3 perlunicode - Unicode support in Perl
7 =head2 Important Caveats
9 Unicode support is an extensive requirement. While Perl does not
10 implement the Unicode standard or the accompanying technical reports
11 from cover to cover, Perl does support many Unicode features.
13 People who want to learn to use Unicode in Perl, should probably read
14 the L<Perl Unicode tutorial, perlunitut|perlunitut> and
15 L<perluniintro>, before reading
16 this reference document.
18 Also, the use of Unicode may present security issues that aren't obvious.
19 Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>.
23 =item Safest if you C<use feature 'unicode_strings'>
25 In order to preserve backward compatibility, Perl does not turn
26 on full internal Unicode support unless the pragma
27 C<use feature 'unicode_strings'> is specified. (This is automatically
28 selected if you use C<use 5.012> or higher.) Failure to do this can
29 trigger unexpected surprises. See L</The "Unicode Bug"> below.
31 This pragma doesn't affect I/O. Nor does it change the internal
32 representation of strings, only their interpretation. There are still
33 several places where Unicode isn't fully supported, such as in
36 =item Input and Output Layers
38 Perl knows when a filehandle uses Perl's internal Unicode encodings
39 (UTF-8, or UTF-EBCDIC if in EBCDIC) if the filehandle is opened with
40 the C<:encoding(utf8)> layer. Other encodings can be converted to Perl's
41 encoding on input or from Perl's encoding on output by use of the
42 C<:encoding(...)> layer. See L<open>.
44 To indicate that Perl source itself is in UTF-8, use C<use utf8;>.
46 =item C<use utf8> still needed to enable UTF-8/UTF-EBCDIC in scripts
48 As a compatibility measure, the C<use utf8> pragma must be explicitly
49 included to enable recognition of UTF-8 in the Perl scripts themselves
50 (in string or regular expression literals, or in identifier names) on
51 ASCII-based machines or to recognize UTF-EBCDIC on EBCDIC-based
52 machines. B<These are the only times when an explicit C<use utf8>
53 is needed.> See L<utf8>.
55 =item C<BOM>-marked scripts and UTF-16 scripts autodetected
57 If a Perl script begins marked with the Unicode C<BOM> (UTF-16LE, UTF16-BE,
58 or UTF-8), or if the script looks like non-C<BOM>-marked UTF-16 of either
59 endianness, Perl will correctly read in the script as Unicode.
60 (C<BOM>less UTF-8 cannot be effectively recognized or differentiated from
61 ISO 8859-1 or other eight-bit encodings.)
63 =item C<use encoding> needed to upgrade non-Latin-1 byte strings
65 By default, there is a fundamental asymmetry in Perl's Unicode model:
66 implicit upgrading from byte strings to Unicode strings assumes that
67 they were encoded in I<ISO 8859-1 (Latin-1)>, but Unicode strings are
68 downgraded with UTF-8 encoding. This happens because the first 256
69 codepoints in Unicode happens to agree with Latin-1.
71 See L</"Byte and Character Semantics"> for more details.
75 =head2 Byte and Character Semantics
77 Perl uses logically-wide characters to represent strings internally.
79 Starting in Perl 5.14, Perl-level operations work with
80 characters rather than bytes within the scope of a
81 C<L<use feature 'unicode_strings'|feature>> (or equivalently
82 C<use 5.012> or higher). (This is not true if bytes have been
83 explicitly requested by C<L<use bytes|bytes>>, nor necessarily true
84 for interactions with the platform's operating system.)
86 For earlier Perls, and when C<unicode_strings> is not in effect, Perl
87 provides a fairly safe environment that can handle both types of
88 semantics in programs. For operations where Perl can unambiguously
89 decide that the input data are characters, Perl switches to character
90 semantics. For operations where this determination cannot be made
91 without additional information from the user, Perl decides in favor of
92 compatibility and chooses to use byte semantics.
94 When C<use locale> (but not C<use locale ':not_characters'>) is in
95 effect, Perl uses the rules associated with the current locale.
96 (C<use locale> overrides C<use feature 'unicode_strings'> in the same scope;
97 while C<use locale ':not_characters'> effectively also selects
98 C<use feature 'unicode_strings'> in its scope; see L<perllocale>.)
99 Otherwise, Perl uses the platform's native
100 byte semantics for characters whose code points are less than 256, and
101 Unicode rules for those greater than 255. That means that non-ASCII
102 characters are undefined except for their
103 ordinal numbers. This means that none have case (upper and lower), nor are any
104 a member of character classes, like C<[:alpha:]> or C<\w>. (But all do belong
105 to the C<\W> class or the Perl regular expression extension C<[:^alpha:]>.)
107 This behavior preserves compatibility with earlier versions of Perl,
108 which allowed byte semantics in Perl operations only if
109 none of the program's inputs were marked as being a source of Unicode
110 character data. Such data may come from filehandles, from calls to
111 external programs, from information provided by the system (such as C<%ENV>),
112 or from literals and constants in the source text.
114 The C<utf8> pragma is primarily a compatibility device that enables
115 recognition of UTF-(8|EBCDIC) in literals encountered by the parser.
116 Note that this pragma is only required while Perl defaults to byte
117 semantics; when character semantics become the default, this pragma
118 may become a no-op. See L<utf8>.
120 If strings operating under byte semantics and strings with Unicode
121 character data are concatenated, the new string will have
122 character semantics. This can cause surprises: See L</BUGS>, below.
123 You can choose to be warned when this happens. See C<L<encoding::warnings>>.
125 Under character semantics, many operations that formerly operated on
126 bytes now operate on characters. A character in Perl is
127 logically just a number ranging from 0 to 2**31 or so. Larger
128 characters may encode into longer sequences of bytes internally, but
129 this internal detail is mostly hidden for Perl code.
130 See L<perluniintro> for more.
132 =head2 Effects of Character Semantics
134 Character semantics have the following effects:
140 Strings--including hash keys--and regular expression patterns may
141 contain characters that have an ordinal value larger than 255.
143 If you use a Unicode editor to edit your program, Unicode characters may
144 occur directly within the literal strings in UTF-8 encoding, or UTF-16.
145 (The former requires a C<BOM> or C<use utf8>, the latter requires a C<BOM>.)
147 Unicode characters can also be added to a string by using the C<\N{U+...}>
148 notation. The Unicode code for the desired character, in hexadecimal,
149 should be placed in the braces, after the C<U>. For instance, a smiley face is
152 Alternatively, you can use the C<\x{...}> notation for characters C<0x100> and
153 above. For characters below C<0x100> you may get byte semantics instead of
154 character semantics; see L</The "Unicode Bug">. On EBCDIC machines there is
155 the additional problem that the value for such characters gives the EBCDIC
156 character rather than the Unicode one, thus it is more portable to use
157 C<\N{U+...}> instead.
159 Additionally, you can use the C<\N{...}> notation and put the official
160 Unicode character name within the braces, such as
161 C<\N{WHITE SMILING FACE}>. This automatically loads the L<charnames>
162 module with the C<:full> and C<:short> options. If you prefer different
163 options for this module, you can instead, before the C<\N{...}>,
164 explicitly load it with your desired options; for example,
166 use charnames ':loose';
170 If an appropriate L<encoding> is specified, identifiers within the
171 Perl script may contain Unicode alphanumeric characters, including
172 ideographs. Perl does not currently attempt to canonicalize variable
177 Regular expressions match characters instead of bytes. C<"."> matches
178 a character instead of a byte.
182 Bracketed character classes in regular expressions match characters instead of
183 bytes and match against the character properties specified in the
184 Unicode properties database. C<\w> can be used to match a Japanese
185 ideograph, for instance.
189 Named Unicode properties, scripts, and block ranges may be used (like bracketed
190 character classes) by using the C<\p{}> "matches property" construct and
191 the C<\P{}> negation, "doesn't match property".
192 See L</"Unicode Character Properties"> for more details.
194 You can define your own character properties and use them
195 in the regular expression with the C<\p{}> or C<\P{}> construct.
196 See L</"User-Defined Character Properties"> for more details.
200 The special pattern C<\X> matches a logical character, an "extended grapheme
201 cluster" in Standardese. In Unicode what appears to the user to be a single
202 character, for example an accented C<G>, may in fact be composed of a sequence
203 of characters, in this case a C<G> followed by an accent character. C<\X>
204 will match the entire sequence.
208 The C<tr///> operator translates characters instead of bytes. Note
209 that the C<tr///CU> functionality has been removed. For similar
210 functionality see pack('U0', ...) and pack('C0', ...).
214 Case translation operators use the Unicode case translation tables
215 when character input is provided. Note that C<uc()>, or C<\U> in
216 interpolated strings, translates to uppercase, while C<ucfirst>,
217 or C<\u> in interpolated strings, translates to titlecase in languages
218 that make the distinction (which is equivalent to uppercase in languages
219 without the distinction).
223 Most operators that deal with positions or lengths in a string will
224 automatically switch to using character positions, including
225 C<chop()>, C<chomp()>, C<substr()>, C<pos()>, C<index()>, C<rindex()>,
226 C<sprintf()>, C<write()>, and C<length()>. An operator that
227 specifically does not switch is C<vec()>. Operators that really don't
228 care include operators that treat strings as a bucket of bits such as
229 C<sort()>, and operators dealing with filenames.
233 The C<pack()>/C<unpack()> letter C<C> does I<not> change, since it is often
234 used for byte-oriented formats. Again, think C<char> in the C language.
