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.
15 =item Input and Output Layers
17 Perl knows when a filehandle uses Perl's internal Unicode encodings
18 (UTF-8, or UTF-EBCDIC if in EBCDIC) if the filehandle is opened with
19 the ":utf8" layer. Other encodings can be converted to Perl's
20 encoding on input or from Perl's encoding on output by use of the
21 ":encoding(...)" layer. See L<open>.
23 To indicate that Perl source itself is using a particular encoding,
26 =item Regular Expressions
28 The regular expression compiler produces polymorphic opcodes. That is,
29 the pattern adapts to the data and automatically switches to the Unicode
30 character scheme when presented with Unicode data--or instead uses
31 a traditional byte scheme when presented with byte data.
33 =item C<use utf8> still needed to enable UTF-8/UTF-EBCDIC in scripts
35 As a compatibility measure, the C<use utf8> pragma must be explicitly
36 included to enable recognition of UTF-8 in the Perl scripts themselves
37 (in string or regular expression literals, or in identifier names) on
38 ASCII-based machines or to recognize UTF-EBCDIC on EBCDIC-based
39 machines. B<These are the only times when an explicit C<use utf8>
40 is needed.> See L<utf8>.
42 You can also use the C<encoding> pragma to change the default encoding
43 of the data in your script; see L<encoding>.
45 =item BOM-marked scripts and UTF-16 scripts autodetected
47 If a Perl script begins marked with the Unicode BOM (UTF-16LE, UTF16-BE,
48 or UTF-8), or if the script looks like non-BOM-marked UTF-16 of either
49 endianness, Perl will correctly read in the script as Unicode.
50 (BOMless UTF-8 cannot be effectively recognized or differentiated from
51 ISO 8859-1 or other eight-bit encodings.)
53 =item C<use encoding> needed to upgrade non-Latin-1 byte strings
55 By default, there is a fundamental asymmetry in Perl's unicode model:
56 implicit upgrading from byte strings to Unicode strings assumes that
57 they were encoded in I<ISO 8859-1 (Latin-1)>, but Unicode strings are
58 downgraded with UTF-8 encoding. This happens because the first 256
59 codepoints in Unicode happens to agree with Latin-1.
61 If you wish to interpret byte strings as UTF-8 instead, use the
66 See L</"Byte and Character Semantics"> for more details.
70 =head2 Byte and Character Semantics
72 Beginning with version 5.6, Perl uses logically-wide characters to
73 represent strings internally.
75 In future, Perl-level operations will be expected to work with
76 characters rather than bytes.
78 However, as an interim compatibility measure, Perl aims to
79 provide a safe migration path from byte semantics to character
80 semantics for programs. For operations where Perl can unambiguously
81 decide that the input data are characters, Perl switches to
82 character semantics. For operations where this determination cannot
83 be made without additional information from the user, Perl decides in
84 favor of compatibility and chooses to use byte semantics.
86 This behavior preserves compatibility with earlier versions of Perl,
87 which allowed byte semantics in Perl operations only if
88 none of the program's inputs were marked as being as source of Unicode
89 character data. Such data may come from filehandles, from calls to
90 external programs, from information provided by the system (such as %ENV),
91 or from literals and constants in the source text.
93 The C<bytes> pragma will always, regardless of platform, force byte
94 semantics in a particular lexical scope. See L<bytes>.
96 The C<utf8> pragma is primarily a compatibility device that enables
97 recognition of UTF-(8|EBCDIC) in literals encountered by the parser.
98 Note that this pragma is only required while Perl defaults to byte
99 semantics; when character semantics become the default, this pragma
100 may become a no-op. See L<utf8>.
102 Unless explicitly stated, Perl operators use character semantics
103 for Unicode data and byte semantics for non-Unicode data.
104 The decision to use character semantics is made transparently. If
105 input data comes from a Unicode source--for example, if a character
106 encoding layer is added to a filehandle or a literal Unicode
107 string constant appears in a program--character semantics apply.
108 Otherwise, byte semantics are in effect. The C<bytes> pragma should
109 be used to force byte semantics on Unicode data.
111 If strings operating under byte semantics and strings with Unicode
112 character data are concatenated, the new string will be created by
113 decoding the byte strings as I<ISO 8859-1 (Latin-1)>, even if the
114 old Unicode string used EBCDIC. This translation is done without
115 regard to the system's native 8-bit encoding. To change this for
116 systems with non-Latin-1 and non-EBCDIC native encodings, use the
117 C<encoding> pragma. See L<encoding>.
119 Under character semantics, many operations that formerly operated on
120 bytes now operate on characters. A character in Perl is
121 logically just a number ranging from 0 to 2**31 or so. Larger
122 characters may encode into longer sequences of bytes internally, but
123 this internal detail is mostly hidden for Perl code.
124 See L<perluniintro> for more.
126 =head2 Effects of Character Semantics
128 Character semantics have the following effects:
134 Strings--including hash keys--and regular expression patterns may
135 contain characters that have an ordinal value larger than 255.
137 If you use a Unicode editor to edit your program, Unicode characters
138 may occur directly within the literal strings in one of the various
139 Unicode encodings (UTF-8, UTF-EBCDIC, UCS-2, etc.), but will be recognized
140 as such and converted to Perl's internal representation only if the
141 appropriate L<encoding> is specified.
143 Unicode characters can also be added to a string by using the
144 C<\x{...}> notation. The Unicode code for the desired character, in
145 hexadecimal, should be placed in the braces. For instance, a smiley
146 face is C<\x{263A}>. This encoding scheme only works for characters
147 with a code of 0x100 or above.
