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 L<the Perl Unicode tutorial, perlunitut|perlunitut>, before reading
15 this reference document.
19 =item Input and Output Layers
21 Perl knows when a filehandle uses Perl's internal Unicode encodings
22 (UTF-8, or UTF-EBCDIC if in EBCDIC) if the filehandle is opened with
23 the ":utf8" layer. Other encodings can be converted to Perl's
24 encoding on input or from Perl's encoding on output by use of the
25 ":encoding(...)" layer. See L<open>.
27 To indicate that Perl source itself is in UTF-8, use C<use utf8;>.
29 =item Regular Expressions
31 The regular expression compiler produces polymorphic opcodes. That is,
32 the pattern adapts to the data and automatically switches to the Unicode
33 character scheme when presented with data that is internally encoded in
34 UTF-8, or instead uses a traditional byte scheme when presented with
37 =item C<use utf8> still needed to enable UTF-8/UTF-EBCDIC in scripts
39 As a compatibility measure, the C<use utf8> pragma must be explicitly
40 included to enable recognition of UTF-8 in the Perl scripts themselves
41 (in string or regular expression literals, or in identifier names) on
42 ASCII-based machines or to recognize UTF-EBCDIC on EBCDIC-based
43 machines. B<These are the only times when an explicit C<use utf8>
44 is needed.> See L<utf8>.
46 =item BOM-marked scripts and UTF-16 scripts autodetected
48 If a Perl script begins marked with the Unicode BOM (UTF-16LE, UTF16-BE,
49 or UTF-8), or if the script looks like non-BOM-marked UTF-16 of either
50 endianness, Perl will correctly read in the script as Unicode.
51 (BOMless UTF-8 cannot be effectively recognized or differentiated from
52 ISO 8859-1 or other eight-bit encodings.)
54 =item C<use encoding> needed to upgrade non-Latin-1 byte strings
56 By default, there is a fundamental asymmetry in Perl's Unicode model:
57 implicit upgrading from byte strings to Unicode strings assumes that
58 they were encoded in I<ISO 8859-1 (Latin-1)>, but Unicode strings are
59 downgraded with UTF-8 encoding. This happens because the first 256
60 codepoints in Unicode happens to agree with Latin-1.
62 See L</"Byte and Character Semantics"> for more details.
66 =head2 Byte and Character Semantics
68 Beginning with version 5.6, Perl uses logically-wide characters to
69 represent strings internally.
71 In future, Perl-level operations will be expected to work with
72 characters rather than bytes.
74 However, as an interim compatibility measure, Perl aims to
75 provide a safe migration path from byte semantics to character
76 semantics for programs. For operations where Perl can unambiguously
77 decide that the input data are characters, Perl switches to
78 character semantics. For operations where this determination cannot
79 be made without additional information from the user, Perl decides in
80 favor of compatibility and chooses to use byte semantics.
82 Under byte semantics, when C<use locale> is in effect, Perl uses the
83 semantics associated with the current locale. Absent a C<use locale>, and
84 absent a C<use feature 'unicode_strings'> pragma, Perl currently uses US-ASCII
85 (or Basic Latin in Unicode terminology) byte semantics, meaning that characters
86 whose ordinal numbers are in the range 128 - 255 are undefined except for their
87 ordinal numbers. This means that none have case (upper and lower), nor are any
88 a member of character classes, like C<[:alpha:]> or C<\w>. (But all do belong
89 to the C<\W> class or the Perl regular expression extension C<[:^alpha:]>.)
91 This behavior preserves compatibility with earlier versions of Perl,
92 which allowed byte semantics in Perl operations only if
93 none of the program's inputs were marked as being a source of Unicode
94 character data. Such data may come from filehandles, from calls to
95 external programs, from information provided by the system (such as %ENV),
96 or from literals and constants in the source text.
98 The C<bytes> pragma will always, regardless of platform, force byte
99 semantics in a particular lexical scope. See L<bytes>.
101 The C<use feature 'unicode_strings'> pragma is intended to always, regardless
102 of platform, force Unicode semantics in a particular lexical scope. In
103 release 5.12, it is partially implemented, applying only to case changes.
104 See L</The "Unicode Bug"> below.
106 The C<utf8> pragma is primarily a compatibility device that enables
107 recognition of UTF-(8|EBCDIC) in literals encountered by the parser.
108 Note that this pragma is only required while Perl defaults to byte
109 semantics; when character semantics become the default, this pragma
110 may become a no-op. See L<utf8>.
112 Unless explicitly stated, Perl operators use character semantics
113 for Unicode data and byte semantics for non-Unicode data.
114 The decision to use character semantics is made transparently. If
115 input data comes from a Unicode source--for example, if a character
116 encoding layer is added to a filehandle or a literal Unicode
117 string constant appears in a program--character semantics apply.
118 Otherwise, byte semantics are in effect. The C<bytes> pragma should
119 be used to force byte semantics on Unicode data, and the C<use feature
120 'unicode_strings'> pragma to force Unicode semantics on byte data (though in
121 5.12 it isn't fully implemented).
123 If strings operating under byte semantics and strings with Unicode
124 character data are concatenated, the new string will have
125 character semantics. This can cause surprises: See L</BUGS>, below.
126 You can choose to be warned when this happens. See L<encoding::warnings>.
128 Under character semantics, many operations that formerly operated on
129 bytes now operate on characters. A character in Perl is
130 logically just a number ranging from 0 to 2**31 or so. Larger
131 characters may encode into longer sequences of bytes internally, but
132 this internal detail is mostly hidden for Perl code.
133 See L<perluniintro> for more.
135 =head2 Effects of Character Semantics
137 Character semantics have the following effects:
143 Strings--including hash keys--and regular expression patterns may
144 contain characters that have an ordinal value larger than 255.
146 If you use a Unicode editor to edit your program, Unicode characters may
147 occur directly within the literal strings in UTF-8 encoding, or UTF-16.
148 (The former requires a BOM or C<use utf8>, the latter requires a BOM.)
150 Unicode characters can also be added to a string by using the C<\N{U+...}>
151 notation. The Unicode code for the desired character, in hexadecimal,
152 should be placed in the braces, after the C<U>. For instance, a smiley face is
155 Alternatively, you can use the C<\x{...}> notation for characters 0x100 and
156 above. For characters below 0x100 you may get byte semantics instead of
157 character semantics; see L</The "Unicode Bug">. On EBCDIC machines there is
158 the additional problem that the value for such characters gives the EBCDIC
159 character rather than the Unicode one.
163 use charnames ':full';
165 you can use the C<\N{...}> notation and put the official Unicode
166 character name within the braces, such as C<\N{WHITE SMILING FACE}>.
171 If an appropriate L<encoding> is specified, identifiers within the
172 Perl script may contain Unicode alphanumeric characters, including
173 ideographs. Perl does not currently attempt to canonicalize variable
178 Regular expressions match characters instead of bytes. "." matches
179 a character instead of a byte.
183 Character classes in regular expressions match characters instead of
184 bytes and match against the character properties specified in the
185 Unicode properties database. C<\w> can be used to match a Japanese
186 ideograph, for instance.
190 Named Unicode properties, scripts, and block ranges may be used like
191 character classes via the C<\p{}> "matches property" construct and
192 the C<\P{}> negation, "doesn't match property".
193 See L</"Unicode Character Properties"> for more details.
195 You can define your own character properties and use them
196 in the regular expression with the C<\p{}> or C<\P{}> construct.
197 See L</"User-Defined Character Properties"> for more details.
201 The special pattern C<\X> matches a logical character, an "extended grapheme
202 cluster" in Standardese. In Unicode what appears to the user to be a single
203 character, for example an accented C<G>, may in fact be composed of a sequence
204 of characters, in this case a C<G> followed by an accent character. C<\X>
205 will match the entire sequence.