236 There is a new C<U> specifier that converts between Unicode characters
237 and code points. There is also a C<W> specifier that is the equivalent of
238 C<chr>/C<ord> and properly handles character values even if they are above 255.
242 The C<chr()> and C<ord()> functions work on characters, similar to
243 C<pack("W")> and C<unpack("W")>, I<not> C<pack("C")> and
244 C<unpack("C")>. C<pack("C")> and C<unpack("C")> are methods for
245 emulating byte-oriented C<chr()> and C<ord()> on Unicode strings.
246 While these methods reveal the internal encoding of Unicode strings,
247 that is not something one normally needs to care about at all.
251 The bit string operators, C<& | ^ ~>, can operate on character data.
252 However, for backward compatibility, such as when using bit string
253 operations when characters are all less than 256 in ordinal value, one
254 should not use C<~> (the bit complement) with characters of both
255 values less than 256 and values greater than 256. Most importantly,
256 DeMorgan's laws (C<~($x|$y) eq ~$x&~$y> and C<~($x&$y) eq ~$x|~$y>)
257 will not hold. The reason for this mathematical I<faux pas> is that
258 the complement cannot return B<both> the 8-bit (byte-wide) bit
259 complement B<and> the full character-wide bit complement.
263 There is a CPAN module, C<L<Unicode::Casing>>, which allows you to define
264 your own mappings to be used in C<lc()>, C<lcfirst()>, C<uc()>,
265 C<ucfirst()>, and C<fc> (or their double-quoted string inlined
266 versions such as C<\U>).
267 (Prior to Perl 5.16, this functionality was partially provided
268 in the Perl core, but suffered from a number of insurmountable
269 drawbacks, so the CPAN module was written instead.)
277 And finally, C<scalar reverse()> reverses by character rather than by byte.
281 =head2 Unicode Character Properties
283 (The only time that Perl considers a sequence of individual code
284 points as a single logical character is in the C<\X> construct, already
285 mentioned above. Therefore "character" in this discussion means a single
288 Very nearly all Unicode character properties are accessible through
289 regular expressions by using the C<\p{}> "matches property" construct
290 and the C<\P{}> "doesn't match property" for its negation.
292 For instance, C<\p{Uppercase}> matches any single character with the Unicode
293 C<"Uppercase"> property, while C<\p{L}> matches any character with a
294 C<General_Category> of C<"L"> (letter) property (see
295 L</General_Category> below). Brackets are not
296 required for single letter property names, so C<\p{L}> is equivalent to C<\pL>.
298 More formally, C<\p{Uppercase}> matches any single character whose Unicode
299 C<Uppercase> property value is C<True>, and C<\P{Uppercase}> matches any character
300 whose C<Uppercase> property value is C<False>, and they could have been written as
301 C<\p{Uppercase=True}> and C<\p{Uppercase=False}>, respectively.
303 This formality is needed when properties are not binary; that is, if they can
304 take on more values than just C<True> and C<False>. For example, the
305 C<Bidi_Class> property (see L</"Bidirectional Character Types"> below),
306 can take on several different
307 values, such as C<Left>, C<Right>, C<Whitespace>, and others. To match these, one needs
308 to specify both the property name (C<Bidi_Class>), AND the value being
310 (C<Left>, C<Right>, etc.). This is done, as in the examples above, by having the
311 two components separated by an equal sign (or interchangeably, a colon), like
312 C<\p{Bidi_Class: Left}>.
314 All Unicode-defined character properties may be written in these compound forms
315 of C<\p{I<property>=I<value>}> or C<\p{I<property>:I<value>}>, but Perl provides some
316 additional properties that are written only in the single form, as well as
317 single-form short-cuts for all binary properties and certain others described
318 below, in which you may omit the property name and the equals or colon
321 Most Unicode character properties have at least two synonyms (or aliases if you
322 prefer): a short one that is easier to type and a longer one that is more
323 descriptive and hence easier to understand. Thus the C<"L"> and
324 C<"Letter"> properties above are equivalent and can be used
325 interchangeably. Likewise, C<"Upper"> is a synonym for C<"Uppercase">,
326 and we could have written C<\p{Uppercase}> equivalently as C<\p{Upper}>.
327 Also, there are typically various synonyms for the values the property
328 can be. For binary properties, C<"True"> has 3 synonyms: C<"T">,
329 C<"Yes">, and C<"Y">; and C<"False"> has correspondingly C<"F">,
330 C<"No">, and C<"N">. But be careful. A short form of a value for one
331 property may not mean the same thing as the same short form for another.
332 Thus, for the C<L</General_Category>> property, C<"L"> means
333 C<"Letter">, but for the L<C<Bidi_Class>|/Bidirectional Character Types>
334 property, C<"L"> means C<"Left">. A complete list of properties and
335 synonyms is in L<perluniprops>.
337 Upper/lower case differences in property names and values are irrelevant;
338 thus C<\p{Upper}> means the same thing as C<\p{upper}> or even C<\p{UpPeR}>.
339 Similarly, you can add or subtract underscores anywhere in the middle of a
340 word, so that these are also equivalent to C<\p{U_p_p_e_r}>. And white space
341 is irrelevant adjacent to non-word characters, such as the braces and the equals
342 or colon separators, so C<\p{ Upper }> and C<\p{ Upper_case : Y }> are
343 equivalent to these as well. In fact, white space and even
344 hyphens can usually be added or deleted anywhere. So even C<\p{ Up-per case = Yes}> is
345 equivalent. All this is called "loose-matching" by Unicode. The few places
346 where stricter matching is used is in the middle of numbers, and in the Perl
347 extension properties that begin or end with an underscore. Stricter matching
348 cares about white space (except adjacent to non-word characters),
349 hyphens, and non-interior underscores.
351 You can also use negation in both C<\p{}> and C<\P{}> by introducing a caret
352 (C<^>) between the first brace and the property name: C<\p{^Tamil}> is
353 equal to C<\P{Tamil}>.
355 Almost all properties are immune to case-insensitive matching. That is,
356 adding a C</i> regular expression modifier does not change what they
357 match. There are two sets that are affected.
361 and C<Titlecase_Letter>,
362 all of which match C<Cased_Letter> under C</i> matching.
363 And the second set is
367 all of which match C<Cased> under C</i> matching.
368 This set also includes its subsets C<PosixUpper> and C<PosixLower> both
369 of which under C</i> match C<PosixAlpha>.
370 (The difference between these sets is that some things, such as Roman
371 numerals, come in both upper and lower case so they are C<Cased>, but aren't considered
372 letters, so they aren't C<Cased_Letter>s.)
374 See L</Beyond Unicode code points> for special considerations when
375 matching Unicode properties against non-Unicode code points.
377 =head3 B<General_Category>
379 Every Unicode character is assigned a general category, which is the "most
380 usual categorization of a character" (from
381 L<http://www.unicode.org/reports/tr44>).
383 The compound way of writing these is like C<\p{General_Category=Number}>
384 (short, C<\p{gc:n}>). But Perl furnishes shortcuts in which everything up
385 through the equal or colon separator is omitted. So you can instead just write
388 Here are the short and long forms of the values the C<General Category> property
394 LC, L& Cased_Letter (that is: [\p{Ll}\p{Lu}\p{Lt}])
407 Nd Decimal_Number (also Digit)
411 P Punctuation (also Punct)
412 Pc Connector_Punctuation
416 Pi Initial_Punctuation
417 (may behave like Ps or Pe depending on usage)
419 (may behave like Ps or Pe depending on usage)
431 Zp Paragraph_Separator
434 Cc Control (also Cntrl)
440 Single-letter properties match all characters in any of the
441 two-letter sub-properties starting with the same letter.
442 C<LC> and C<L&> are special: both are aliases for the set consisting of everything matched by C<Ll>, C<Lu>, and C<Lt>.
444 =head3 B<Bidirectional Character Types>
446 Because scripts differ in their directionality (Hebrew and Arabic are
447 written right to left, for example) Unicode supplies a C<Bidi_Class> property.
448 Some of the values this property can have are:
453 LRE Left-to-Right Embedding
454 LRO Left-to-Right Override
457 RLE Right-to-Left Embedding
458 RLO Right-to-Left Override
459 PDF Pop Directional Format
461 ES European Separator
462 ET European Terminator
467 B Paragraph Separator
472 This property is always written in the compound form.
473 For example, C<\p{Bidi_Class:R}> matches characters that are normally
474 written right to left. Unlike the
475 C<L</General_Category>> property, this
476 property can have more values added in a future Unicode release. Those
477 listed above comprised the complete set for many Unicode releases, but
478 others were added in Unicode 6.3; you can always find what the
479 current ones are in in L<perluniprops>. And
480 L<http://www.unicode.org/reports/tr9/> describes how to use them.
484 The world's languages are written in many different scripts. This sentence
485 (unless you're reading it in translation) is written in Latin, while Russian is
486 written in Cyrillic, and Greek is written in, well, Greek; Japanese mainly in
487 Hiragana or Katakana. There are many more.
489 The Unicode Script and Script_Extensions properties give what script a
490 given character is in. Either property can be specified with the
492 C<\p{Script=Hebrew}> (short: C<\p{sc=hebr}>), or
493 C<\p{Script_Extensions=Javanese}> (short: C<\p{scx=java}>).
494 In addition, Perl furnishes shortcuts for all
495 C<Script> property names. You can omit everything up through the equals
496 (or colon), and simply write C<\p{Latin}> or C<\P{Cyrillic}>.