151 use charnames ':full';
153 you can use the C<\N{...}> notation and put the official Unicode
154 character name within the braces, such as C<\N{WHITE SMILING FACE}>.
159 If an appropriate L<encoding> is specified, identifiers within the
160 Perl script may contain Unicode alphanumeric characters, including
161 ideographs. Perl does not currently attempt to canonicalize variable
166 Regular expressions match characters instead of bytes. "." matches
167 a character instead of a byte. The C<\C> pattern is provided to force
168 a match a single byte--a C<char> in C, hence C<\C>.
172 Character classes in regular expressions match characters instead of
173 bytes and match against the character properties specified in the
174 Unicode properties database. C<\w> can be used to match a Japanese
175 ideograph, for instance.
177 (However, and as a limitation of the current implementation, using
178 C<\w> or C<\W> I<inside> a C<[...]> character class will still match
179 with byte semantics.)
183 Named Unicode properties, scripts, and block ranges may be used like
184 character classes via the C<\p{}> "matches property" construct and
185 the C<\P{}> negation, "doesn't match property".
187 For instance, C<\p{Lu}> matches any character with the Unicode "Lu"
188 (Letter, uppercase) property, while C<\p{M}> matches any character
189 with an "M" (mark--accents and such) property. Brackets are not
190 required for single letter properties, so C<\p{M}> is equivalent to
191 C<\pM>. Many predefined properties are available, such as
192 C<\p{Mirrored}> and C<\p{Tibetan}>.
194 The official Unicode script and block names have spaces and dashes as
195 separators, but for convenience you can use dashes, spaces, or
196 underbars, and case is unimportant. It is recommended, however, that
197 for consistency you use the following naming: the official Unicode
198 script, property, or block name (see below for the additional rules
199 that apply to block names) with whitespace and dashes removed, and the
200 words "uppercase-first-lowercase-rest". C<Latin-1 Supplement> thus
201 becomes C<Latin1Supplement>.
203 You can also use negation in both C<\p{}> and C<\P{}> by introducing a caret
204 (^) between the first brace and the property name: C<\p{^Tamil}> is
205 equal to C<\P{Tamil}>.
207 B<NOTE: the properties, scripts, and blocks listed here are as of
208 Unicode 3.2.0, March 2002, or Perl 5.8.0, July 2002. Unicode 4.0.0
209 came out in April 2003, and Perl 5.8.1 in September 2003.>
211 Here are the basic Unicode General Category properties, followed by their
212 long form. You can use either; C<\p{Lu}> and C<\p{UppercaseLetter}>,
213 for instance, are identical.
236 Pc ConnectorPunctuation
240 Pi InitialPunctuation
241 (may behave like Ps or Pe depending on usage)
243 (may behave like Ps or Pe depending on usage)
255 Zp ParagraphSeparator
260 Cs Surrogate (not usable)
264 Single-letter properties match all characters in any of the
265 two-letter sub-properties starting with the same letter.
266 C<LC> and C<L&> are special cases, which are aliases for the set of
267 C<Ll>, C<Lu>, and C<Lt>.
269 Because Perl hides the need for the user to understand the internal
270 representation of Unicode characters, there is no need to implement
271 the somewhat messy concept of surrogates. C<Cs> is therefore not
274 Because scripts differ in their directionality--Hebrew is
275 written right to left, for example--Unicode supplies these properties in
281 LRE Left-to-Right Embedding
282 LRO Left-to-Right Override
284 AL Right-to-Left Arabic
285 RLE Right-to-Left Embedding
286 RLO Right-to-Left Override
287 PDF Pop Directional Format
289 ES European Number Separator
290 ET European Number Terminator
292 CS Common Number Separator
295 B Paragraph Separator
300 For example, C<\p{BidiClass:R}> matches characters that are normally
301 written right to left.
307 The script names which can be used by C<\p{...}> and C<\P{...}>,
308 such as in C<\p{Latin}> or C<\p{Cyrillic}>, are as follows:
355 Extended property classes can supplement the basic
356 properties, defined by the F<PropList> Unicode database:
371 LogicalOrderException
372 NoncharacterCodePoint
374 OtherDefaultIgnorableCodePoint
386 and there are further derived properties:
388 Alphabetic Lu + Ll + Lt + Lm + Lo + OtherAlphabetic
389 Lowercase Ll + OtherLowercase
390 Uppercase Lu + OtherUppercase
393 ID_Start Lu + Ll + Lt + Lm + Lo + Nl
394 ID_Continue ID_Start + Mn + Mc + Nd + Pc
397 Assigned Any non-Cn character (i.e. synonym for \P{Cn})
398 Unassigned Synonym for \p{Cn}
399 Common Any character (or unassigned code point)
400 not explicitly assigned to a script
402 For backward compatibility (with Perl 5.6), all properties mentioned
403 so far may have C<Is> prepended to their name, so C<\P{IsLu}>, for
404 example, is equal to C<\P{Lu}>.
408 In addition to B<scripts>, Unicode also defines B<blocks> of
409 characters. The difference between scripts and blocks is that the
410 concept of scripts is closer to natural languages, while the concept
411 of blocks is more of an artificial grouping based on groups of 256
412 Unicode characters. For example, the C<Latin> script contains letters
413 from many blocks but does not contain all the characters from those
414 blocks. It does not, for example, contain digits, because digits are
415 shared across many scripts. Digits and similar groups, like
416 punctuation, are in a category called C<Common>.