209 The C<tr///> operator translates characters instead of bytes. Note
210 that the C<tr///CU> functionality has been removed. For similar
211 functionality see pack('U0', ...) and pack('C0', ...).
215 Case translation operators use the Unicode case translation tables
216 when character input is provided. Note that C<uc()>, or C<\U> in
217 interpolated strings, translates to uppercase, while C<ucfirst>,
218 or C<\u> in interpolated strings, translates to titlecase in languages
219 that make the distinction (which is equivalent to uppercase in languages
220 without the distinction).
224 Most operators that deal with positions or lengths in a string will
225 automatically switch to using character positions, including
226 C<chop()>, C<chomp()>, C<substr()>, C<pos()>, C<index()>, C<rindex()>,
227 C<sprintf()>, C<write()>, and C<length()>. An operator that
228 specifically does not switch is C<vec()>. Operators that really don't
229 care include operators that treat strings as a bucket of bits such as
230 C<sort()>, and operators dealing with filenames.
234 The C<pack()>/C<unpack()> letter C<C> does I<not> change, since it is often
235 used for byte-oriented formats. Again, think C<char> in the C language.
237 There is a new C<U> specifier that converts between Unicode characters
238 and code points. There is also a C<W> specifier that is the equivalent of
239 C<chr>/C<ord> and properly handles character values even if they are above 255.
243 The C<chr()> and C<ord()> functions work on characters, similar to
244 C<pack("W")> and C<unpack("W")>, I<not> C<pack("C")> and
245 C<unpack("C")>. C<pack("C")> and C<unpack("C")> are methods for
246 emulating byte-oriented C<chr()> and C<ord()> on Unicode strings.
247 While these methods reveal the internal encoding of Unicode strings,
248 that is not something one normally needs to care about at all.
252 The bit string operators, C<& | ^ ~>, can operate on character data.
253 However, for backward compatibility, such as when using bit string
254 operations when characters are all less than 256 in ordinal value, one
255 should not use C<~> (the bit complement) with characters of both
256 values less than 256 and values greater than 256. Most importantly,
257 DeMorgan's laws (C<~($x|$y) eq ~$x&~$y> and C<~($x&$y) eq ~$x|~$y>)
258 will not hold. The reason for this mathematical I<faux pas> is that
259 the complement cannot return B<both> the 8-bit (byte-wide) bit
260 complement B<and> the full character-wide bit complement.
264 You can define your own mappings to be used in lc(),
265 lcfirst(), uc(), and ucfirst() (or their string-inlined versions).
266 See L</"User-Defined Case Mappings"> for more details.
274 And finally, C<scalar reverse()> reverses by character rather than by byte.
278 =head2 Unicode Character Properties
280 Most Unicode character properties are accessible by using regular expressions.
281 They are used like character classes via the C<\p{}> "matches property"
282 construct and the C<\P{}> negation, "doesn't match property".
284 For instance, C<\p{Uppercase}> matches any character with the Unicode
285 "Uppercase" property, while C<\p{L}> matches any character with a
286 General_Category of "L" (letter) property. Brackets are not
287 required for single letter properties, so C<\p{L}> is equivalent to C<\pL>.
289 More formally, C<\p{Uppercase}> matches any character whose Unicode Uppercase
290 property value is True, and C<\P{Uppercase}> matches any character whose
291 Uppercase property value is False, and they could have been written as
292 C<\p{Uppercase=True}> and C<\p{Uppercase=False}>, respectively
294 This formality is needed when properties are not binary, that is if they can
295 take on more values than just True and False. For example, the Bidi_Class (see
296 L</"Bidirectional Character Types"> below), can take on a number of different
297 values, such as Left, Right, Whitespace, and others. To match these, one needs
298 to specify the property name (Bidi_Class), and the value being matched against
299 (Left, Right, I<etc.>). This is done, as in the examples above, by having the
300 two components separated by an equal sign (or interchangeably, a colon), like
301 C<\p{Bidi_Class: Left}>.
303 All Unicode-defined character properties may be written in these compound forms
304 of C<\p{property=value}> or C<\p{property:value}>, but Perl provides some
305 additional properties that are written only in the single form, as well as
306 single-form short-cuts for all binary properties and certain others described
307 below, in which you may omit the property name and the equals or colon
310 Most Unicode character properties have at least two synonyms (or aliases if you
311 prefer), a short one that is easier to type, and a longer one which is more
312 descriptive and hence it is easier to understand what it means. Thus the "L"
313 and "Letter" above are equivalent and can be used interchangeably. Likewise,
314 "Upper" is a synonym for "Uppercase", and we could have written
315 C<\p{Uppercase}> equivalently as C<\p{Upper}>. Also, there are typically
316 various synonyms for the values the property can be. For binary properties,
317 "True" has 3 synonyms: "T", "Yes", and "Y"; and "False has correspondingly "F",
318 "No", and "N". But be careful. A short form of a value for one property may
319 not mean the same thing as the same short form for another. Thus, for the
320 General_Category property, "L" means "Letter", but for the Bidi_Class property,
321 "L" means "Left". A complete list of properties and synonyms is in
324 Upper/lower case differences in the property names and values are irrelevant,
325 thus C<\p{Upper}> means the same thing as C<\p{upper}> or even C<\p{UpPeR}>.
326 Similarly, you can add or subtract underscores anywhere in the middle of a
327 word, so that these are also equivalent to C<\p{U_p_p_e_r}>. And white space
328 is irrelevant adjacent to non-word characters, such as the braces and the equals
329 or colon separators so C<\p{ Upper }> and C<\p{ Upper_case : Y }> are
330 equivalent to these as well. In fact, in most cases, white space and even
331 hyphens can be added or deleted anywhere. So even C<\p{ Up-per case = Yes}> is
332 equivalent. All this is called "loose-matching" by Unicode. The few places
333 where stricter matching is employed is in the middle of numbers, and the Perl
334 extension properties that begin or end with an underscore. Stricter matching
335 cares about white space (except adjacent to the non-word characters) and
336 hyphens, and non-interior underscores.
338 You can also use negation in both C<\p{}> and C<\P{}> by introducing a caret
339 (^) between the first brace and the property name: C<\p{^Tamil}> is
340 equal to C<\P{Tamil}>.
342 =head3 B<General_Category>
344 Every Unicode character is assigned a general category, which is the "most
345 usual categorization of a character" (from
346 L<http://www.unicode.org/reports/tr44>).
348 The compound way of writing these is like C<\p{General_Category=Number}>
349 (short, C<\p{gc:n}>). But Perl furnishes shortcuts in which everything up
350 through the equal or colon separator is omitted. So you can instead just write
353 Here are the short and long forms of the General Category properties:
358 LC, L& Cased_Letter (that is: [\p{Ll}\p{Lu}\p{Lt}])
371 Nd Decimal_Number (also Digit)
375 P Punctuation (also Punct)
376 Pc Connector_Punctuation
380 Pi Initial_Punctuation
381 (may behave like Ps or Pe depending on usage)
383 (may behave like Ps or Pe depending on usage)
395 Zp Paragraph_Separator
398 Cc Control (also Cntrl)
400 Cs Surrogate (not usable)
404 Single-letter properties match all characters in any of the
405 two-letter sub-properties starting with the same letter.
406 C<LC> and C<L&> are special cases, which are aliases for the set of
407 C<Ll>, C<Lu>, and C<Lt>.
409 Because Perl hides the need for the user to understand the internal
410 representation of Unicode characters, there is no need to implement
411 the somewhat messy concept of surrogates. C<Cs> is therefore not
414 =head3 B<Bidirectional Character Types>
416 Because scripts differ in their directionality--Hebrew is
417 written right to left, for example--Unicode supplies these properties in
418 the Bidi_Class class:
423 LRE Left-to-Right Embedding
424 LRO Left-to-Right Override
427 RLE Right-to-Left Embedding
428 RLO Right-to-Left Override
429 PDF Pop Directional Format
431 ES European Separator
432 ET European Terminator
437 B Paragraph Separator
442 This property is always written in the compound form.
443 For example, C<\p{Bidi_Class:R}> matches characters that are normally
444 written right to left.