497 (This is not true for C<Script_Extensions>, which is required to be
498 written in the compound form.)
500 The difference between these two properties involves characters that are
501 used in multiple scripts. For example the digits '0' through '9' are
502 used in many parts of the world. These are placed in a script named
503 C<Common>. Other characters are used in just a few scripts. For
504 example, the C<"KATAKANA-HIRAGANA DOUBLE HYPHEN"> is used in both Japanese
505 scripts, Katakana and Hiragana, but nowhere else. The C<Script>
506 property places all characters that are used in multiple scripts in the
507 C<Common> script, while the C<Script_Extensions> property places those
508 that are used in only a few scripts into each of those scripts; while
509 still using C<Common> for those used in many scripts. Thus both these
512 "0" =~ /\p{sc=Common}/ # Matches
513 "0" =~ /\p{scx=Common}/ # Matches
515 and only the first of these match:
517 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{sc=Common} # Matches
518 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{scx=Common} # No match
520 And only the last two of these match:
522 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{sc=Hiragana} # No match
523 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{sc=Katakana} # No match
524 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{scx=Hiragana} # Matches
525 "\N{KATAKANA-HIRAGANA DOUBLE HYPHEN}" =~ /\p{scx=Katakana} # Matches
527 C<Script_Extensions> is thus an improved C<Script>, in which there are
528 fewer characters in the C<Common> script, and correspondingly more in
529 other scripts. It is new in Unicode version 6.0, and its data are likely
530 to change significantly in later releases, as things get sorted out.
532 (Actually, besides C<Common>, the C<Inherited> script, contains
533 characters that are used in multiple scripts. These are modifier
534 characters which modify other characters, and inherit the script value
535 of the controlling character. Some of these are used in many scripts,
536 and so go into C<Inherited> in both C<Script> and C<Script_Extensions>.
537 Others are used in just a few scripts, so are in C<Inherited> in
538 C<Script>, but not in C<Script_Extensions>.)
540 It is worth stressing that there are several different sets of digits in
541 Unicode that are equivalent to 0-9 and are matchable by C<\d> in a
542 regular expression. If they are used in a single language only, they
543 are in that language's C<Script> and C<Script_Extension>. If they are
544 used in more than one script, they will be in C<sc=Common>, but only
545 if they are used in many scripts should they be in C<scx=Common>.
547 A complete list of scripts and their shortcuts is in L<perluniprops>.
549 =head3 B<Use of the C<"Is"> Prefix>
551 For backward compatibility (with Perl 5.6), all properties mentioned
552 so far may have C<Is> or C<Is_> prepended to their name, so C<\P{Is_Lu}>, for
553 example, is equal to C<\P{Lu}>, and C<\p{IsScript:Arabic}> is equal to
558 In addition to B<scripts>, Unicode also defines B<blocks> of
559 characters. The difference between scripts and blocks is that the
560 concept of scripts is closer to natural languages, while the concept
561 of blocks is more of an artificial grouping based on groups of Unicode
562 characters with consecutive ordinal values. For example, the C<"Basic Latin">
563 block is all characters whose ordinals are between 0 and 127, inclusive; in
564 other words, the ASCII characters. The C<"Latin"> script contains some letters
565 from this as well as several other blocks, like C<"Latin-1 Supplement">,
566 C<"Latin Extended-A">, etc., but it does not contain all the characters from
567 those blocks. It does not, for example, contain the digits 0-9, because
568 those digits are shared across many scripts, and hence are in the
571 For more about scripts versus blocks, see UAX#24 "Unicode Script Property":
572 L<http://www.unicode.org/reports/tr24>
574 The C<Script> or C<Script_Extensions> properties are likely to be the
575 ones you want to use when processing
576 natural language; the C<Block> property may occasionally be useful in working
577 with the nuts and bolts of Unicode.
579 Block names are matched in the compound form, like C<\p{Block: Arrows}> or
580 C<\p{Blk=Hebrew}>. Unlike most other properties, only a few block names have a
581 Unicode-defined short name. But Perl does provide a (slight) shortcut: You
582 can say, for example C<\p{In_Arrows}> or C<\p{In_Hebrew}>. For backwards
583 compatibility, the C<In> prefix may be omitted if there is no naming conflict
584 with a script or any other property, and you can even use an C<Is> prefix
585 instead in those cases. But it is not a good idea to do this, for a couple
592 It is confusing. There are many naming conflicts, and you may forget some.
593 For example, C<\p{Hebrew}> means the I<script> Hebrew, and NOT the I<block>
594 Hebrew. But would you remember that 6 months from now?
598 It is unstable. A new version of Unicode may preempt the current meaning by
599 creating a property with the same name. There was a time in very early Unicode
600 releases when C<\p{Hebrew}> would have matched the I<block> Hebrew; now it
605 Some people prefer to always use C<\p{Block: foo}> and C<\p{Script: bar}>
606 instead of the shortcuts, whether for clarity, because they can't remember the
607 difference between 'In' and 'Is' anyway, or they aren't confident that those who
608 eventually will read their code will know that difference.
610 A complete list of blocks and their shortcuts is in L<perluniprops>.
612 =head3 B<Other Properties>
614 There are many more properties than the very basic ones described here.
615 A complete list is in L<perluniprops>.
617 Unicode defines all its properties in the compound form, so all single-form
618 properties are Perl extensions. Most of these are just synonyms for the
619 Unicode ones, but some are genuine extensions, including several that are in
620 the compound form. And quite a few of these are actually recommended by Unicode
621 (in L<http://www.unicode.org/reports/tr18>).
623 This section gives some details on all extensions that aren't just
624 synonyms for compound-form Unicode properties
625 (for those properties, you'll have to refer to the
626 L<Unicode Standard|http://www.unicode.org/reports/tr44>.
632 This matches every possible code point. It is equivalent to C<qr/./s>.
633 Unlike all the other non-user-defined C<\p{}> property matches, no
634 warning is ever generated if this is property is matched against a
635 non-Unicode code point (see L</Beyond Unicode code points> below).
637 =item B<C<\p{Alnum}>>
639 This matches any C<\p{Alphabetic}> or C<\p{Decimal_Number}> character.
643 This matches any of the 1_114_112 Unicode code points. It is a synonym
646 =item B<C<\p{ASCII}>>
648 This matches any of the 128 characters in the US-ASCII character set,
649 which is a subset of Unicode.
651 =item B<C<\p{Assigned}>>
653 This matches any assigned code point; that is, any code point whose L<general
654 category|/General_Category> is not C<Unassigned> (or equivalently, not C<Cn>).
656 =item B<C<\p{Blank}>>
658 This is the same as C<\h> and C<\p{HorizSpace}>: A character that changes the
659 spacing horizontally.
661 =item B<C<\p{Decomposition_Type: Non_Canonical}>> (Short: C<\p{Dt=NonCanon}>)
663 Matches a character that has a non-canonical decomposition.
665 To understand the use of this rarely used I<property=value> combination, it is
666 necessary to know some basics about decomposition.
667 Consider a character, say H. It could appear with various marks around it,
668 such as an acute accent, or a circumflex, or various hooks, circles, arrows,
669 I<etc.>, above, below, to one side or the other, etc. There are many
670 possibilities among the world's languages. The number of combinations is
671 astronomical, and if there were a character for each combination, it would
672 soon exhaust Unicode's more than a million possible characters. So Unicode
673 took a different approach: there is a character for the base H, and a
674 character for each of the possible marks, and these can be variously combined
675 to get a final logical character. So a logical character--what appears to be a
676 single character--can be a sequence of more than one individual characters.
677 This is called an "extended grapheme cluster"; Perl furnishes the C<\X>
678 regular expression construct to match such sequences.
680 But Unicode's intent is to unify the existing character set standards and
681 practices, and several pre-existing standards have single characters that
682 mean the same thing as some of these combinations. An example is ISO-8859-1,
683 which has quite a few of these in the Latin-1 range, an example being C<"LATIN
684 CAPITAL LETTER E WITH ACUTE">. Because this character was in this pre-existing
685 standard, Unicode added it to its repertoire. But this character is considered
686 by Unicode to be equivalent to the sequence consisting of the character
687 C<"LATIN CAPITAL LETTER E"> followed by the character C<"COMBINING ACUTE ACCENT">.
689 C<"LATIN CAPITAL LETTER E WITH ACUTE"> is called a "pre-composed" character, and
690 its equivalence with the sequence is called canonical equivalence. All
691 pre-composed characters are said to have a decomposition (into the equivalent
692 sequence), and the decomposition type is also called canonical.
694 However, many more characters have a different type of decomposition, a
695 "compatible" or "non-canonical" decomposition. The sequences that form these
696 decompositions are not considered canonically equivalent to the pre-composed
697 character. An example, again in the Latin-1 range, is the C<"SUPERSCRIPT ONE">.
698 It is somewhat like a regular digit 1, but not exactly; its decomposition
699 into the digit 1 is called a "compatible" decomposition, specifically a
700 "super" decomposition. There are several such compatibility
701 decompositions (see L<http://www.unicode.org/reports/tr44>), including one
702 called "compat", which means some miscellaneous type of decomposition
703 that doesn't fit into the decomposition categories that Unicode has chosen.
705 Note that most Unicode characters don't have a decomposition, so their
706 decomposition type is C<"None">.