418 For more about scripts, see the UTR #24:
420 http://www.unicode.org/unicode/reports/tr24/
422 For more about blocks, see:
424 http://www.unicode.org/Public/UNIDATA/Blocks.txt
426 Block names are given with the C<In> prefix. For example, the
427 Katakana block is referenced via C<\p{InKatakana}>. The C<In>
428 prefix may be omitted if there is no naming conflict with a script
429 or any other property, but it is recommended that C<In> always be used
430 for block tests to avoid confusion.
432 These block names are supported:
434 InAlphabeticPresentationForms
436 InArabicPresentationFormsA
437 InArabicPresentationFormsB
448 InByzantineMusicalSymbols
450 InCJKCompatibilityForms
451 InCJKCompatibilityIdeographs
452 InCJKCompatibilityIdeographsSupplement
453 InCJKRadicalsSupplement
454 InCJKSymbolsAndPunctuation
455 InCJKUnifiedIdeographs
456 InCJKUnifiedIdeographsExtensionA
457 InCJKUnifiedIdeographsExtensionB
459 InCombiningDiacriticalMarks
460 InCombiningDiacriticalMarksforSymbols
465 InCyrillicSupplementary
469 InEnclosedAlphanumerics
470 InEnclosedCJKLettersAndMonths
480 InHalfwidthAndFullwidthForms
481 InHangulCompatibilityJamo
486 InHighPrivateUseSurrogates
490 InIdeographicDescriptionCharacters
495 InKatakanaPhoneticExtensions
500 InLatinExtendedAdditional
505 InMathematicalAlphanumericSymbols
506 InMathematicalOperators
507 InMiscellaneousMathematicalSymbolsA
508 InMiscellaneousMathematicalSymbolsB
509 InMiscellaneousSymbols
510 InMiscellaneousTechnical
517 InOpticalCharacterRecognition
523 InSpacingModifierLetters
525 InSuperscriptsAndSubscripts
526 InSupplementalArrowsA
527 InSupplementalArrowsB
528 InSupplementalMathematicalOperators
529 InSupplementaryPrivateUseAreaA
530 InSupplementaryPrivateUseAreaB
540 InUnifiedCanadianAboriginalSyllabics
549 The special pattern C<\X> matches any extended Unicode
550 sequence--"a combining character sequence" in Standardese--where the
551 first character is a base character and subsequent characters are mark
552 characters that apply to the base character. C<\X> is equivalent to
557 The C<tr///> operator translates characters instead of bytes. Note
558 that the C<tr///CU> functionality has been removed. For similar
559 functionality see pack('U0', ...) and pack('C0', ...).
563 Case translation operators use the Unicode case translation tables
564 when character input is provided. Note that C<uc()>, or C<\U> in
565 interpolated strings, translates to uppercase, while C<ucfirst>,
566 or C<\u> in interpolated strings, translates to titlecase in languages
567 that make the distinction.
571 Most operators that deal with positions or lengths in a string will
572 automatically switch to using character positions, including
573 C<chop()>, C<chomp()>, C<substr()>, C<pos()>, C<index()>, C<rindex()>,
574 C<sprintf()>, C<write()>, and C<length()>. Operators that
575 specifically do not switch include C<vec()>, C<pack()>, and
576 C<unpack()>. Operators that really don't care include
577 operators that treats strings as a bucket of bits such as C<sort()>,
578 and operators dealing with filenames.
582 The C<pack()>/C<unpack()> letters C<c> and C<C> do I<not> change,
583 since they are often used for byte-oriented formats. Again, think
584 C<char> in the C language.
586 There is a new C<U> specifier that converts between Unicode characters
591 The C<chr()> and C<ord()> functions work on characters, similar to
592 C<pack("U")> and C<unpack("U")>, I<not> C<pack("C")> and
593 C<unpack("C")>. C<pack("C")> and C<unpack("C")> are methods for
594 emulating byte-oriented C<chr()> and C<ord()> on Unicode strings.
595 While these methods reveal the internal encoding of Unicode strings,
596 that is not something one normally needs to care about at all.
600 The bit string operators, C<& | ^ ~>, can operate on character data.
601 However, for backward compatibility, such as when using bit string
602 operations when characters are all less than 256 in ordinal value, one
603 should not use C<~> (the bit complement) with characters of both
604 values less than 256 and values greater than 256. Most importantly,
605 DeMorgan's laws (C<~($x|$y) eq ~$x&~$y> and C<~($x&$y) eq ~$x|~$y>)
606 will not hold. The reason for this mathematical I<faux pas> is that
607 the complement cannot return B<both> the 8-bit (byte-wide) bit
608 complement B<and> the full character-wide bit complement.
612 lc(), uc(), lcfirst(), and ucfirst() work for the following cases:
618 the case mapping is from a single Unicode character to another
619 single Unicode character, or
623 the case mapping is from a single Unicode character to more
624 than one Unicode character.
628 Things to do with locales (Lithuanian, Turkish, Azeri) do B<not> work
629 since Perl does not understand the concept of Unicode locales.
631 See the Unicode Technical Report #21, Case Mappings, for more details.
639 And finally, C<scalar reverse()> reverses by character rather than by byte.
643 =head2 User-Defined Character Properties
645 You can define your own character properties by defining subroutines
646 whose names begin with "In" or "Is". The subroutines can be defined in
647 any package. The user-defined properties can be used in the regular
648 expression C<\p> and C<\P> constructs; if you are using a user-defined
649 property from a package other than the one you are in, you must specify
650 its package in the C<\p> or C<\P> construct.
652 # assuming property IsForeign defined in Lang::
653 package main; # property package name required
654 if ($txt =~ /\p{Lang::IsForeign}+/) { ... }
656 package Lang; # property package name not required
657 if ($txt =~ /\p{IsForeign}+/) { ... }
660 Note that the effect is compile-time and immutable once defined.
662 The subroutines must return a specially-formatted string, with one
663 or more newline-separated lines. Each line must be one of the following:
669 Two hexadecimal numbers separated by horizontal whitespace (space or
670 tabular characters) denoting a range of Unicode code points to include.
