut>, L<perlvar/"${^UNICODE}">
1760 L<http://www.unicode.org/reports/tr44>).