708 For your convenience, Perl has added the C<Non_Canonical> decomposition
709 type to mean any of the several compatibility decompositions.
711 =item B<C<\p{Graph}>>
713 Matches any character that is graphic. Theoretically, this means a character
714 that on a printer would cause ink to be used.
716 =item B<C<\p{HorizSpace}>>
718 This is the same as C<\h> and C<\p{Blank}>: a character that changes the
719 spacing horizontally.
723 This is a synonym for C<\p{Present_In=*}>
725 =item B<C<\p{PerlSpace}>>
727 This is the same as C<\s>, restricted to ASCII, namely C<S<[ \f\n\r\t]>>
728 and starting in Perl v5.18, experimentally, a vertical tab.
730 Mnemonic: Perl's (original) space
732 =item B<C<\p{PerlWord}>>
734 This is the same as C<\w>, restricted to ASCII, namely C<[A-Za-z0-9_]>
736 Mnemonic: Perl's (original) word.
738 =item B<C<\p{Posix...}>>
740 There are several of these, which are equivalents using the C<\p{}>
741 notation for Posix classes and are described in
742 L<perlrecharclass/POSIX Character Classes>.
744 =item B<C<\p{Present_In: *}>> (Short: C<\p{In=*}>)
746 This property is used when you need to know in what Unicode version(s) a
749 The "*" above stands for some two digit Unicode version number, such as
750 C<1.1> or C<4.0>; or the "*" can also be C<Unassigned>. This property will
751 match the code points whose final disposition has been settled as of the
752 Unicode release given by the version number; C<\p{Present_In: Unassigned}>
753 will match those code points whose meaning has yet to be assigned.
755 For example, C<U+0041> C<"LATIN CAPITAL LETTER A"> was present in the very first
756 Unicode release available, which is C<1.1>, so this property is true for all
757 valid "*" versions. On the other hand, C<U+1EFF> was not assigned until version
758 5.1 when it became C<"LATIN SMALL LETTER Y WITH LOOP">, so the only "*" that
759 would match it are 5.1, 5.2, and later.
761 Unicode furnishes the C<Age> property from which this is derived. The problem
762 with Age is that a strict interpretation of it (which Perl takes) has it
763 matching the precise release a code point's meaning is introduced in. Thus
764 C<U+0041> would match only 1.1; and C<U+1EFF> only 5.1. This is not usually what
767 Some non-Perl implementations of the Age property may change its meaning to be
768 the same as the Perl C<Present_In> property; just be aware of that.
770 Another confusion with both these properties is that the definition is not
771 that the code point has been I<assigned>, but that the meaning of the code point
772 has been I<determined>. This is because 66 code points will always be
773 unassigned, and so the C<Age> for them is the Unicode version in which the decision
774 to make them so was made. For example, C<U+FDD0> is to be permanently
775 unassigned to a character, and the decision to do that was made in version 3.1,
776 so C<\p{Age=3.1}> matches this character, as also does C<\p{Present_In: 3.1}> and up.
778 =item B<C<\p{Print}>>
780 This matches any character that is graphical or blank, except controls.
782 =item B<C<\p{SpacePerl}>>
784 This is the same as C<\s>, including beyond ASCII.
786 Mnemonic: Space, as modified by Perl. (It doesn't include the vertical tab
787 which both the Posix standard and Unicode consider white space.)
789 =item B<C<\p{Title}>> and B<C<\p{Titlecase}>>
791 Under case-sensitive matching, these both match the same code points as
792 C<\p{General Category=Titlecase_Letter}> (C<\p{gc=lt}>). The difference
793 is that under C</i> caseless matching, these match the same as
794 C<\p{Cased}>, whereas C<\p{gc=lt}> matches C<\p{Cased_Letter>).
796 =item B<C<\p{Unicode}>>
798 This matches any of the 1_114_112 Unicode code points.
801 =item B<C<\p{VertSpace}>>
803 This is the same as C<\v>: A character that changes the spacing vertically.
807 This is the same as C<\w>, including over 100_000 characters beyond ASCII.
809 =item B<C<\p{XPosix...}>>
811 There are several of these, which are the standard Posix classes
812 extended to the full Unicode range. They are described in
813 L<perlrecharclass/POSIX Character Classes>.
818 =head2 User-Defined Character Properties
820 You can define your own binary character properties by defining subroutines
821 whose names begin with C<"In"> or C<"Is">. (The experimental feature
822 L<perlre/(?[ ])> provides an alternative which allows more complex
823 definitions.) The subroutines can be defined in any
824 package. The user-defined properties can be used in the regular expression
825 C<\p{}> and C<\P{}> constructs; if you are using a user-defined property from a
826 package other than the one you are in, you must specify its package in the
827 C<\p{}> or C<\P{}> construct.
829 # assuming property Is_Foreign defined in Lang::
830 package main; # property package name required
831 if ($txt =~ /\p{Lang::IsForeign}+/) { ... }
833 package Lang; # property package name not required
834 if ($txt =~ /\p{IsForeign}+/) { ... }
837 Note that the effect is compile-time and immutable once defined.
838 However, the subroutines are passed a single parameter, which is 0 if
839 case-sensitive matching is in effect and non-zero if caseless matching
840 is in effect. The subroutine may return different values depending on
841 the value of the flag, and one set of values will immutably be in effect
842 for all case-sensitive matches, and the other set for all case-insensitive
845 Note that if the regular expression is tainted, then Perl will die rather
846 than calling the subroutine when the name of the subroutine is
847 determined by the tainted data.
849 The subroutines must return a specially-formatted string, with one
850 or more newline-separated lines. Each line must be one of the following:
856 A single hexadecimal number denoting a code point to include.
860 Two hexadecimal numbers separated by horizontal whitespace (space or
861 tabular characters) denoting a range of code points to include.
865 Something to include, prefixed by C<"+">: a built-in character
866 property (prefixed by C<"utf8::">) or a fully qualified (including package
867 name) user-defined character property,
868 to represent all the characters in that property; two hexadecimal code
869 points for a range; or a single hexadecimal code point.
873 Something to exclude, prefixed by C<"-">: an existing character
874 property (prefixed by C<"utf8::">) or a fully qualified (including package
875 name) user-defined character property,
876 to represent all the characters in that property; two hexadecimal code
877 points for a range; or a single hexadecimal code point.
881 Something to negate, prefixed C<"!">: an existing character
882 property (prefixed by C<"utf8::">) or a fully qualified (including package
883 name) user-defined character property,
884 to represent all the characters in that property; two hexadecimal code
885 points for a range; or a single hexadecimal code point.
889 Something to intersect with, prefixed by C<"&">: an existing character
890 property (prefixed by C<"utf8::">) or a fully qualified (including package
891 name) user-defined character property,
892 for all the characters except the characters in the property; two
893 hexadecimal code points for a range; or a single hexadecimal code point.
897 For example, to define a property that covers both the Japanese
898 syllabaries (hiragana and katakana), you can define
907 Imagine that the here-doc end marker is at the beginning of the line.
908 Now you can use C<\p{InKana}> and C<\P{InKana}>.
910 You could also have used the existing block property names:
919 Suppose you wanted to match only the allocated characters,
920 not the raw block ranges: in other words, you want to remove
931 The negation is useful for defining (surprise!) negated classes.
941 This will match all non-Unicode code points, since every one of them is
942 not in Kana. You can use intersection to exclude these, if desired, as
943 this modified example shows:
954 C<&utf8::Any> must be the last line in the definition.
956 Intersection is used generally for getting the common characters matched
957 by two (or more) classes. It's important to remember not to use C<"&"> for
958 the first set; that would be intersecting with nothing, resulting in an
961 Unlike non-user-defined C<\p{}> property matches, no warning is ever
962 generated if these properties are matched against a non-Unicode code
963 point (see L</Beyond Unicode code points> below).
965 =head2 User-Defined Case Mappings (for serious hackers only)
967 B<This feature has been removed as of Perl 5.16.>
968 The CPAN module C<L<Unicode::Casing>> provides better functionality without
969 the drawbacks that this feature had. If you are using a Perl earlier
970 than 5.16, this feature was most fully documented in the 5.14 version of
972 L<http://perldoc.perl.org/5.14.0/perlunicode.html#User-Defined-Case-Mappings-%28for-serious-hackers-only%29>
974 =head2 Character Encodings for Input and Output
978 =head2 Unicode Regular Expression Support Level
980 The following list of Unicode supported features for regular expressions describes
981 all features currently directly supported by core Perl. The references to "Level N"
982 and the section numbers refer to the Unicode Technical Standard #18,
983 "Unicode Regular Expressions", version 13, from August 2008.
989 Level 1 - Basic Unicode Support
991 RL1.1 Hex Notation - done [1]
992 RL1.2 Properties - done [2][3]
993 RL1.2a Compatibility Properties - done [4]
994 RL1.3 Subtraction and Intersection - experimental [5]
995 RL1.4 Simple Word Boundaries - done [6]
996 RL1.5 Simple Loose Matches - done [7]
997 RL1.6 Line Boundaries - MISSING [8][9]
998 RL1.7 Supplementary Code Points - done [10]
1008 C<\p{...}> C<\P{...}>
1012 supports not only minimal list, but all Unicode character properties (see Unicode Character Properties above)
1016 C<\d> C<\D> C<\s> C<\S> C<\w> C<\W> C<\X> C<[:I<prop>:]> C<[:^I<prop>:]>
1020 The experimental feature in v5.18 C<"(?[...])"> accomplishes this. See
1021 L<perlre/(?[ ])>. If you don't want to use an experimental feature,
1022 you can use one of the following:
1026 =item * Regular expression look-ahead
1028 You can mimic class subtraction using lookahead.
1029 For example, what UTS#18 might write as
1031 [{Block=Greek}-[{UNASSIGNED}]]
1033 in Perl can be written as:
1035 (?!\p{Unassigned})\p{Block=Greek}
1036 (?=\p{Assigned})\p{Block=Greek}
1038 But in this particular example, you probably really want
1042 which will match assigned characters known to be part of the Greek script.