674 Something to include, prefixed by "+": a built-in character
675 property (prefixed by "utf8::") or a user-defined character property,
676 to represent all the characters in that property; two hexadecimal code
677 points for a range; or a single hexadecimal code point.
681 Something to exclude, prefixed by "-": an existing character
682 property (prefixed by "utf8::") or a user-defined character property,
683 to represent all the characters in that property; two hexadecimal code
684 points for a range; or a single hexadecimal code point.
688 Something to negate, prefixed "!": an existing character
689 property (prefixed by "utf8::") or a user-defined character property,
690 to represent all the characters in that property; two hexadecimal code
691 points for a range; or a single hexadecimal code point.
695 Something to intersect with, prefixed by "&": an existing character
696 property (prefixed by "utf8::") or a user-defined character property,
697 for all the characters except the characters in the property; two
698 hexadecimal code points for a range; or a single hexadecimal code point.
702 For example, to define a property that covers both the Japanese
703 syllabaries (hiragana and katakana), you can define
712 Imagine that the here-doc end marker is at the beginning of the line.
713 Now you can use C<\p{InKana}> and C<\P{InKana}>.
715 You could also have used the existing block property names:
724 Suppose you wanted to match only the allocated characters,
725 not the raw block ranges: in other words, you want to remove
736 The negation is useful for defining (surprise!) negated classes.
746 Intersection is useful for getting the common characters matched by
747 two (or more) classes.
756 It's important to remember not to use "&" for the first set -- that
757 would be intersecting with nothing (resulting in an empty set).
759 You can also define your own mappings to be used in the lc(),
760 lcfirst(), uc(), and ucfirst() (or their string-inlined versions).
761 The principle is the same: define subroutines in the C<main> package
762 with names like C<ToLower> (for lc() and lcfirst()), C<ToTitle> (for
763 the first character in ucfirst()), and C<ToUpper> (for uc(), and the
764 rest of the characters in ucfirst()).
766 The string returned by the subroutines needs now to be three
767 hexadecimal numbers separated by tabulators: start of the source
768 range, end of the source range, and start of the destination range.
777 defines an uc() mapping that causes only the characters "a", "b", and
778 "c" to be mapped to "A", "B", "C", all other characters will remain
781 If there is no source range to speak of, that is, the mapping is from
782 a single character to another single character, leave the end of the
783 source range empty, but the two tabulator characters are still needed.
792 defines a lc() mapping that causes only "A" to be mapped to "a", all
793 other characters will remain unchanged.
795 (For serious hackers only) If you want to introspect the default
796 mappings, you can find the data in the directory
797 C<$Config{privlib}>/F<unicore/To/>. The mapping data is returned as
798 the here-document, and the C<utf8::ToSpecFoo> are special exception
799 mappings derived from <$Config{privlib}>/F<unicore/SpecialCasing.txt>.
800 The C<Digit> and C<Fold> mappings that one can see in the directory
801 are not directly user-accessible, one can use either the
802 C<Unicode::UCD> module, or just match case-insensitively (that's when
803 the C<Fold> mapping is used).
805 A final note on the user-defined property tests and mappings: they
806 will be used only if the scalar has been marked as having Unicode
807 characters. Old byte-style strings will not be affected.
809 =head2 Character Encodings for Input and Output
813 =head2 Unicode Regular Expression Support Level
815 The following list of Unicode support for regular expressions describes
816 all the features currently supported. The references to "Level N"
817 and the section numbers refer to the Unicode Technical Report 18,
818 "Unicode Regular Expression Guidelines", version 6 (Unicode 3.2.0,
825 Level 1 - Basic Unicode Support
827 2.1 Hex Notation - done [1]
828 Named Notation - done [2]
829 2.2 Categories - done [3][4]
830 2.3 Subtraction - MISSING [5][6]
831 2.4 Simple Word Boundaries - done [7]
832 2.5 Simple Loose Matches - done [8]
833 2.6 End of Line - MISSING [9][10]
837 [ 3] . \p{...} \P{...}
838 [ 4] support for scripts (see UTR#24 Script Names), blocks,
839 binary properties, enumerated non-binary properties, and
840 numeric properties (as listed in UTR#18 Other Properties)
842 [ 6] can use regular expression look-ahead [a]
843 or user-defined character properties [b] to emulate subtraction
844 [ 7] include Letters in word characters
845 [ 8] note that Perl does Full case-folding in matching, not Simple:
846 for example U+1F88 is equivalent with U+1F00 U+03B9,
847 not with 1F80. This difference matters for certain Greek
848 capital letters with certain modifiers: the Full case-folding
849 decomposes the letter, while the Simple case-folding would map
850 it to a single character.
851 [ 9] see UTR #13 Unicode Newline Guidelines
852 [10] should do ^ and $ also on \x{85}, \x{2028} and \x{2029}
853 (should also affect <>, $., and script line numbers)
854 (the \x{85}, \x{2028} and \x{2029} do match \s)
856 [a] You can mimic class subtraction using lookahead.
857 For example, what UTR #18 might write as
859 [{Greek}-[{UNASSIGNED}]]
861 in Perl can be written as:
863 (?!\p{Unassigned})\p{InGreekAndCoptic}
864 (?=\p{Assigned})\p{InGreekAndCoptic}
866 But in this particular example, you probably really want
870 which will match assigned characters known to be part of the Greek script.