448 The world's languages are written in a number of scripts. This sentence
449 (unless you're reading it in translation) is written in Latin, while Russian is
450 written in Cyrllic, and Greek is written in, well, Greek; Japanese mainly in
451 Hiragana or Katakana. There are many more.
453 The Unicode Script property gives what script a given character is in,
454 and can be matched with the compound form like C<\p{Script=Hebrew}> (short:
455 C<\p{sc=hebr}>). Perl furnishes shortcuts for all script names. You can omit
456 everything up through the equals (or colon), and simply write C<\p{Latin}> or
459 A complete list of scripts and their shortcuts is in L<perluniprops>.
461 =head3 B<Use of "Is" Prefix>
463 For backward compatibility (with Perl 5.6), all properties mentioned
464 so far may have C<Is> or C<Is_> prepended to their name, so C<\P{Is_Lu}>, for
465 example, is equal to C<\P{Lu}>, and C<\p{IsScript:Arabic}> is equal to
470 In addition to B<scripts>, Unicode also defines B<blocks> of
471 characters. The difference between scripts and blocks is that the
472 concept of scripts is closer to natural languages, while the concept
473 of blocks is more of an artificial grouping based on groups of Unicode
474 characters with consecutive ordinal values. For example, the "Basic Latin"
475 block is all characters whose ordinals are between 0 and 127, inclusive, in
476 other words, the ASCII characters. The "Latin" script contains some letters
477 from this block as well as several more, like "Latin-1 Supplement",
478 "Latin Extended-A", I<etc.>, but it does not contain all the characters from
479 those blocks. It does not, for example, contain digits, because digits are
480 shared across many scripts. Digits and similar groups, like punctuation, are in
481 the script called C<Common>. There is also a script called C<Inherited> for
482 characters that modify other characters, and inherit the script value of the
483 controlling character.
485 For more about scripts versus blocks, see UAX#24 "Unicode Script Property":
486 L<http://www.unicode.org/reports/tr24>
488 The Script property is likely to be the one you want to use when processing
489 natural language; the Block property may be useful in working with the nuts and
492 Block names are matched in the compound form, like C<\p{Block: Arrows}> or
493 C<\p{Blk=Hebrew}>. Unlike most other properties only a few block names have a
494 Unicode-defined short name. But Perl does provide a (slight) shortcut: You
495 can say, for example C<\p{In_Arrows}> or C<\p{In_Hebrew}>. For backwards
496 compatibility, the C<In> prefix may be omitted if there is no naming conflict
497 with a script or any other property, and you can even use an C<Is> prefix
498 instead in those cases. But it is not a good idea to do this, for a couple
505 It is confusing. There are many naming conflicts, and you may forget some.
506 For example, C<\p{Hebrew}> means the I<script> Hebrew, and NOT the I<block>
507 Hebrew. But would you remember that 6 months from now?
511 It is unstable. A new version of Unicode may pre-empt the current meaning by
512 creating a property with the same name. There was a time in very early Unicode
513 releases when C<\p{Hebrew}> would have matched the I<block> Hebrew; now it
518 Some people just prefer to always use C<\p{Block: foo}> and C<\p{Script: bar}>
519 instead of the shortcuts, for clarity, and because they can't remember the
520 difference between 'In' and 'Is' anyway (or aren't confident that those who
521 eventually will read their code will know).
523 A complete list of blocks and their shortcuts is in L<perluniprops>.
525 =head3 B<Other Properties>
527 There are many more properties than the very basic ones described here.
528 A complete list is in L<perluniprops>.
530 Unicode defines all its properties in the compound form, so all single-form
531 properties are Perl extensions. A number of these are just synonyms for the
532 Unicode ones, but some are genunine extensions, including a couple that are in
533 the compound form. And quite a few of these are actually recommended by Unicode
534 (in L<http://www.unicode.org/reports/tr18>).
536 This section gives some details on all the extensions that aren't synonyms for
537 compound-form Unicode properties (for those, you'll have to refer to the
538 L<Unicode Standard|http://www.unicode.org/reports/tr44>.
544 This matches any of the 1_114_112 Unicode code points. It is a synonym for
547 =item B<C<\p{Alnum}>>
549 This matches any C<\p{Alphabetic}> or C<\p{Decimal_Number}> character.
553 This matches any of the 1_114_112 Unicode code points. It is a synonym for
556 =item B<C<\p{Assigned}>>
558 This matches any assigned code point; that is, any code point whose general
559 category is not Unassigned (or equivalently, not Cn).
561 =item B<C<\p{Blank}>>
563 This is the same as C<\h> and C<\p{HorizSpace}>: A character that changes the
564 spacing horizontally.
566 =item B<C<\p{Decomposition_Type: Non_Canonical}>> (Short: C<\p{Dt=NonCanon}>)
568 Matches a character that has a non-canonical decomposition.
570 To understand the use of this rarely used property=value combination, it is
571 necessary to know some basics about decomposition.
572 Consider a character, say H. It could appear with various marks around it,
573 such as an acute accent, or a circumflex, or various hooks, circles, arrows,
574 I<etc.>, above, below, to one side and/or the other, I<etc.> There are many
575 possibilities among the world's languages. The number of combinations is
576 astronomical, and if there were a character for each combination, it would
577 soon exhaust Unicode's more than a million possible characters. So Unicode
578 took a different approach: there is a character for the base H, and a
579 character for each of the possible marks, and they can be combined variously
580 to get a final logical character. So a logical character--what appears to be a
581 single character--can be a sequence of more than one individual characters.
582 This is called an "extended grapheme cluster". (Perl furnishes the C<\X>
583 construct to match such sequences.)
585 But Unicode's intent is to unify the existing character set standards and
586 practices, and a number of pre-existing standards have single characters that
587 mean the same thing as some of these combinations. An example is ISO-8859-1,
588 which has quite a few of these in the Latin-1 range, an example being "LATIN
589 CAPITAL LETTER E WITH ACUTE". Because this character was in this pre-existing
590 standard, Unicode added it to its repertoire. But this character is considered
591 by Unicode to be equivalent to the sequence consisting of first the character
592 "LATIN CAPITAL LETTER E", then the character "COMBINING ACUTE ACCENT".
594 "LATIN CAPITAL LETTER E WITH ACUTE" is called a "pre-composed" character, and
595 the equivalence with the sequence is called canonical equivalence. All
596 pre-composed characters are said to have a decomposition (into the equivalent
597 sequence) and the decomposition type is also called canonical.
599 However, many more characters have a different type of decomposition, a
600 "compatible" or "non-canonical" decomposition. The sequences that form these
601 decompositions are not considered canonically equivalent to the pre-composed
602 character. An example, again in the Latin-1 range, is the "SUPERSCRIPT ONE".
603 It is kind of like a regular digit 1, but not exactly; its decomposition
604 into the digit 1 is called a "compatible" decomposition, specifically a
605 "super" decomposition. There are several such compatibility
606 decompositions (see L<http://www.unicode.org/reports/tr44>), including one
607 called "compat" which means some miscellaneous type of decomposition
608 that doesn't fit into the decomposition categories that Unicode has chosen.
610 Note that most Unicode characters don't have a decomposition, so their
611 decomposition type is "None".
613 Perl has added the C<Non_Canonical> type, for your convenience, to mean any of
614 the compatibility decompositions.
616 =item B<C<\p{Graph}>>
618 Matches any character that is graphic. Theoretically, this means a character
619 that on a printer would cause ink to be used.
621 =item B<C<\p{HorizSpace}>>
623 This is the same as C<\h> and C<\p{Blank}>: A character that changes the
624 spacing horizontally.