1044 =item * CPAN module C<L<Unicode::Regex::Set>>
1046 It does implement the full UTS#18 grouping, intersection, union, and
1047 removal (subtraction) syntax.
1049 =item * L</"User-Defined Character Properties">
1051 C<"+"> for union, C<"-"> for removal (set-difference), C<"&"> for intersection
1061 Note that Perl does Full case-folding in matching (but with bugs), not
1062 Simple: for example C<U+1F88> is equivalent to C<U+1F00 U+03B9>, instead of
1063 just C<U+1F80>. This difference matters mainly for certain Greek capital
1064 letters with certain modifiers: the Full case-folding decomposes the
1065 letter, while the Simple case-folding would map it to a single
1070 Should do C<^> and C<$> also on C<U+000B> (C<\v> in C), C<FF> (C<\f>),
1071 C<CR> (C<\r>), C<CRLF> (C<\r\n>), C<NEL> (C<U+0085>), C<LS> (C<U+2028>),
1072 and C<PS> (C<U+2029>); should also affect C<E<lt>E<gt>>, C<$.>, and
1073 script line numbers; should not split lines within C<CRLF> (i.e. there
1074 is no empty line between C<\r> and C<\n>). For C<CRLF>, try the
1075 C<:crlf> layer (see L<PerlIO>).
1079 Linebreaking conformant with L<UAX#14 "Unicode Line Breaking
1080 Algorithm"|http://www.unicode.org/reports/tr14>
1081 is available through the C<L<Unicode::LineBreak>> module.
1085 UTF-8/UTF-EBDDIC used in Perl allows not only C<U+10000> to
1086 C<U+10FFFF> but also beyond C<U+10FFFF>
1092 Level 2 - Extended Unicode Support
1094 RL2.1 Canonical Equivalents - MISSING [10][11]
1095 RL2.2 Default Grapheme Clusters - MISSING [12]
1096 RL2.3 Default Word Boundaries - MISSING [14]
1097 RL2.4 Default Loose Matches - MISSING [15]
1098 RL2.5 Name Properties - DONE
1099 RL2.6 Wildcard Properties - MISSING
1101 [10] see UAX#15 "Unicode Normalization Forms"
1102 [11] have Unicode::Normalize but not integrated to regexes
1103 [12] have \X but we don't have a "Grapheme Cluster Mode"
1104 [14] see UAX#29, Word Boundaries
1105 [15] This is covered in Chapter 3.13 (in Unicode 6.0)
1109 Level 3 - Tailored Support
1111 RL3.1 Tailored Punctuation - MISSING
1112 RL3.2 Tailored Grapheme Clusters - MISSING [17][18]
1113 RL3.3 Tailored Word Boundaries - MISSING
1114 RL3.4 Tailored Loose Matches - MISSING
1115 RL3.5 Tailored Ranges - MISSING
1116 RL3.6 Context Matching - MISSING [19]
1117 RL3.7 Incremental Matches - MISSING
1118 ( RL3.8 Unicode Set Sharing )
1119 RL3.9 Possible Match Sets - MISSING
1120 RL3.10 Folded Matching - MISSING [20]
1121 RL3.11 Submatchers - MISSING
1123 [17] see UAX#10 "Unicode Collation Algorithms"
1124 [18] have Unicode::Collate but not integrated to regexes
1125 [19] have (?<=x) and (?=x), but look-aheads or look-behinds
1126 should see outside of the target substring
1127 [20] need insensitive matching for linguistic features other
1128 than case; for example, hiragana to katakana, wide and
1129 narrow, simplified Han to traditional Han (see UTR#30
1130 "Character Foldings")
1134 =head2 Unicode Encodings
1136 Unicode characters are assigned to I<code points>, which are abstract
1137 numbers. To use these numbers, various encodings are needed.
1145 UTF-8 is a variable-length (1 to 4 bytes), byte-order independent
1146 encoding. For ASCII (and we really do mean 7-bit ASCII, not another
1147 8-bit encoding), UTF-8 is transparent.
1149 The following table is from Unicode 3.2.
1151 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1153 U+0000..U+007F 00..7F
1154 U+0080..U+07FF * C2..DF 80..BF
1155 U+0800..U+0FFF E0 * A0..BF 80..BF
1156 U+1000..U+CFFF E1..EC 80..BF 80..BF
1157 U+D000..U+D7FF ED 80..9F 80..BF
1158 U+D800..U+DFFF +++++ utf16 surrogates, not legal utf8 +++++
1159 U+E000..U+FFFF EE..EF 80..BF 80..BF
1160 U+10000..U+3FFFF F0 * 90..BF 80..BF 80..BF
1161 U+40000..U+FFFFF F1..F3 80..BF 80..BF 80..BF
1162 U+100000..U+10FFFF F4 80..8F 80..BF 80..BF
1164 Note the gaps marked by "*" before several of the byte entries above. These are
1165 caused by legal UTF-8 avoiding non-shortest encodings: it is technically
1166 possible to UTF-8-encode a single code point in different ways, but that is
1167 explicitly forbidden, and the shortest possible encoding should always be used
1168 (and that is what Perl does).
1170 Another way to look at it is via bits:
1172 Code Points 1st Byte 2nd Byte 3rd Byte 4th Byte
1175 00000bbbbbaaaaaa 110bbbbb 10aaaaaa
1176 ccccbbbbbbaaaaaa 1110cccc 10bbbbbb 10aaaaaa
1177 00000dddccccccbbbbbbaaaaaa 11110ddd 10cccccc 10bbbbbb 10aaaaaa
1179 As you can see, the continuation bytes all begin with C<"10">, and the
1180 leading bits of the start byte tell how many bytes there are in the
1183 The original UTF-8 specification allowed up to 6 bytes, to allow
1184 encoding of numbers up to C<0x7FFF_FFFF>. Perl continues to allow those,
1185 and has extended that up to 13 bytes to encode code points up to what
1186 can fit in a 64-bit word. However, Perl will warn if you output any of
1187 these as being non-portable; and under strict UTF-8 input protocols,
1190 The Unicode non-character code points are also disallowed in UTF-8 in
1191 "open interchange". See L</Non-character code points>.
1197 Like UTF-8 but EBCDIC-safe, in the way that UTF-8 is ASCII-safe.
1201 UTF-16, UTF-16BE, UTF-16LE, Surrogates, and C<BOM>s (Byte Order Marks)
1203 The followings items are mostly for reference and general Unicode
1204 knowledge, Perl doesn't use these constructs internally.
1206 Like UTF-8, UTF-16 is a variable-width encoding, but where
1207 UTF-8 uses 8-bit code units, UTF-16 uses 16-bit code units.
1208 All code points occupy either 2 or 4 bytes in UTF-16: code points
1209 C<U+0000..U+FFFF> are stored in a single 16-bit unit, and code
1210 points C<U+10000..U+10FFFF> in two 16-bit units. The latter case is
1211 using I<surrogates>, the first 16-bit unit being the I<high
1212 surrogate>, and the second being the I<low surrogate>.
1214 Surrogates are code points set aside to encode the C<U+10000..U+10FFFF>
1215 range of Unicode code points in pairs of 16-bit units. The I<high
1216 surrogates> are the range C<U+D800..U+DBFF> and the I<low surrogates>
1217 are the range C<U+DC00..U+DFFF>. The surrogate encoding is
1219 $hi = ($uni - 0x10000) / 0x400 + 0xD800;
1220 $lo = ($uni - 0x10000) % 0x400 + 0xDC00;
1224 $uni = 0x10000 + ($hi - 0xD800) * 0x400 + ($lo - 0xDC00);
1226 Because of the 16-bitness, UTF-16 is byte-order dependent. UTF-16
1227 itself can be used for in-memory computations, but if storage or
1228 transfer is required either UTF-16BE (big-endian) or UTF-16LE
1229 (little-endian) encodings must be chosen.
1231 This introduces another problem: what if you just know that your data
1232 is UTF-16, but you don't know which endianness? Byte Order Marks, or
1233 C<BOM>s, are a solution to this. A special character has been reserved
1234 in Unicode to function as a byte order marker: the character with the
1235 code point C<U+FEFF> is the C<BOM>.
1237 The trick is that if you read a C<BOM>, you will know the byte order,
1238 since if it was written on a big-endian platform, you will read the
1239 bytes C<0xFE 0xFF>, but if it was written on a little-endian platform,
1240 you will read the bytes C<0xFF 0xFE>. (And if the originating platform
1241 was writing in UTF-8, you will read the bytes C<0xEF 0xBB 0xBF>.)