872 Also see the Unicode::Regex::Set module, it does implement the full
873 UTR #18 grouping, intersection, union, and removal (subtraction) syntax.
875 [b] See L</"User-Defined Character Properties">.
879 Level 2 - Extended Unicode Support
881 3.1 Surrogates - MISSING [11]
882 3.2 Canonical Equivalents - MISSING [12][13]
883 3.3 Locale-Independent Graphemes - MISSING [14]
884 3.4 Locale-Independent Words - MISSING [15]
885 3.5 Locale-Independent Loose Matches - MISSING [16]
887 [11] Surrogates are solely a UTF-16 concept and Perl's internal
888 representation is UTF-8. The Encode module does UTF-16, though.
889 [12] see UTR#15 Unicode Normalization
890 [13] have Unicode::Normalize but not integrated to regexes
891 [14] have \X but at this level . should equal that
892 [15] need three classes, not just \w and \W
893 [16] see UTR#21 Case Mappings
897 Level 3 - Locale-Sensitive Support
899 4.1 Locale-Dependent Categories - MISSING
900 4.2 Locale-Dependent Graphemes - MISSING [16][17]
901 4.3 Locale-Dependent Words - MISSING
902 4.4 Locale-Dependent Loose Matches - MISSING
903 4.5 Locale-Dependent Ranges - MISSING
905 [16] see UTR#10 Unicode Collation Algorithms
906 [17] have Unicode::Collate but not integrated to regexes
910 =head2 Unicode Encodings
912 Unicode characters are assigned to I<code points>, which are abstract
913 numbers. To use these numbers, various encodings are needed.
921 UTF-8 is a variable-length (1 to 6 bytes, current character allocations
922 require 4 bytes), byte-order independent encoding. For ASCII (and we
923 really do mean 7-bit ASCII, not another 8-bit encoding), UTF-8 is
926 The following table is from Unicode 3.2.
928 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
930 U+0000..U+007F 00..7F
931 U+0080..U+07FF C2..DF 80..BF
932 U+0800..U+0FFF E0 A0..BF 80..BF
933 U+1000..U+CFFF E1..EC 80..BF 80..BF
934 U+D000..U+D7FF ED 80..9F 80..BF
935 U+D800..U+DFFF ******* ill-formed *******
936 U+E000..U+FFFF EE..EF 80..BF 80..BF
937 U+10000..U+3FFFF F0 90..BF 80..BF 80..BF
938 U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
939 U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
941 Note the C<A0..BF> in C<U+0800..U+0FFF>, the C<80..9F> in
942 C<U+D000...U+D7FF>, the C<90..B>F in C<U+10000..U+3FFFF>, and the
943 C<80...8F> in C<U+100000..U+10FFFF>. The "gaps" are caused by legal
944 UTF-8 avoiding non-shortest encodings: it is technically possible to
945 UTF-8-encode a single code point in different ways, but that is
946 explicitly forbidden, and the shortest possible encoding should always
947 be used. So that's what Perl does.
949 Another way to look at it is via bits:
951 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
954 00000bbbbbaaaaaa 110bbbbb 10aaaaaa
955 ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
956 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
958 As you can see, the continuation bytes all begin with C<10>, and the
959 leading bits of the start byte tell how many bytes the are in the
966 Like UTF-8 but EBCDIC-safe, in the way that UTF-8 is ASCII-safe.
970 UTF-16, UTF-16BE, UTF-16LE, Surrogates, and BOMs (Byte Order Marks)
972 The followings items are mostly for reference and general Unicode
973 knowledge, Perl doesn't use these constructs internally.
975 UTF-16 is a 2 or 4 byte encoding. The Unicode code points
976 C<U+0000..U+FFFF> are stored in a single 16-bit unit, and the code
977 points C<U+10000..U+10FFFF> in two 16-bit units. The latter case is
978 using I<surrogates>, the first 16-bit unit being the I<high
979 surrogate>, and the second being the I<low surrogate>.
981 Surrogates are code points set aside to encode the C<U+10000..U+10FFFF>
982 range of Unicode code points in pairs of 16-bit units. The I<high
983 surrogates> are the range C<U+D800..U+DBFF>, and the I<low surrogates>
984 are the range C<U+DC00..U+DFFF>. The surrogate encoding is
986 $hi = ($uni - 0x10000) / 0x400 + 0xD800;
987 $lo = ($uni - 0x10000) % 0x400 + 0xDC00;
991 $uni = 0x10000 + ($hi - 0xD800) * 0x400 + ($lo - 0xDC00);
993 If you try to generate surrogates (for example by using chr()), you
994 will get a warning if warnings are turned on, because those code
995 points are not valid for a Unicode character.
997 Because of the 16-bitness, UTF-16 is byte-order dependent. UTF-16
998 itself can be used for in-memory computations, but if storage or
999 transfer is required either UTF-16BE (big-endian) or UTF-16LE
1000 (little-endian) encodings must be chosen.
1002 This introduces another problem: what if you just know that your data
1003 is UTF-16, but you don't know which endianness? Byte Order Marks, or
1004 BOMs, are a solution to this. A special character has been reserved
1005 in Unicode to function as a byte order marker: the character with the
1006 code point C<U+FEFF> is the BOM.
1008 The trick is that if you read a BOM, you will know the byte order,
1009 since if it was written on a big-endian platform, you will read the
1010 bytes C<0xFE 0xFF>, but if it was written on a little-endian platform,
1011 you will read the bytes C<0xFF 0xFE>. (And if the originating platform
1012 was writing in UTF-8, you will read the bytes C<0xEF 0xBB 0xBF>.)