628 This is a synonym for C<\p{Present_In=*}>
630 =item B<C<\p{PerlSpace}>>
632 This is the same as C<\s>, restricted to ASCII, namely C<S<[ \f\n\r\t]>>.
634 Mnemonic: Perl's (original) space
636 =item B<C<\p{PerlWord}>>
638 This is the same as C<\w>, restricted to ASCII, namely C<[A-Za-z0-9_]>
640 Mnemonic: Perl's (original) word.
642 =item B<C<\p{PosixAlnum}>>
644 This matches any alphanumeric character in the ASCII range, namely
647 =item B<C<\p{PosixAlpha}>>
649 This matches any alphabetic character in the ASCII range, namely C<[A-Za-z]>.
651 =item B<C<\p{PosixBlank}>>
653 This matches any blank character in the ASCII range, namely C<S<[ \t]>>.
655 =item B<C<\p{PosixCntrl}>>
657 This matches any control character in the ASCII range, namely C<[\x00-\x1F\x7F]>
659 =item B<C<\p{PosixDigit}>>
661 This matches any digit character in the ASCII range, namely C<[0-9]>.
663 =item B<C<\p{PosixGraph}>>
665 This matches any graphical character in the ASCII range, namely C<[\x21-\x7E]>.
667 =item B<C<\p{PosixLower}>>
669 This matches any lowercase character in the ASCII range, namely C<[a-z]>.
671 =item B<C<\p{PosixPrint}>>
673 This matches any printable character in the ASCII range, namely C<[\x20-\x7E]>.
674 These are the graphical characters plus SPACE.
676 =item B<C<\p{PosixPunct}>>
678 This matches any punctuation character in the ASCII range, namely
679 C<[\x21-\x2F\x3A-\x40\x5B-\x60\x7B-\x7E]>. These are the
680 graphical characters that aren't word characters. Note that the Posix standard
681 includes in its definition of punctuation, those characters that Unicode calls
684 =item B<C<\p{PosixSpace}>>
686 This matches any space character in the ASCII range, namely
687 C<S<[ \f\n\r\t\x0B]>> (the last being a vertical tab).
689 =item B<C<\p{PosixUpper}>>
691 This matches any uppercase character in the ASCII range, namely C<[A-Z]>.
693 =item B<C<\p{Present_In: *}>> (Short: C<\p{In=*}>)
695 This property is used when you need to know in what Unicode version(s) a
698 The "*" above stands for some two digit Unicode version number, such as
699 C<1.1> or C<4.0>; or the "*" can also be C<Unassigned>. This property will
700 match the code points whose final disposition has been settled as of the
701 Unicode release given by the version number; C<\p{Present_In: Unassigned}>
702 will match those code points whose meaning has yet to be assigned.
704 For example, C<U+0041> "LATIN CAPITAL LETTER A" was present in the very first
705 Unicode release available, which is C<1.1>, so this property is true for all
706 valid "*" versions. On the other hand, C<U+1EFF> was not assigned until version
707 5.1 when it became "LATIN SMALL LETTER Y WITH LOOP", so the only "*" that
708 would match it are 5.1, 5.2, and later.
710 Unicode furnishes the C<Age> property from which this is derived. The problem
711 with Age is that a strict interpretation of it (which Perl takes) has it
712 matching the precise release a code point's meaning is introduced in. Thus
713 C<U+0041> would match only 1.1; and C<U+1EFF> only 5.1. This is not usually what
716 Some non-Perl implementations of the Age property may change its meaning to be
717 the same as the Perl Present_In property; just be aware of that.
719 Another confusion with both these properties is that the definition is not
720 that the code point has been assigned, but that the meaning of the code point
721 has been determined. This is because 66 code points will always be
722 unassigned, and, so the Age for them is the Unicode version the decision to
723 make them so was made in. For example, C<U+FDD0> is to be permanently
724 unassigned to a character, and the decision to do that was made in version 3.1,
725 so C<\p{Age=3.1}> matches this character and C<\p{Present_In: 3.1}> and up
728 =item B<C<\p{Print}>>
730 This matches any character that is graphical or blank, except controls.
732 =item B<C<\p{SpacePerl}>>
734 This is the same as C<\s>, including beyond ASCII.
736 Mnemonic: Space, as modified by Perl. (It doesn't include the vertical tab
737 which both the Posix standard and Unicode consider to be space.)
739 =item B<C<\p{VertSpace}>>
741 This is the same as C<\v>: A character that changes the spacing vertically.
745 This is the same as C<\w>, including beyond ASCII.
749 =head2 User-Defined Character Properties
751 You can define your own binary character properties by defining subroutines
752 whose names begin with "In" or "Is". The subroutines can be defined in any
753 package. The user-defined properties can be used in the regular expression
754 C<\p> and C<\P> constructs; if you are using a user-defined property from a
755 package other than the one you are in, you must specify its package in the
756 C<\p> or C<\P> construct.
758 # assuming property Is_Foreign defined in Lang::
759 package main; # property package name required
760 if ($txt =~ /\p{Lang::IsForeign}+/) { ... }
762 package Lang; # property package name not required
763 if ($txt =~ /\p{IsForeign}+/) { ... }
766 Note that the effect is compile-time and immutable once defined.
768 The subroutines must return a specially-formatted string, with one
769 or more newline-separated lines. Each line must be one of the following:
775 A single hexadecimal number denoting a Unicode code point to include.
779 Two hexadecimal numbers separated by horizontal whitespace (space or
780 tabular characters) denoting a range of Unicode code points to include.
784 Something to include, prefixed by "+": a built-in character
785 property (prefixed by "utf8::") or a user-defined character property,
786 to represent all the characters in that property; two hexadecimal code
787 points for a range; or a single hexadecimal code point.
791 Something to exclude, prefixed by "-": an existing character
792 property (prefixed by "utf8::") or a user-defined character property,
793 to represent all the characters in that property; two hexadecimal code
794 points for a range; or a single hexadecimal code point.
798 Something to negate, prefixed "!": an existing character
799 property (prefixed by "utf8::") or a user-defined character property,
800 to represent all the characters in that property; two hexadecimal code
801 points for a range; or a single hexadecimal code point.
805 Something to intersect with, prefixed by "&": an existing character
806 property (prefixed by "utf8::") or a user-defined character property,
807 for all the characters except the characters in the property; two
808 hexadecimal code points for a range; or a single hexadecimal code point.
812 For example, to define a property that covers both the Japanese
813 syllabaries (hiragana and katakana), you can define
822 Imagine that the here-doc end marker is at the beginning of the line.
823 Now you can use C<\p{InKana}> and C<\P{InKana}>.
825 You could also have used the existing block property names:
834 Suppose you wanted to match only the allocated characters,
835 not the raw block ranges: in other words, you want to remove
846 The negation is useful for defining (surprise!) negated classes.
856 Intersection is useful for getting the common characters matched by
857 two (or more) classes.
866 It's important to remember not to use "&" for the first set; that
867 would be intersecting with nothing (resulting in an empty set).
869 =head2 User-Defined Case Mappings
871 You can also define your own mappings to be used in the lc(),
872 lcfirst(), uc(), and ucfirst() (or their string-inlined versions).
873 The principle is similar to that of user-defined character
874 properties: to define subroutines
875 with names like C<ToLower> (for lc() and lcfirst()), C<ToTitle> (for
876 the first character in ucfirst()), and C<ToUpper> (for uc(), and the
877 rest of the characters in ucfirst()).
879 The string returned by the subroutines needs to be two hexadecimal numbers
880 separated by two tabulators: the two numbers being, respectively, the source
881 code point and the destination code point. For example:
889 defines an uc() mapping that causes only the character "a"
890 to be mapped to "A"; all other characters will remain unchanged.