1243 The way this trick works is that the character with the code point
1244 C<U+FFFE> is not supposed to be in input streams, so the
1245 sequence of bytes C<0xFF 0xFE> is unambiguously "C<BOM>, represented in
1246 little-endian format" and cannot be C<U+FFFE>, represented in big-endian
1249 Surrogates have no meaning in Unicode outside their use in pairs to
1250 represent other code points. However, Perl allows them to be
1251 represented individually internally, for example by saying
1252 C<chr(0xD801)>, so that all code points, not just those valid for open
1254 representable. Unicode does define semantics for them, such as their
1255 C<L</General_Category>> is C<"Cs">. But because their use is somewhat dangerous,
1256 Perl will warn (using the warning category C<"surrogate">, which is a
1257 sub-category of C<"utf8">) if an attempt is made
1258 to do things like take the lower case of one, or match
1259 case-insensitively, or to output them. (But don't try this on Perls
1264 UTF-32, UTF-32BE, UTF-32LE
1266 The UTF-32 family is pretty much like the UTF-16 family, expect that
1267 the units are 32-bit, and therefore the surrogate scheme is not
1268 needed. UTF-32 is a fixed-width encoding. The C<BOM> signatures are
1269 C<0x00 0x00 0xFE 0xFF> for BE and C<0xFF 0xFE 0x00 0x00> for LE.
1275 Legacy, fixed-width encodings defined by the ISO 10646 standard. UCS-2 is a 16-bit
1276 encoding. Unlike UTF-16, UCS-2 is not extensible beyond C<U+FFFF>,
1277 because it does not use surrogates. UCS-4 is a 32-bit encoding,
1278 functionally identical to UTF-32 (the difference being that
1279 UCS-4 forbids neither surrogates nor code points larger than C<0x10_FFFF>).
1285 A seven-bit safe (non-eight-bit) encoding, which is useful if the
1286 transport or storage is not eight-bit safe. Defined by RFC 2152.
1290 =head2 Non-character code points
1292 66 code points are set aside in Unicode as "non-character code points".
1293 These all have the C<Unassigned> (C<Cn>) C<L</General_Category>>, and
1295 be assigned. These are never supposed to be in legal Unicode input
1296 streams, so that code can use them as sentinels that can be mixed in
1297 with character data, and they always will be distinguishable from that data.
1298 To keep them out of Perl input streams, strict UTF-8 should be
1299 specified, such as by using the layer C<:encoding('UTF-8')>. The
1300 non-character code points are the 32 between C<U+FDD0> and C<U+FDEF>, and the
1301 34 code points C<U+FFFE>, C<U+FFFF>, C<U+1FFFE>, C<U+1FFFF>, ... C<U+10FFFE>, C<U+10FFFF>.
1302 Some people are under the mistaken impression that these are "illegal",
1303 but that is not true. An application or cooperating set of applications
1304 can legally use them at will internally; but these code points are
1305 "illegal for open interchange". Therefore, Perl will not accept these
1306 from input streams unless lax rules are being used, and will warn
1307 (using the warning category C<"nonchar">, which is a sub-category of C<"utf8">) if
1308 an attempt is made to output them.
1310 =head2 Beyond Unicode code points
1312 The maximum Unicode code point is C<U+10FFFF>, and Unicode only defines
1313 operations on code points up through that. But Perl works on code
1314 points up to the maximum permissible unsigned number available on the
1315 platform. However, Perl will not accept these from input streams unless
1316 lax rules are being used, and will warn (using the warning category
1317 C<"non_unicode">, which is a sub-category of C<"utf8">) if any are output.
1319 Since Unicode rules are not defined on these code points, if a
1320 Unicode-defined operation is done on them, Perl uses what we believe are
1321 sensible rules, while generally warning, using the C<"non_unicode">
1322 category. For example, C<uc("\x{11_0000}")> will generate such a
1323 warning, returning the input parameter as its result, since Perl defines
1324 the uppercase of every non-Unicode code point to be the code point
1325 itself. In fact, all the case changing operations, not just
1326 uppercasing, work this way.
1328 The situation with matching Unicode properties in regular expressions,
1329 the C<\p{}> and C<\P{}> constructs, against these code points is not as
1330 clear cut, and how these are handled has changed as we've gained
1333 One possibility is to treat any match against these code points as
1334 undefined. But since Perl doesn't have the concept of a match being
1335 undefined, it converts this to failing or C<FALSE>. This is almost, but
1336 not quite, what Perl did from v5.14 (when use of these code points
1337 became generally reliable) through v5.18. The difference is that Perl
1338 treated all C<\p{}> matches as failing, but all C<\P{}> matches as
1341 One problem with this is that it leads to unexpected, and confusting
1342 results in some cases:
1344 chr(0x110000) =~ \p{ASCII_Hex_Digit=True} # Failed on <= v5.18
1345 chr(0x110000) =~ \p{ASCII_Hex_Digit=False} # Failed! on <= v5.18
1347 That is, it treated both matches as undefined, and converted that to
1348 false (raising a warning on each). The first case is the expected
1349 result, but the second is likely counterintuitive: "How could both be
1350 false when they are complements?" Another problem was that the
1351 implementation optimized many Unicode property matches down to already
1352 existing simpler, faster operations, which don't raise the warning. We
1353 chose to not forgo those optimizations, which help the vast majority of
1354 matches, just to generate a warning for the unlikely event that an
1355 above-Unicode code point is being matched against.
1357 As a result of these problems, starting in v5.20, what Perl does is
1358 to treat non-Unicode code points as just typical unassigned Unicode
1359 characters, and matches accordingly. (Note: Unicode has atypical
1360 unassigned code points. For example, it has non-character code points,
1361 and ones that, when they do get assigned, are destined to be written
1362 Right-to-left, as Arabic and Hebrew are. Perl assumes that no
1363 non-Unicode code point has any atypical properties.)
1365 Perl, in most cases, will raise a warning when matching an above-Unicode
1366 code point against a Unicode property when the result is C<TRUE> for
1367 C<\p{}>, and C<FALSE> for C<\P{}>. For example:
1369 chr(0x110000) =~ \p{ASCII_Hex_Digit=True} # Fails, no warning
1370 chr(0x110000) =~ \p{ASCII_Hex_Digit=False} # Succeeds, with warning
1372 In both these examples, the character being matched is non-Unicode, so
1373 Unicode doesn't define how it should match. It clearly isn't an ASCII
1374 hex digit, so the first example clearly should fail, and so it does,
1375 with no warning. But it is arguable that the second example should have
1376 an undefined, hence C<FALSE>, result. So a warning is raised for it.
1378 Thus the warning is raised for many fewer cases than in earlier Perls,
1379 and only when what the result is could be arguable. It turns out that
1380 none of the optimizations made by Perl (or are ever likely to be made)
1381 cause the warning to be skipped, so it solves both problems of Perl's
1382 earlier approach. The most commonly used property that is affected by
1383 this change is C<\p{Unassigned}> which is a short form for
1384 C<\p{General_Category=Unassigned}>. Starting in v5.20, all non-Unicode
1385 code points are considered C<Unassigned>. In earlier releases the
1386 matches failed because the result was considered undefined.
1388 The only place where the warning is not raised when it might ought to
1389 have been is if optimizations cause the whole pattern match to not even
1390 be attempted. For example, Perl may figure out that for a string to
1391 match a certain regular expression pattern, the string has to contain
1392 the substring C<"foobar">. Before attempting the match, Perl may look
1393 for that substring, and if not found, immediately fail the match without
1394 actually trying it; so no warning gets generated even if the string
1395 contains an above-Unicode code point.
1397 This behavior is more "Do what I mean" than in earlier Perls for most
1398 applications. But it catches fewer issues for code that needs to be
1399 strictly Unicode compliant. Therefore there is an additional mode of
1400 operation available to accommodate such code. This mode is enabled if a
1401 regular expression pattern is compiled within the lexical scope where
1402 the C<"non_unicode"> warning class has been made fatal, say by:
1404 use warnings FATAL => "non_unicode"
1406 (see L<warnings>). In this mode of operation, Perl will raise the
1407 warning for all matches against a non-Unicode code point (not just the
1408 arguable ones), and it skips the optimizations that might cause the
1409 warning to not be output. (It currently still won't warn if the match
1410 isn't even attempted, like in the C<"foobar"> example above.)
1412 In summary, Perl now normally treats non-Unicode code points as typical
1413 Unicode unassigned code points for regular expression matches, raising a
1414 warning only when it is arguable what the result should be. However, if
1415 this warning has been made fatal, it isn't skipped.
1417 There is one exception to all this. C<\p{All}> looks like a Unicode
1418 property, but it is a Perl extension that is defined to be true for all
1419 possible code points, Unicode or not, so no warning is ever generated
1420 when matching this against a non-Unicode code point. (Prior to v5.20,
1421 it was an exact synonym for C<\p{Any}>, matching code points C<0>
1422 through C<0x10FFFF>.)
1424 =head2 Security Implications of Unicode
1426 Read L<Unicode Security Considerations|http://www.unicode.org/reports/tr36>.
1427 Also, note the following:
1435 Unfortunately, the original specification of UTF-8 leaves some room for
1436 interpretation of how many bytes of encoded output one should generate
1437 from one input Unicode character. Strictly speaking, the shortest
1438 possible sequence of UTF-8 bytes should be generated,
1439 because otherwise there is potential for an input buffer overflow at
1440 the receiving end of a UTF-8 connection. Perl always generates the
1441 shortest length UTF-8, and with warnings on, Perl will warn about
1442 non-shortest length UTF-8 along with other malformations, such as the
1443 surrogates, which are not Unicode code points valid for interchange.