1014 The way this trick works is that the character with the code point
1015 C<U+FFFE> is guaranteed not to be a valid Unicode character, so the
1016 sequence of bytes C<0xFF 0xFE> is unambiguously "BOM, represented in
1017 little-endian format" and cannot be C<U+FFFE>, represented in big-endian
1022 UTF-32, UTF-32BE, UTF-32LE
1024 The UTF-32 family is pretty much like the UTF-16 family, expect that
1025 the units are 32-bit, and therefore the surrogate scheme is not
1026 needed. The BOM signatures will be C<0x00 0x00 0xFE 0xFF> for BE and
1027 C<0xFF 0xFE 0x00 0x00> for LE.
1033 Encodings defined by the ISO 10646 standard. UCS-2 is a 16-bit
1034 encoding. Unlike UTF-16, UCS-2 is not extensible beyond C<U+FFFF>,
1035 because it does not use surrogates. UCS-4 is a 32-bit encoding,
1036 functionally identical to UTF-32.
1042 A seven-bit safe (non-eight-bit) encoding, which is useful if the
1043 transport or storage is not eight-bit safe. Defined by RFC 2152.
1047 =head2 Security Implications of Unicode
1055 Unfortunately, the specification of UTF-8 leaves some room for
1056 interpretation of how many bytes of encoded output one should generate
1057 from one input Unicode character. Strictly speaking, the shortest
1058 possible sequence of UTF-8 bytes should be generated,
1059 because otherwise there is potential for an input buffer overflow at
1060 the receiving end of a UTF-8 connection. Perl always generates the
1061 shortest length UTF-8, and with warnings on Perl will warn about
1062 non-shortest length UTF-8 along with other malformations, such as the
1063 surrogates, which are not real Unicode code points.
1067 Regular expressions behave slightly differently between byte data and
1068 character (Unicode) data. For example, the "word character" character
1069 class C<\w> will work differently depending on if data is eight-bit bytes
1072 In the first case, the set of C<\w> characters is either small--the
1073 default set of alphabetic characters, digits, and the "_"--or, if you
1074 are using a locale (see L<perllocale>), the C<\w> might contain a few
1075 more letters according to your language and country.
1077 In the second case, the C<\w> set of characters is much, much larger.
1078 Most importantly, even in the set of the first 256 characters, it will
1079 probably match different characters: unlike most locales, which are
1080 specific to a language and country pair, Unicode classifies all the
1081 characters that are letters I<somewhere> as C<\w>. For example, your
1082 locale might not think that LATIN SMALL LETTER ETH is a letter (unless
1083 you happen to speak Icelandic), but Unicode does.
1085 As discussed elsewhere, Perl has one foot (two hooves?) planted in
1086 each of two worlds: the old world of bytes and the new world of
1087 characters, upgrading from bytes to characters when necessary.
1088 If your legacy code does not explicitly use Unicode, no automatic
1089 switch-over to characters should happen. Characters shouldn't get
1090 downgraded to bytes, either. It is possible to accidentally mix bytes
1091 and characters, however (see L<perluniintro>), in which case C<\w> in
1092 regular expressions might start behaving differently. Review your
1093 code. Use warnings and the C<strict> pragma.
1097 =head2 Unicode in Perl on EBCDIC
1099 The way Unicode is handled on EBCDIC platforms is still
1100 experimental. On such platforms, references to UTF-8 encoding in this
1101 document and elsewhere should be read as meaning the UTF-EBCDIC
1102 specified in Unicode Technical Report 16, unless ASCII vs. EBCDIC issues
1103 are specifically discussed. There is no C<utfebcdic> pragma or
1104 ":utfebcdic" layer; rather, "utf8" and ":utf8" are reused to mean
1105 the platform's "natural" 8-bit encoding of Unicode. See L<perlebcdic>
1106 for more discussion of the issues.
1110 Usually locale settings and Unicode do not affect each other, but
1111 there are a couple of exceptions:
1117 You can enable automatic UTF-8-ification of your standard file
1118 handles, default C<open()> layer, and C<@ARGV> by using either
1119 the C<-C> command line switch or the C<PERL_UNICODE> environment
1120 variable, see L<perlrun> for the documentation of the C<-C> switch.
1124 Perl tries really hard to work both with Unicode and the old
1125 byte-oriented world. Most often this is nice, but sometimes Perl's
1126 straddling of the proverbial fence causes problems.
1130 =head2 When Unicode Does Not Happen
1132 While Perl does have extensive ways to input and output in Unicode,
1133 and few other 'entry points' like the @ARGV which can be interpreted
1134 as Unicode (UTF-8), there still are many places where Unicode (in some
1135 encoding or another) could be given as arguments or received as
1136 results, or both, but it is not.
1138 The following are such interfaces. For all of these interfaces Perl
1139 currently (as of 5.8.3) simply assumes byte strings both as arguments
1140 and results, or UTF-8 strings if the C<encoding> pragma has been used.
1142 One reason why Perl does not attempt to resolve the role of Unicode in
1143 this cases is that the answers are highly dependent on the operating
1144 system and the file system(s). For example, whether filenames can be
1145 in Unicode, and in exactly what kind of encoding, is not exactly a
1146 portable concept. Similarly for the qx and system: how well will the
1147 'command line interface' (and which of them?) handle Unicode?