892 (For serious hackers only) The above means you have to furnish a complete
893 mapping; you can't just override a couple of characters and leave the rest
894 unchanged. You can find all the mappings in the directory
895 C<$Config{privlib}>/F<unicore/To/>. The mapping data is returned as the
896 here-document, and the C<utf8::ToSpecFoo> are special exception mappings
897 derived from <$Config{privlib}>/F<unicore/SpecialCasing.txt>. The "Digit" and
898 "Fold" mappings that one can see in the directory are not directly
899 user-accessible, one can use either the C<Unicode::UCD> module, or just match
900 case-insensitively (that's when the "Fold" mapping is used).
902 The mappings will only take effect on scalars that have been marked as having
903 Unicode characters, for example by using C<utf8::upgrade()>.
904 Old byte-style strings are not affected.
906 The mappings are in effect for the package they are defined in.
908 =head2 Character Encodings for Input and Output
912 =head2 Unicode Regular Expression Support Level
914 The following list of Unicode support for regular expressions describes
915 all the features currently supported. The references to "Level N"
916 and the section numbers refer to the Unicode Technical Standard #18,
917 "Unicode Regular Expressions", version 11, in May 2005.
923 Level 1 - Basic Unicode Support
925 RL1.1 Hex Notation - done [1]
926 RL1.2 Properties - done [2][3]
927 RL1.2a Compatibility Properties - done [4]
928 RL1.3 Subtraction and Intersection - MISSING [5]
929 RL1.4 Simple Word Boundaries - done [6]
930 RL1.5 Simple Loose Matches - done [7]
931 RL1.6 Line Boundaries - MISSING [8]
932 RL1.7 Supplementary Code Points - done [9]
936 [3] supports not only minimal list, but all Unicode character
937 properties (see L</Unicode Character Properties>)
938 [4] \d \D \s \S \w \W \X [:prop:] [:^prop:]
939 [5] can use regular expression look-ahead [a] or
940 user-defined character properties [b] to emulate set operations
942 [7] note that Perl does Full case-folding in matching (but with bugs),
943 not Simple: for example U+1F88 is equivalent to U+1F00 U+03B9,
944 not with 1F80. This difference matters mainly for certain Greek
945 capital letters with certain modifiers: the Full case-folding
946 decomposes the letter, while the Simple case-folding would map
947 it to a single character.
948 [8] should do ^ and $ also on U+000B (\v in C), FF (\f), CR (\r),
949 CRLF (\r\n), NEL (U+0085), LS (U+2028), and PS (U+2029);
950 should also affect <>, $., and script line numbers;
951 should not split lines within CRLF [c] (i.e. there is no empty
952 line between \r and \n)
953 [9] UTF-8/UTF-EBDDIC used in perl allows not only U+10000 to U+10FFFF
954 but also beyond U+10FFFF [d]
956 [a] You can mimic class subtraction using lookahead.
957 For example, what UTS#18 might write as
959 [{Greek}-[{UNASSIGNED}]]
961 in Perl can be written as:
963 (?!\p{Unassigned})\p{InGreekAndCoptic}
964 (?=\p{Assigned})\p{InGreekAndCoptic}
966 But in this particular example, you probably really want
970 which will match assigned characters known to be part of the Greek script.
972 Also see the Unicode::Regex::Set module, it does implement the full
973 UTS#18 grouping, intersection, union, and removal (subtraction) syntax.
975 [b] '+' for union, '-' for removal (set-difference), '&' for intersection
976 (see L</"User-Defined Character Properties">)
978 [c] Try the C<:crlf> layer (see L<PerlIO>).
980 [d] U+FFFF will currently generate a warning message if 'utf8' warnings are
985 Level 2 - Extended Unicode Support
987 RL2.1 Canonical Equivalents - MISSING [10][11]
988 RL2.2 Default Grapheme Clusters - MISSING [12]
989 RL2.3 Default Word Boundaries - MISSING [14]
990 RL2.4 Default Loose Matches - MISSING [15]
991 RL2.5 Name Properties - MISSING [16]
992 RL2.6 Wildcard Properties - MISSING
994 [10] see UAX#15 "Unicode Normalization Forms"
995 [11] have Unicode::Normalize but not integrated to regexes
996 [12] have \X but we don't have a "Grapheme Cluster Mode"
997 [14] see UAX#29, Word Boundaries
998 [15] see UAX#21 "Case Mappings"
999 [16] have \N{...} but neither compute names of CJK Ideographs
1000 and Hangul Syllables nor use a loose match [e]
1002 [e] C<\N{...}> allows namespaces (see L<charnames>).
1006 Level 3 - Tailored Support
1008 RL3.1 Tailored Punctuation - MISSING
1009 RL3.2 Tailored Grapheme Clusters - MISSING [17][18]
1010 RL3.3 Tailored Word Boundaries - MISSING
1011 RL3.4 Tailored Loose Matches - MISSING
1012 RL3.5 Tailored Ranges - MISSING
1013 RL3.6 Context Matching - MISSING [19]
1014 RL3.7 Incremental Matches - MISSING
1015 ( RL3.8 Unicode Set Sharing )
1016 RL3.9 Possible Match Sets - MISSING
1017 RL3.10 Folded Matching - MISSING [20]
1018 RL3.11 Submatchers - MISSING
1020 [17] see UAX#10 "Unicode Collation Algorithms"
1021 [18] have Unicode::Collate but not integrated to regexes
1022 [19] have (?<=x) and (?=x), but look-aheads or look-behinds should see
1023 outside of the target substring
1024 [20] need insensitive matching for linguistic features other than case;
1025 for example, hiragana to katakana, wide and narrow, simplified Han
1026 to traditional Han (see UTR#30 "Character Foldings")
1030 =head2 Unicode Encodings
1032 Unicode characters are assigned to I<code points>, which are abstract
1033 numbers. To use these numbers, various encodings are needed.
1041 UTF-8 is a variable-length (1 to 6 bytes, current character allocations
1042 require 4 bytes), byte-order independent encoding. For ASCII (and we
1043 really do mean 7-bit ASCII, not another 8-bit encoding), UTF-8 is
1046 The following table is from Unicode 3.2.
1048 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1050 U+0000..U+007F 00..7F
1051 U+0080..U+07FF * C2..DF 80..BF
1052 U+0800..U+0FFF E0 * A0..BF 80..BF
1053 U+1000..U+CFFF E1..EC 80..BF 80..BF
1054 U+D000..U+D7FF ED 80..9F 80..BF
1055 U+D800..U+DFFF +++++++ utf16 surrogates, not legal utf8 +++++++
1056 U+E000..U+FFFF EE..EF 80..BF 80..BF
1057 U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF
1058 U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
1059 U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
1061 Note the gaps before several of the byte entries above marked by '*'. These are
1062 caused by legal UTF-8 avoiding non-shortest encodings: it is technically
1063 possible to UTF-8-encode a single code point in different ways, but that is
1064 explicitly forbidden, and the shortest possible encoding should always be used
1065 (and that is what Perl does).
1067 Another way to look at it is via bits:
1069 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1072 00000bbbbbaaaaaa 110bbbbb 10aaaaaa
1073 ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
1074 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
1076 As you can see, the continuation bytes all begin with "10", and the
1077 leading bits of the start byte tell how many bytes there are in the
1084 Like UTF-8 but EBCDIC-safe, in the way that UTF-8 is ASCII-safe.
1088 UTF-16, UTF-16BE, UTF-16LE, Surrogates, and BOMs (Byte Order Marks)
1090 The followings items are mostly for reference and general Unicode
1091 knowledge, Perl doesn't use these constructs internally.
1093 UTF-16 is a 2 or 4 byte encoding. The Unicode code points
1094 C<U+0000..U+FFFF> are stored in a single 16-bit unit, and the code
1095 points C<U+10000..U+10FFFF> in two 16-bit units. The latter case is
1096 using I<surrogates>, the first 16-bit unit being the I<high
1097 surrogate>, and the second being the I<low surrogate>.