1447 Regular expression pattern matching may surprise you if you're not
1448 accustomed to Unicode. Starting in Perl 5.14, several pattern
1449 modifiers are available to control this, called the character set
1450 modifiers. Details are given in L<perlre/Character set modifiers>.
1454 As discussed elsewhere, Perl has one foot (two hooves?) planted in
1455 each of two worlds: the old world of bytes and the new world of
1456 characters, upgrading from bytes to characters when necessary.
1457 If your legacy code does not explicitly use Unicode, no automatic
1458 switch-over to characters should happen. Characters shouldn't get
1459 downgraded to bytes, either. It is possible to accidentally mix bytes
1460 and characters, however (see L<perluniintro>), in which case C<\w> in
1461 regular expressions might start behaving differently (unless the C</a>
1462 modifier is in effect). Review your code. Use warnings and the C<strict> pragma.
1464 =head2 Unicode in Perl on EBCDIC
1466 The way Unicode is handled on EBCDIC platforms is still
1467 experimental. On such platforms, references to UTF-8 encoding in this
1468 document and elsewhere should be read as meaning the UTF-EBCDIC
1469 specified in Unicode Technical Report 16, unless ASCII vs. EBCDIC issues
1470 are specifically discussed. There is no C<utfebcdic> pragma or
1471 C<":utfebcdic"> layer; rather, C<"utf8"> and C<":utf8"> are reused to mean
1472 the platform's "natural" 8-bit encoding of Unicode. See L<perlebcdic>
1473 for more discussion of the issues.
1477 See L<perllocale/Unicode and UTF-8>
1479 =head2 When Unicode Does Not Happen
1481 While Perl does have extensive ways to input and output in Unicode,
1482 and a few other "entry points" like the C<@ARGV> array (which can sometimes be
1483 interpreted as UTF-8), there are still many places where Unicode
1484 (in some encoding or another) could be given as arguments or received as
1485 results, or both, but it is not.
1487 The following are such interfaces. Also, see L</The "Unicode Bug">.
1488 For all of these interfaces Perl
1489 currently (as of v5.16.0) simply assumes byte strings both as arguments
1490 and results, or UTF-8 strings if the (problematic) C<encoding> pragma has been used.
1492 One reason that Perl does not attempt to resolve the role of Unicode in
1493 these situations is that the answers are highly dependent on the operating
1494 system and the file system(s). For example, whether filenames can be
1495 in Unicode and in exactly what kind of encoding, is not exactly a
1496 portable concept. Similarly for C<qx> and C<system>: how well will the
1497 "command-line interface" (and which of them?) handle Unicode?
1503 C<chdir>, C<chmod>, C<chown>, C<chroot>, C<exec>, C<link>, C<lstat>, C<mkdir>,
1504 C<rename>, C<rmdir>, C<stat>, C<symlink>, C<truncate>, C<unlink>, C<utime>, C<-X>
1512 C<glob> (aka the C<E<lt>*E<gt>>)
1516 C<open>, C<opendir>, C<sysopen>
1520 C<qx> (aka the backtick operator), C<system>
1524 C<readdir>, C<readlink>
1528 =head2 The "Unicode Bug"
1530 The term, "Unicode bug" has been applied to an inconsistency
1531 on ASCII platforms with the
1532 Unicode code points in the C<Latin-1 Supplement> block, that
1533 is, between 128 and 255. Without a locale specified, unlike all other
1534 characters or code points, these characters have very different semantics in
1535 byte semantics versus character semantics, unless
1536 C<use feature 'unicode_strings'> is specified, directly or indirectly.
1537 (It is indirectly specified by a C<use v5.12> or higher.)
1539 In character semantics these upper-Latin1 characters are interpreted as
1540 Unicode code points, which means
1541 they have the same semantics as Latin-1 (ISO-8859-1).
1543 In byte semantics (without C<unicode_strings>), they are considered to
1544 be unassigned characters, meaning that the only semantics they have is
1545 their ordinal numbers, and that they are
1546 not members of various character classes. None are considered to match C<\w>
1547 for example, but all match C<\W>.
1549 Perl 5.12.0 added C<unicode_strings> to force character semantics on
1550 these code points in some circumstances, which fixed portions of the
1551 bug; Perl 5.14.0 fixed almost all of it; and Perl 5.16.0 fixed the
1552 remainder (so far as we know, anyway). The lesson here is to enable
1553 C<unicode_strings> to avoid the headaches described below.
1555 The old, problematic behavior affects these areas:
1561 Changing the case of a scalar, that is, using C<uc()>, C<ucfirst()>, C<lc()>,
1562 and C<lcfirst()>, or C<\L>, C<\U>, C<\u> and C<\l> in double-quotish
1563 contexts, such as regular expression substitutions.
1564 Under C<unicode_strings> starting in Perl 5.12.0, character semantics are
1565 generally used. See L<perlfunc/lc> for details on how this works
1566 in combination with various other pragmas.
1570 Using caseless (C</i>) regular expression matching.
1571 Starting in Perl 5.14.0, regular expressions compiled within
1572 the scope of C<unicode_strings> use character semantics
1573 even when executed or compiled into larger
1574 regular expressions outside the scope.
1578 Matching any of several properties in regular expressions, namely C<\b>,
1579 C<\B>, C<\s>, C<\S>, C<\w>, C<\W>, and all the Posix character classes
1580 I<except> C<[[:ascii:]]>.
1581 Starting in Perl 5.14.0, regular expressions compiled within
1582 the scope of C<unicode_strings> use character semantics
1583 even when executed or compiled into larger
1584 regular expressions outside the scope.
1588 In C<quotemeta> or its inline equivalent C<\Q>, no code points above 127
1589 are quoted in UTF-8 encoded strings, but in byte encoded strings, code
1590 points between 128-255 are always quoted.
1591 Starting in Perl 5.16.0, consistent quoting rules are used within the
1592 scope of C<unicode_strings>, as described in L<perlfunc/quotemeta>.
1596 This behavior can lead to unexpected results in which a string's semantics
1597 suddenly change if a code point above 255 is appended to or removed from it,
1598 which changes the string's semantics from byte to character or vice versa. As
1599 an example, consider the following program and its output:
1602 no feature 'unicode_strings';
1605 for ($s1, $s2, $s1.$s2) {
1613 If there's no C<\w> in C<s1> or in C<s2>, why does their concatenation have one?
1615 This anomaly stems from Perl's attempt to not disturb older programs that
1616 didn't use Unicode, and hence had no semantics for characters outside of the
1617 ASCII range (except in a locale), along with Perl's desire to add Unicode
1618 support seamlessly. The result wasn't seamless: these characters were
1621 For Perls earlier than those described above, or when a string is passed
1622 to a function outside the subpragma's scope, a workaround is to always
1623 call L<C<utf8::upgrade($string)>|utf8/Utility functions>,
1624 or to use the standard module L<Encode>. Also, a scalar that has any characters
1625 whose ordinal is C<0x100> or above, or which were specified using either of the
1626 C<\N{...}> notations, will automatically have character semantics.
1628 =head2 Forcing Unicode in Perl (Or Unforcing Unicode in Perl)
1630 Sometimes (see L</"When Unicode Does Not Happen"> or L</The "Unicode Bug">)
1631 there are situations where you simply need to force a byte
1632 string into UTF-8, or vice versa. The low-level calls
1633 L<C<utf8::upgrade($bytestring)>|utf8/Utility functions> and
1634 L<C<utf8::downgrade($utf8string[, FAIL_OK])>|utf8/Utility functions> are
1637 Note that C<utf8::downgrade()> can fail if the string contains characters
1638 that don't fit into a byte.
1640 Calling either function on a string that already is in the desired state is a
1643 =head2 Using Unicode in XS
1645 If you want to handle Perl Unicode in XS extensions, you may find the
1646 following C APIs useful. See also L<perlguts/"Unicode Support"> for an
1647 explanation about Unicode at the XS level, and L<perlapi> for the API
1654 C<DO_UTF8(sv)> returns true if the C<UTF8> flag is on and the bytes
1655 pragma is not in effect. C<SvUTF8(sv)> returns true if the C<UTF8>
1656 flag is on; the C<bytes> pragma is ignored. The C<UTF8> flag being on
1657 does B<not> mean that there are any characters of code points greater
1658 than 255 (or 127) in the scalar or that there are even any characters
1659 in the scalar. What the C<UTF8> flag means is that the sequence of
1660 octets in the representation of the scalar is the sequence of UTF-8
1661 encoded code points of the characters of a string. The C<UTF8> flag
1662 being off means that each octet in this representation encodes a
1663 single character with code point 0..255 within the string. Perl's
1664 Unicode model is not to use UTF-8 until it is absolutely necessary.
1668 C<uvchr_to_utf8(buf, chr)> writes a Unicode character code point into
1669 a buffer encoding the code point as UTF-8, and returns a pointer
1670 pointing after the UTF-8 bytes. It works appropriately on EBCDIC machines.
1674 C<utf8_to_uvchr_buf(buf, bufend, lenp)> reads UTF-8 encoded bytes from a
1676 returns the Unicode character code point and, optionally, the length of
1677 the UTF-8 byte sequence. It works appropriately on EBCDIC machines.