1153 chmod, chmod, chown, chroot, exec, link, lstat, mkdir,
1154 rename, rmdir, stat, symlink, truncate, unlink, utime, -X
1166 open, opendir, sysopen
1170 qx (aka the backtick operator), system
1178 =head2 Forcing Unicode in Perl (Or Unforcing Unicode in Perl)
1180 Sometimes (see L</"When Unicode Does Not Happen">) there are
1181 situations where you simply need to force Perl to believe that a byte
1182 string is UTF-8, or vice versa. The low-level calls
1183 utf8::upgrade($bytestring) and utf8::downgrade($utf8string) are
1186 Do not use them without careful thought, though: Perl may easily get
1187 very confused, angry, or even crash, if you suddenly change the 'nature'
1188 of scalar like that. Especially careful you have to be if you use the
1189 utf8::upgrade(): any random byte string is not valid UTF-8.
1191 =head2 Using Unicode in XS
1193 If you want to handle Perl Unicode in XS extensions, you may find the
1194 following C APIs useful. See also L<perlguts/"Unicode Support"> for an
1195 explanation about Unicode at the XS level, and L<perlapi> for the API
1202 C<DO_UTF8(sv)> returns true if the C<UTF8> flag is on and the bytes
1203 pragma is not in effect. C<SvUTF8(sv)> returns true is the C<UTF8>
1204 flag is on; the bytes pragma is ignored. The C<UTF8> flag being on
1205 does B<not> mean that there are any characters of code points greater
1206 than 255 (or 127) in the scalar or that there are even any characters
1207 in the scalar. What the C<UTF8> flag means is that the sequence of
1208 octets in the representation of the scalar is the sequence of UTF-8
1209 encoded code points of the characters of a string. The C<UTF8> flag
1210 being off means that each octet in this representation encodes a
1211 single character with code point 0..255 within the string. Perl's
1212 Unicode model is not to use UTF-8 until it is absolutely necessary.
1216 C<uvuni_to_utf8(buf, chr)> writes a Unicode character code point into
1217 a buffer encoding the code point as UTF-8, and returns a pointer
1218 pointing after the UTF-8 bytes.
1222 C<utf8_to_uvuni(buf, lenp)> reads UTF-8 encoded bytes from a buffer and
1223 returns the Unicode character code point and, optionally, the length of
1224 the UTF-8 byte sequence.
1228 C<utf8_length(start, end)> returns the length of the UTF-8 encoded buffer
1229 in characters. C<sv_len_utf8(sv)> returns the length of the UTF-8 encoded
1234 C<sv_utf8_upgrade(sv)> converts the string of the scalar to its UTF-8
1235 encoded form. C<sv_utf8_downgrade(sv)> does the opposite, if
1236 possible. C<sv_utf8_encode(sv)> is like sv_utf8_upgrade except that
1237 it does not set the C<UTF8> flag. C<sv_utf8_decode()> does the
1238 opposite of C<sv_utf8_encode()>. Note that none of these are to be
1239 used as general-purpose encoding or decoding interfaces: C<use Encode>
1240 for that. C<sv_utf8_upgrade()> is affected by the encoding pragma
1241 but C<sv_utf8_downgrade()> is not (since the encoding pragma is
1242 designed to be a one-way street).
1246 C<is_utf8_char(s)> returns true if the pointer points to a valid UTF-8
1251 C<is_utf8_string(buf, len)> returns true if C<len> bytes of the buffer
1256 C<UTF8SKIP(buf)> will return the number of bytes in the UTF-8 encoded
1257 character in the buffer. C<UNISKIP(chr)> will return the number of bytes
1258 required to UTF-8-encode the Unicode character code point. C<UTF8SKIP()>
1259 is useful for example for iterating over the characters of a UTF-8
1260 encoded buffer; C<UNISKIP()> is useful, for example, in computing
1261 the size required for a UTF-8 encoded buffer.
1265 C<utf8_distance(a, b)> will tell the distance in characters between the
1266 two pointers pointing to the same UTF-8 encoded buffer.
1270 C<utf8_hop(s, off)> will return a pointer to an UTF-8 encoded buffer
1271 that is C<off> (positive or negative) Unicode characters displaced
1272 from the UTF-8 buffer C<s>. Be careful not to overstep the buffer:
1273 C<utf8_hop()> will merrily run off the end or the beginning of the
1274 buffer if told to do so.
1278 C<pv_uni_display(dsv, spv, len, pvlim, flags)> and
1279 C<sv_uni_display(dsv, ssv, pvlim, flags)> are useful for debugging the
1280 output of Unicode strings and scalars. By default they are useful
1281 only for debugging--they display B<all> characters as hexadecimal code
1282 points--but with the flags C<UNI_DISPLAY_ISPRINT>,
1283 C<UNI_DISPLAY_BACKSLASH>, and C<UNI_DISPLAY_QQ> you can make the
1284 output more readable.
1288 C<ibcmp_utf8(s1, pe1, u1, l1, u1, s2, pe2, l2, u2)> can be used to
1289 compare two strings case-insensitively in Unicode. For case-sensitive
1290 comparisons you can just use C<memEQ()> and C<memNE()> as usual.
1294 For more information, see L<perlapi>, and F<utf8.c> and F<utf8.h>
1295 in the Perl source code distribution.
1299 =head2 Interaction with Locales
1301 Use of locales with Unicode data may lead to odd results. Currently,
1302 Perl attempts to attach 8-bit locale info to characters in the range
1303 0..255, but this technique is demonstrably incorrect for locales that
1304 use characters above that range when mapped into Unicode. Perl's
1305 Unicode support will also tend to run slower. Use of locales with
1306 Unicode is discouraged.
1308 =head2 Interaction with Extensions
1310 When Perl exchanges data with an extension, the extension should be
1311 able to understand the UTF-8 flag and act accordingly. If the
1312 extension doesn't know about the flag, it's likely that the extension
1313 will return incorrectly-flagged data.