1099 Surrogates are code points set aside to encode the C<U+10000..U+10FFFF>
1100 range of Unicode code points in pairs of 16-bit units. The I<high
1101 surrogates> are the range C<U+D800..U+DBFF> and the I<low surrogates>
1102 are the range C<U+DC00..U+DFFF>. The surrogate encoding is
1104 $hi = ($uni - 0x10000) / 0x400 + 0xD800;
1105 $lo = ($uni - 0x10000) % 0x400 + 0xDC00;
1109 $uni = 0x10000 + ($hi - 0xD800) * 0x400 + ($lo - 0xDC00);
1111 If you try to generate surrogates (for example by using chr()), you
1112 will get a warning, if warnings are turned on, because those code
1113 points are not valid for a Unicode character.
1115 Because of the 16-bitness, UTF-16 is byte-order dependent. UTF-16
1116 itself can be used for in-memory computations, but if storage or
1117 transfer is required either UTF-16BE (big-endian) or UTF-16LE
1118 (little-endian) encodings must be chosen.
1120 This introduces another problem: what if you just know that your data
1121 is UTF-16, but you don't know which endianness? Byte Order Marks, or
1122 BOMs, are a solution to this. A special character has been reserved
1123 in Unicode to function as a byte order marker: the character with the
1124 code point C<U+FEFF> is the BOM.
1126 The trick is that if you read a BOM, you will know the byte order,
1127 since if it was written on a big-endian platform, you will read the
1128 bytes C<0xFE 0xFF>, but if it was written on a little-endian platform,
1129 you will read the bytes C<0xFF 0xFE>. (And if the originating platform
1130 was writing in UTF-8, you will read the bytes C<0xEF 0xBB 0xBF>.)
1132 The way this trick works is that the character with the code point
1133 C<U+FFFE> is guaranteed not to be a valid Unicode character, so the
1134 sequence of bytes C<0xFF 0xFE> is unambiguously "BOM, represented in
1135 little-endian format" and cannot be C<U+FFFE>, represented in big-endian
1136 format". (Actually, C<U+FFFE> is legal for use by your program, even for
1137 input/output, but better not use it if you need a BOM. But it is "illegal for
1138 interchange", so that an unsuspecting program won't get confused.)
1142 UTF-32, UTF-32BE, UTF-32LE
1144 The UTF-32 family is pretty much like the UTF-16 family, expect that
1145 the units are 32-bit, and therefore the surrogate scheme is not
1146 needed. The BOM signatures will be C<0x00 0x00 0xFE 0xFF> for BE and
1147 C<0xFF 0xFE 0x00 0x00> for LE.
1153 Encodings defined by the ISO 10646 standard. UCS-2 is a 16-bit
1154 encoding. Unlike UTF-16, UCS-2 is not extensible beyond C<U+FFFF>,
1155 because it does not use surrogates. UCS-4 is a 32-bit encoding,
1156 functionally identical to UTF-32.
1162 A seven-bit safe (non-eight-bit) encoding, which is useful if the
1163 transport or storage is not eight-bit safe. Defined by RFC 2152.
1167 =head2 Security Implications of Unicode
1169 Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>.
1170 Also, note the following:
1178 Unfortunately, the specification of UTF-8 leaves some room for
1179 interpretation of how many bytes of encoded output one should generate
1180 from one input Unicode character. Strictly speaking, the shortest
1181 possible sequence of UTF-8 bytes should be generated,
1182 because otherwise there is potential for an input buffer overflow at
1183 the receiving end of a UTF-8 connection. Perl always generates the
1184 shortest length UTF-8, and with warnings on, Perl will warn about
1185 non-shortest length UTF-8 along with other malformations, such as the
1186 surrogates, which are not real Unicode code points.
1190 Regular expressions behave slightly differently between byte data and
1191 character (Unicode) data. For example, the "word character" character
1192 class C<\w> will work differently depending on if data is eight-bit bytes
1195 In the first case, the set of C<\w> characters is either small--the
1196 default set of alphabetic characters, digits, and the "_"--or, if you
1197 are using a locale (see L<perllocale>), the C<\w> might contain a few
1198 more letters according to your language and country.
1200 In the second case, the C<\w> set of characters is much, much larger.
1201 Most importantly, even in the set of the first 256 characters, it will
1202 probably match different characters: unlike most locales, which are
1203 specific to a language and country pair, Unicode classifies all the
1204 characters that are letters I<somewhere> as C<\w>. For example, your
1205 locale might not think that LATIN SMALL LETTER ETH is a letter (unless
1206 you happen to speak Icelandic), but Unicode does.
1208 As discussed elsewhere, Perl has one foot (two hooves?) planted in
1209 each of two worlds: the old world of bytes and the new world of
1210 characters, upgrading from bytes to characters when necessary.
1211 If your legacy code does not explicitly use Unicode, no automatic
1212 switch-over to characters should happen. Characters shouldn't get
1213 downgraded to bytes, either. It is possible to accidentally mix bytes
1214 and characters, however (see L<perluniintro>), in which case C<\w> in
1215 regular expressions might start behaving differently. Review your
1216 code. Use warnings and the C<strict> pragma.
1220 =head2 Unicode in Perl on EBCDIC
1222 The way Unicode is handled on EBCDIC platforms is still
1223 experimental. On such platforms, references to UTF-8 encoding in this
1224 document and elsewhere should be read as meaning the UTF-EBCDIC
1225 specified in Unicode Technical Report 16, unless ASCII vs. EBCDIC issues
1226 are specifically discussed. There is no C<utfebcdic> pragma or
1227 ":utfebcdic" layer; rather, "utf8" and ":utf8" are reused to mean
1228 the platform's "natural" 8-bit encoding of Unicode. See L<perlebcdic>
1229 for more discussion of the issues.
1233 Usually locale settings and Unicode do not affect each other, but
1234 there are a couple of exceptions:
1240 You can enable automatic UTF-8-ification of your standard file
1241 handles, default C<open()> layer, and C<@ARGV> by using either
1242 the C<-C> command line switch or the C<PERL_UNICODE> environment
1243 variable, see L<perlrun> for the documentation of the C<-C> switch.
1247 Perl tries really hard to work both with Unicode and the old
1248 byte-oriented world. Most often this is nice, but sometimes Perl's
1249 straddling of the proverbial fence causes problems.
1253 =head2 When Unicode Does Not Happen
1255 While Perl does have extensive ways to input and output in Unicode,
1256 and few other 'entry points' like the @ARGV which can be interpreted
1257 as Unicode (UTF-8), there still are many places where Unicode (in some
1258 encoding or another) could be given as arguments or received as
1259 results, or both, but it is not.
1261 The following are such interfaces. Also, see L</The "Unicode Bug">.
1262 For all of these interfaces Perl
1263 currently (as of 5.8.3) simply assumes byte strings both as arguments
1264 and results, or UTF-8 strings if the C<encoding> pragma has been used.
1266 One reason why Perl does not attempt to resolve the role of Unicode in
1267 these cases is that the answers are highly dependent on the operating
1268 system and the file system(s). For example, whether filenames can be
1269 in Unicode, and in exactly what kind of encoding, is not exactly a
1270 portable concept. Similarly for the qx and system: how well will the
1271 'command line interface' (and which of them?) handle Unicode?
1277 chdir, chmod, chown, chroot, exec, link, lstat, mkdir,
1278 rename, rmdir, stat, symlink, truncate, unlink, utime, -X
1290 open, opendir, sysopen
1294 qx (aka the backtick operator), system
1302 =head2 The "Unicode Bug"
1304 The term, the "Unicode bug" has been applied to an inconsistency with the
1305 Unicode characters whose ordinals are in the Latin-1 Supplement block, that
1306 is, between 128 and 255. Without a locale specified, unlike all other
1307 characters or code points, these characters have very different semantics in
1308 byte semantics versus character semantics.
1310 In character semantics they are interpreted as Unicode code points, which means
1311 they have the same semantics as Latin-1 (ISO-8859-1).
1313 In byte semantics, they are considered to be unassigned characters, meaning
1314 that the only semantics they have is their ordinal numbers, and that they are
1315 not members of various character classes. None are considered to match C<\w>
1316 for example, but all match C<\W>. (On EBCDIC platforms, the behavior may
1317 be different from this, depending on the underlying C language library
1320 The behavior is known to have effects on these areas:
1326 Changing the case of a scalar, that is, using C<uc()>, C<ucfirst()>, C<lc()>,
1327 and C<lcfirst()>, or C<\L>, C<\U>, C<\u> and C<\l> in regular expression
1332 Using caseless (C</i>) regular expression matching
1336 Matching a number of properties in regular expressions, such as C<\w>
1340 User-defined case change mappings. You can create a C<ToUpper()> function, for
1341 example, which overrides Perl's built-in case mappings. The scalar must be
1342 encoded in utf8 for your function to actually be invoked.
1346 This behavior can lead to unexpected results in which a string's semantics
1347 suddenly change if a code point above 255 is appended to or removed from it,
1348 which changes the string's semantics from byte to character or vice versa. As
1349 an example, consider the following program and its output:
1354 for ($s1, $s2, $s1.$s2) {
1362 If there's no C<\w> in C<s1> or in C<s2>, why does their concatenation have one?
1364 This anomaly stems from Perl's attempt to not disturb older programs that
1365 didn't use Unicode, and hence had no semantics for characters outside of the
1366 ASCII range (except in a locale), along with Perl's desire to add Unicode
1367 support seamlessly. The result wasn't seamless: these characters were
1370 Work is being done to correct this, but only some of it was complete in time
1371 for the 5.12 release. What has been finished is the important part of the case
1372 changing component. Due to concerns, and some evidence, that older code might
1373 have come to rely on the existing behavior, the new behavior must be explicitly
1374 enabled by the feature C<unicode_strings> in the L<feature> pragma, even though
1375 no new syntax is involved.
1377 See L<perlfunc/lc> for details on how this pragma works in combination with
1378 various others for casing. Even though the pragma only affects casing
1379 operations in the 5.12 release, it is planned to have it affect all the
1380 problematic behaviors in later releases: you can't have one without them all.
1382 In the meantime, a workaround is to always call utf8::upgrade($string), or to
1383 use the standard module L<Encode>. Also, a scalar that has any characters
1384 whose ordinal is above 0x100, or which were specified using either of the
1385 C<\N{...}> notations will automatically have character semantics.
1387 =head2 Forcing Unicode in Perl (Or Unforcing Unicode in Perl)
1389 Sometimes (see L</"When Unicode Does Not Happen"> or L</The "Unicode Bug">)
1390 there are situations where you simply need to force a byte
1391 string into UTF-8, or vice versa. The low-level calls
1392 utf8::upgrade($bytestring) and utf8::downgrade($utf8string[, FAIL_OK]) are
1395 Note that utf8::downgrade() can fail if the string contains characters
1396 that don't fit into a byte.
1398 Calling either function on a string that already is in the desired state is a
1401 =head2 Using Unicode in XS
1403 If you want to handle Perl Unicode in XS extensions, you may find the
1404 following C APIs useful. See also L<perlguts/"Unicode Support"> for an
1405 explanation about Unicode at the XS level, and L<perlapi> for the API
1412 C<DO_UTF8(sv)> returns true if the C<UTF8> flag is on and the bytes
1413 pragma is not in effect. C<SvUTF8(sv)> returns true if the C<UTF8>
1414 flag is on; the bytes pragma is ignored. The C<UTF8> flag being on
1415 does B<not> mean that there are any characters of code points greater
1416 than 255 (or 127) in the scalar or that there are even any characters
1417 in the scalar. What the C<UTF8> flag means is that the sequence of
1418 octets in the representation of the scalar is the sequence of UTF-8
1419 encoded code points of the characters of a string. The C<UTF8> flag
1420 being off means that each octet in this representation encodes a
1421 single character with code point 0..255 within the string. Perl's
1422 Unicode model is not to use UTF-8 until it is absolutely necessary.
1426 C<uvchr_to_utf8(buf, chr)> writes a Unicode character code point into
1427 a buffer encoding the code point as UTF-8, and returns a pointer
1428 pointing after the UTF-8 bytes. It works appropriately on EBCDIC machines.
1432 C<utf8_to_uvchr(buf, lenp)> reads UTF-8 encoded bytes from a buffer and
1433 returns the Unicode character code point and, optionally, the length of
1434 the UTF-8 byte sequence. It works appropriately on EBCDIC machines.
1438 C<utf8_length(start, end)> returns the length of the UTF-8 encoded buffer
1439 in characters. C<sv_len_utf8(sv)> returns the length of the UTF-8 encoded
1444 C<sv_utf8_upgrade(sv)> converts the string of the scalar to its UTF-8
1445 encoded form. C<sv_utf8_downgrade(sv)> does the opposite, if
1446 possible. C<sv_utf8_encode(sv)> is like sv_utf8_upgrade except that
1447 it does not set the C<UTF8> flag. C<sv_utf8_decode()> does the
1448 opposite of C<sv_utf8_encode()>. Note that none of these are to be
1449 used as general-purpose encoding or decoding interfaces: C<use Encode>
1450 for that. C<sv_utf8_upgrade()> is affected by the encoding pragma
1451 but C<sv_utf8_downgrade()> is not (since the encoding pragma is
1452 designed to be a one-way street).
1456 C<is_utf8_char(s)> returns true if the pointer points to a valid UTF-8
1461 C<is_utf8_string(buf, len)> returns true if C<len> bytes of the buffer
1466 C<UTF8SKIP(buf)> will return the number of bytes in the UTF-8 encoded
1467 character in the buffer. C<UNISKIP(chr)> will return the number of bytes
1468 required to UTF-8-encode the Unicode character code point. C<UTF8SKIP()>
1469 is useful for example for iterating over the characters of a UTF-8
1470 encoded buffer; C<UNISKIP()> is useful, for example, in computing
1471 the size required for a UTF-8 encoded buffer.
1475 C<utf8_distance(a, b)> will tell the distance in characters between the
1476 two pointers pointing to the same UTF-8 encoded buffer.
1480 C<utf8_hop(s, off)> will return a pointer to a UTF-8 encoded buffer
1481 that is C<off> (positive or negative) Unicode characters displaced
1482 from the UTF-8 buffer C<s>. Be careful not to overstep the buffer:
1483 C<utf8_hop()> will merrily run off the end or the beginning of the
1484 buffer if told to do so.
1488 C<pv_uni_display(dsv, spv, len, pvlim, flags)> and
1489 C<sv_uni_display(dsv, ssv, pvlim, flags)> are useful for debugging the
1490 output of Unicode strings and scalars. By default they are useful
1491 only for debugging--they display B<all> characters as hexadecimal code
1492 points--but with the flags C<UNI_DISPLAY_ISPRINT>,
1493 C<UNI_DISPLAY_BACKSLASH>, and C<UNI_DISPLAY_QQ> you can make the
1494 output more readable.
1498 C<ibcmp_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2)> can be used to
1499 compare two strings case-insensitively in Unicode. For case-sensitive
1500 comparisons you can just use C<memEQ()> and C<memNE()> as usual.
1504 For more information, see L<perlapi>, and F<utf8.c> and F<utf8.h>
1505 in the Perl source code distribution.
1507 =head2 Hacking Perl to work on earlier Unicode versions (for very serious hackers only)
1509 Perl by default comes with the latest supported Unicode version built in, but
1510 you can change to use any earlier one.