1315 So if you're working with Unicode data, consult the documentation of
1316 every module you're using if there are any issues with Unicode data
1317 exchange. If the documentation does not talk about Unicode at all,
1318 suspect the worst and probably look at the source to learn how the
1319 module is implemented. Modules written completely in Perl shouldn't
1320 cause problems. Modules that directly or indirectly access code written
1321 in other programming languages are at risk.
1323 For affected functions, the simple strategy to avoid data corruption is
1324 to always make the encoding of the exchanged data explicit. Choose an
1325 encoding that you know the extension can handle. Convert arguments passed
1326 to the extensions to that encoding and convert results back from that
1327 encoding. Write wrapper functions that do the conversions for you, so
1328 you can later change the functions when the extension catches up.
1330 To provide an example, let's say the popular Foo::Bar::escape_html
1331 function doesn't deal with Unicode data yet. The wrapper function
1332 would convert the argument to raw UTF-8 and convert the result back to
1333 Perl's internal representation like so:
1335 sub my_escape_html ($) {
1337 return unless defined $what;
1338 Encode::decode_utf8(Foo::Bar::escape_html(Encode::encode_utf8($what)));
1341 Sometimes, when the extension does not convert data but just stores
1342 and retrieves them, you will be in a position to use the otherwise
1343 dangerous Encode::_utf8_on() function. Let's say the popular
1344 C<Foo::Bar> extension, written in C, provides a C<param> method that
1345 lets you store and retrieve data according to these prototypes:
1347 $self->param($name, $value); # set a scalar
1348 $value = $self->param($name); # retrieve a scalar
1350 If it does not yet provide support for any encoding, one could write a
1351 derived class with such a C<param> method:
1354 my($self,$name,$value) = @_;
1355 utf8::upgrade($name); # make sure it is UTF-8 encoded
1357 utf8::upgrade($value); # make sure it is UTF-8 encoded
1358 return $self->SUPER::param($name,$value);
1360 my $ret = $self->SUPER::param($name);
1361 Encode::_utf8_on($ret); # we know, it is UTF-8 encoded
1366 Some extensions provide filters on data entry/exit points, such as
1367 DB_File::filter_store_key and family. Look out for such filters in
1368 the documentation of your extensions, they can make the transition to
1369 Unicode data much easier.
1373 Some functions are slower when working on UTF-8 encoded strings than
1374 on byte encoded strings. All functions that need to hop over
1375 characters such as length(), substr() or index(), or matching regular
1376 expressions can work B<much> faster when the underlying data are
1379 In Perl 5.8.0 the slowness was often quite spectacular; in Perl 5.8.1
1380 a caching scheme was introduced which will hopefully make the slowness
1381 somewhat less spectacular, at least for some operations. In general,
1382 operations with UTF-8 encoded strings are still slower. As an example,
1383 the Unicode properties (character classes) like C<\p{Nd}> are known to
1384 be quite a bit slower (5-20 times) than their simpler counterparts
1385 like C<\d> (then again, there 268 Unicode characters matching C<Nd>
1386 compared with the 10 ASCII characters matching C<d>).
1388 =head2 Porting code from perl-5.6.X
1390 Perl 5.8 has a different Unicode model from 5.6. In 5.6 the programmer
1391 was required to use the C<utf8> pragma to declare that a given scope
1392 expected to deal with Unicode data and had to make sure that only
1393 Unicode data were reaching that scope. If you have code that is
1394 working with 5.6, you will need some of the following adjustments to
1395 your code. The examples are written such that the code will continue
1396 to work under 5.6, so you should be safe to try them out.
1402 A filehandle that should read or write UTF-8
1405 binmode $fh, ":utf8";
1410 A scalar that is going to be passed to some extension
1412 Be it Compress::Zlib, Apache::Request or any extension that has no
1413 mention of Unicode in the manpage, you need to make sure that the
1414 UTF-8 flag is stripped off. Note that at the time of this writing
1415 (October 2002) the mentioned modules are not UTF-8-aware. Please
1416 check the documentation to verify if this is still true.
1420 $val = Encode::encode_utf8($val); # make octets
1425 A scalar we got back from an extension
1427 If you believe the scalar comes back as UTF-8, you will most likely
1428 want the UTF-8 flag restored:
1432 $val = Encode::decode_utf8($val);
1437 Same thing, if you are really sure it is UTF-8
1441 Encode::_utf8_on($val);
1446 A wrapper for fetchrow_array and fetchrow_hashref
1448 When the database contains only UTF-8, a wrapper function or method is
1449 a convenient way to replace all your fetchrow_array and
1450 fetchrow_hashref calls. A wrapper function will also make it easier to
1451 adapt to future enhancements in your database driver. Note that at the
1452 time of this writing (October 2002), the DBI has no standardized way
1453 to deal with UTF-8 data. Please check the documentation to verify if
1457 my($self, $sth, $what) = @_; # $what is one of fetchrow_{array,hashref}
1463 my @arr = $sth->$what;
1465 defined && /[^\000-\177]/ && Encode::_utf8_on($_);
1469 my $ret = $sth->$what;
1471 for my $k (keys %$ret) {
1472 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret->{$k};
1476 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret;
1486 A large scalar that you know can only contain ASCII
1488 Scalars that contain only ASCII and are marked as UTF-8 are sometimes
1489 a drag to your program. If you recognize such a situation, just remove
1492 utf8::downgrade($val) if $] > 5.007;
1498 L<perluniintro>, L<encoding>, L<Encode>, L<open>, L<utf8>, L<bytes>,
1499 L<perlretut>, L<perlvar/"${^UNICODE}">
https://perl5.git.perl.org / perl5.git / blob