1512 Download the files in the version of Unicode that you want from the Unicode web
1513 site L<http://www.unicode.org>). These should replace the existing files in
1514 C<\$Config{privlib}>/F<unicore>. (C<\%Config> is available from the Config
1515 module.) Follow the instructions in F<README.perl> in that directory to change
1516 some of their names, and then run F<make>.
1518 It is even possible to download them to a different directory, and then change
1519 F<utf8_heavy.pl> in the directory C<\$Config{privlib}> to point to the new
1520 directory, or maybe make a copy of that directory before making the change, and
1521 using C<@INC> or the C<-I> run-time flag to switch between versions at will
1522 (but because of caching, not in the middle of a process), but all this is
1523 beyond the scope of these instructions.
1527 =head2 Interaction with Locales
1529 Use of locales with Unicode data may lead to odd results. Currently,
1530 Perl attempts to attach 8-bit locale info to characters in the range
1531 0..255, but this technique is demonstrably incorrect for locales that
1532 use characters above that range when mapped into Unicode. Perl's
1533 Unicode support will also tend to run slower. Use of locales with
1534 Unicode is discouraged.
1536 =head2 Problems with characters in the Latin-1 Supplement range
1538 See L</The "Unicode Bug">
1540 =head2 Problems with case-insensitive regular expression matching
1542 There are problems with case-insensitive matches, including those involving
1543 character classes (enclosed in [square brackets]), characters whose fold
1544 is to multiple characters (such as the single character LATIN SMALL LIGATURE
1545 FFL matches case-insensitively with the 3-character string C<ffl>), and
1546 characters in the Latin-1 Supplement.
1548 =head2 Interaction with Extensions
1550 When Perl exchanges data with an extension, the extension should be
1551 able to understand the UTF8 flag and act accordingly. If the
1552 extension doesn't know about the flag, it's likely that the extension
1553 will return incorrectly-flagged data.
1555 So if you're working with Unicode data, consult the documentation of
1556 every module you're using if there are any issues with Unicode data
1557 exchange. If the documentation does not talk about Unicode at all,
1558 suspect the worst and probably look at the source to learn how the
1559 module is implemented. Modules written completely in Perl shouldn't
1560 cause problems. Modules that directly or indirectly access code written
1561 in other programming languages are at risk.
1563 For affected functions, the simple strategy to avoid data corruption is
1564 to always make the encoding of the exchanged data explicit. Choose an
1565 encoding that you know the extension can handle. Convert arguments passed
1566 to the extensions to that encoding and convert results back from that
1567 encoding. Write wrapper functions that do the conversions for you, so
1568 you can later change the functions when the extension catches up.
1570 To provide an example, let's say the popular Foo::Bar::escape_html
1571 function doesn't deal with Unicode data yet. The wrapper function
1572 would convert the argument to raw UTF-8 and convert the result back to
1573 Perl's internal representation like so:
1575 sub my_escape_html ($) {
1577 return unless defined $what;
1578 Encode::decode_utf8(Foo::Bar::escape_html(Encode::encode_utf8($what)));
1581 Sometimes, when the extension does not convert data but just stores
1582 and retrieves them, you will be in a position to use the otherwise
1583 dangerous Encode::_utf8_on() function. Let's say the popular
1584 C<Foo::Bar> extension, written in C, provides a C<param> method that
1585 lets you store and retrieve data according to these prototypes:
1587 $self->param($name, $value); # set a scalar
1588 $value = $self->param($name); # retrieve a scalar
1590 If it does not yet provide support for any encoding, one could write a
1591 derived class with such a C<param> method:
1594 my($self,$name,$value) = @_;
1595 utf8::upgrade($name); # make sure it is UTF-8 encoded
1596 if (defined $value) {
1597 utf8::upgrade($value); # make sure it is UTF-8 encoded
1598 return $self->SUPER::param($name,$value);
1600 my $ret = $self->SUPER::param($name);
1601 Encode::_utf8_on($ret); # we know, it is UTF-8 encoded
1606 Some extensions provide filters on data entry/exit points, such as
1607 DB_File::filter_store_key and family. Look out for such filters in
1608 the documentation of your extensions, they can make the transition to
1609 Unicode data much easier.
1613 Some functions are slower when working on UTF-8 encoded strings than
1614 on byte encoded strings. All functions that need to hop over
1615 characters such as length(), substr() or index(), or matching regular
1616 expressions can work B<much> faster when the underlying data are
1619 In Perl 5.8.0 the slowness was often quite spectacular; in Perl 5.8.1
1620 a caching scheme was introduced which will hopefully make the slowness
1621 somewhat less spectacular, at least for some operations. In general,
1622 operations with UTF-8 encoded strings are still slower. As an example,
1623 the Unicode properties (character classes) like C<\p{Nd}> are known to
1624 be quite a bit slower (5-20 times) than their simpler counterparts
1625 like C<\d> (then again, there 268 Unicode characters matching C<Nd>
1626 compared with the 10 ASCII characters matching C<d>).
1628 =head2 Problems on EBCDIC platforms
1630 There are a number of known problems with Perl on EBCDIC platforms. If you
1631 want to use Perl there, send email to perlbug@perl.org.
1633 In earlier versions, when byte and character data were concatenated,
1634 the new string was sometimes created by
1635 decoding the byte strings as I<ISO 8859-1 (Latin-1)>, even if the
1636 old Unicode string used EBCDIC.
1638 If you find any of these, please report them as bugs.
1640 =head2 Porting code from perl-5.6.X
1642 Perl 5.8 has a different Unicode model from 5.6. In 5.6 the programmer
1643 was required to use the C<utf8> pragma to declare that a given scope
1644 expected to deal with Unicode data and had to make sure that only
1645 Unicode data were reaching that scope. If you have code that is
1646 working with 5.6, you will need some of the following adjustments to
1647 your code. The examples are written such that the code will continue
1648 to work under 5.6, so you should be safe to try them out.
1654 A filehandle that should read or write UTF-8
1657 binmode $fh, ":encoding(utf8)";
1662 A scalar that is going to be passed to some extension
1664 Be it Compress::Zlib, Apache::Request or any extension that has no
1665 mention of Unicode in the manpage, you need to make sure that the
1666 UTF8 flag is stripped off. Note that at the time of this writing
1667 (October 2002) the mentioned modules are not UTF-8-aware. Please
1668 check the documentation to verify if this is still true.
1672 $val = Encode::encode_utf8($val); # make octets
1677 A scalar we got back from an extension
1679 If you believe the scalar comes back as UTF-8, you will most likely
1680 want the UTF8 flag restored:
1684 $val = Encode::decode_utf8($val);
1689 Same thing, if you are really sure it is UTF-8
1693 Encode::_utf8_on($val);
1698 A wrapper for fetchrow_array and fetchrow_hashref
1700 When the database contains only UTF-8, a wrapper function or method is
1701 a convenient way to replace all your fetchrow_array and
1702 fetchrow_hashref calls. A wrapper function will also make it easier to
1703 adapt to future enhancements in your database driver. Note that at the
1704 time of this writing (October 2002), the DBI has no standardized way
1705 to deal with UTF-8 data. Please check the documentation to verify if
1709 my($self, $sth, $what) = @_; # $what is one of fetchrow_{array,hashref}
1715 my @arr = $sth->$what;
1717 defined && /[^\000-\177]/ && Encode::_utf8_on($_);
1721 my $ret = $sth->$what;
1723 for my $k (keys %$ret) {
1724 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret->{$k};
1728 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret;
1738 A large scalar that you know can only contain ASCII
1740 Scalars that contain only ASCII and are marked as UTF-8 are sometimes
1741 a drag to your program. If you recognize such a situation, just remove
1744 utf8::downgrade($val) if $] > 5.007;
1750 L<perlunitut>, L<perluniintro>, L<perluniprops>, L<Encode>, L<open>, L<utf8>, L<bytes>,
1751 L<perlretut>, L<perlvar/"${^UNICODE}">
1752 L<http://www.unicode.org/reports/tr44>).