1681 C<utf8_length(start, end)> returns the length of the UTF-8 encoded buffer
1682 in characters. C<sv_len_utf8(sv)> returns the length of the UTF-8 encoded
1687 C<sv_utf8_upgrade(sv)> converts the string of the scalar to its UTF-8
1688 encoded form. C<sv_utf8_downgrade(sv)> does the opposite, if
1689 possible. C<sv_utf8_encode(sv)> is like sv_utf8_upgrade except that
1690 it does not set the C<UTF8> flag. C<sv_utf8_decode()> does the
1691 opposite of C<sv_utf8_encode()>. Note that none of these are to be
1692 used as general-purpose encoding or decoding interfaces: C<use Encode>
1693 for that. C<sv_utf8_upgrade()> is affected by the encoding pragma
1694 but C<sv_utf8_downgrade()> is not (since the encoding pragma is
1695 designed to be a one-way street).
1699 C<is_utf8_string(buf, len)> returns true if C<len> bytes of the buffer
1704 C<isUTF8_CHAR(buf, buf_end)> returns true if the pointer points to
1705 a valid UTF-8 character.
1709 C<UTF8SKIP(buf)> will return the number of bytes in the UTF-8 encoded
1710 character in the buffer. C<UNISKIP(chr)> will return the number of bytes
1711 required to UTF-8-encode the Unicode character code point. C<UTF8SKIP()>
1712 is useful for example for iterating over the characters of a UTF-8
1713 encoded buffer; C<UNISKIP()> is useful, for example, in computing
1714 the size required for a UTF-8 encoded buffer.
1718 C<utf8_distance(a, b)> will tell the distance in characters between the
1719 two pointers pointing to the same UTF-8 encoded buffer.
1723 C<utf8_hop(s, off)> will return a pointer to a UTF-8 encoded buffer
1724 that is C<off> (positive or negative) Unicode characters displaced
1725 from the UTF-8 buffer C<s>. Be careful not to overstep the buffer:
1726 C<utf8_hop()> will merrily run off the end or the beginning of the
1727 buffer if told to do so.
1731 C<pv_uni_display(dsv, spv, len, pvlim, flags)> and
1732 C<sv_uni_display(dsv, ssv, pvlim, flags)> are useful for debugging the
1733 output of Unicode strings and scalars. By default they are useful
1734 only for debugging--they display B<all> characters as hexadecimal code
1735 points--but with the flags C<UNI_DISPLAY_ISPRINT>,
1736 C<UNI_DISPLAY_BACKSLASH>, and C<UNI_DISPLAY_QQ> you can make the
1737 output more readable.
1741 C<foldEQ_utf8(s1, pe1, l1, u1, s2, pe2, l2, u2)> can be used to
1742 compare two strings case-insensitively in Unicode. For case-sensitive
1743 comparisons you can just use C<memEQ()> and C<memNE()> as usual, except
1744 if one string is in utf8 and the other isn't.
1748 For more information, see L<perlapi>, and F<utf8.c> and F<utf8.h>
1749 in the Perl source code distribution.
1751 =head2 Hacking Perl to work on earlier Unicode versions (for very serious hackers only)
1753 Perl by default comes with the latest supported Unicode version built in, but
1754 you can change to use any earlier one.
1756 Download the files in the desired version of Unicode from the Unicode web
1757 site L<http://www.unicode.org>). These should replace the existing files in
1758 F<lib/unicore> in the Perl source tree. Follow the instructions in
1759 F<README.perl> in that directory to change some of their names, and then build
1760 perl (see L<INSTALL>).
1764 =head2 Interaction with Locales
1766 See L<perllocale/Unicode and UTF-8>
1768 =head2 Problems with characters in the Latin-1 Supplement range
1770 See L</The "Unicode Bug">
1772 =head2 Interaction with Extensions
1774 When Perl exchanges data with an extension, the extension should be
1775 able to understand the UTF8 flag and act accordingly. If the
1776 extension doesn't recognize that flag, it's likely that the extension
1777 will return incorrectly-flagged data.
1779 So if you're working with Unicode data, consult the documentation of
1780 every module you're using if there are any issues with Unicode data
1781 exchange. If the documentation does not talk about Unicode at all,
1782 suspect the worst and probably look at the source to learn how the
1783 module is implemented. Modules written completely in Perl shouldn't
1784 cause problems. Modules that directly or indirectly access code written
1785 in other programming languages are at risk.
1787 For affected functions, the simple strategy to avoid data corruption is
1788 to always make the encoding of the exchanged data explicit. Choose an
1789 encoding that you know the extension can handle. Convert arguments passed
1790 to the extensions to that encoding and convert results back from that
1791 encoding. Write wrapper functions that do the conversions for you, so
1792 you can later change the functions when the extension catches up.
1794 To provide an example, let's say the popular C<Foo::Bar::escape_html>
1795 function doesn't deal with Unicode data yet. The wrapper function
1796 would convert the argument to raw UTF-8 and convert the result back to
1797 Perl's internal representation like so:
1799 sub my_escape_html ($) {
1801 return unless defined $what;
1802 Encode::decode_utf8(Foo::Bar::escape_html(
1803 Encode::encode_utf8($what)));
1806 Sometimes, when the extension does not convert data but just stores
1807 and retrieves them, you will be able to use the otherwise
1808 dangerous L<C<Encode::_utf8_on()>|Encode/_utf8_on> function. Let's say
1809 the popular C<Foo::Bar> extension, written in C, provides a C<param>
1810 method that lets you store and retrieve data according to these prototypes:
1812 $self->param($name, $value); # set a scalar
1813 $value = $self->param($name); # retrieve a scalar
1815 If it does not yet provide support for any encoding, one could write a
1816 derived class with such a C<param> method:
1819 my($self,$name,$value) = @_;
1820 utf8::upgrade($name); # make sure it is UTF-8 encoded
1821 if (defined $value) {
1822 utf8::upgrade($value); # make sure it is UTF-8 encoded
1823 return $self->SUPER::param($name,$value);
1825 my $ret = $self->SUPER::param($name);
1826 Encode::_utf8_on($ret); # we know, it is UTF-8 encoded
1831 Some extensions provide filters on data entry/exit points, such as
1832 C<DB_File::filter_store_key> and family. Look out for such filters in
1833 the documentation of your extensions, they can make the transition to
1834 Unicode data much easier.
1838 Some functions are slower when working on UTF-8 encoded strings than
1839 on byte encoded strings. All functions that need to hop over
1840 characters such as C<length()>, C<substr()> or C<index()>, or matching
1841 regular expressions can work B<much> faster when the underlying data are
1844 In Perl 5.8.0 the slowness was often quite spectacular; in Perl 5.8.1
1845 a caching scheme was introduced which will hopefully make the slowness
1846 somewhat less spectacular, at least for some operations. In general,
1847 operations with UTF-8 encoded strings are still slower. As an example,
1848 the Unicode properties (character classes) like C<\p{Nd}> are known to
1849 be quite a bit slower (5-20 times) than their simpler counterparts
1850 like C<\d> (then again, there are hundreds of Unicode characters matching C<Nd>
1851 compared with the 10 ASCII characters matching C<d>).
1853 =head2 Problems on EBCDIC platforms
1855 There are several known problems with Perl on EBCDIC platforms. If you
1856 want to use Perl there, send email to perlbug@perl.org.
1858 In earlier versions, when byte and character data were concatenated,
1859 the new string was sometimes created by
1860 decoding the byte strings as I<ISO 8859-1 (Latin-1)>, even if the
1861 old Unicode string used EBCDIC.
1863 If you find any of these, please report them as bugs.
1865 =head2 Porting code from perl-5.6.X
1867 Perl 5.8 has a different Unicode model from 5.6. In 5.6 the programmer
1868 was required to use the C<utf8> pragma to declare that a given scope
1869 expected to deal with Unicode data and had to make sure that only
1870 Unicode data were reaching that scope. If you have code that is
1871 working with 5.6, you will need some of the following adjustments to
1872 your code. The examples are written such that the code will continue
1873 to work under 5.6, so you should be safe to try them out.
1879 A filehandle that should read or write UTF-8
1882 binmode $fh, ":encoding(utf8)";
1887 A scalar that is going to be passed to some extension
1889 Be it C<Compress::Zlib>, C<Apache::Request> or any extension that has no
1890 mention of Unicode in the manpage, you need to make sure that the
1891 UTF8 flag is stripped off. Note that at the time of this writing
1892 (January 2012) the mentioned modules are not UTF-8-aware. Please
1893 check the documentation to verify if this is still true.
1897 $val = Encode::encode_utf8($val); # make octets
1902 A scalar we got back from an extension
1904 If you believe the scalar comes back as UTF-8, you will most likely
1905 want the UTF8 flag restored:
1909 $val = Encode::decode_utf8($val);
1914 Same thing, if you are really sure it is UTF-8
1918 Encode::_utf8_on($val);
1923 A wrapper for L<DBI> C<fetchrow_array> and C<fetchrow_hashref>
1925 When the database contains only UTF-8, a wrapper function or method is
1926 a convenient way to replace all your C<fetchrow_array> and
1927 C<fetchrow_hashref> calls. A wrapper function will also make it easier to
1928 adapt to future enhancements in your database driver. Note that at the
1929 time of this writing (January 2012), the DBI has no standardized way
1930 to deal with UTF-8 data. Please check the L<DBI documentation|DBI> to verify if
1934 # $what is one of fetchrow_{array,hashref}
1935 my($self, $sth, $what) = @_;
1941 my @arr = $sth->$what;
1943 defined && /[^\000-\177]/ && Encode::_utf8_on($_);
1947 my $ret = $sth->$what;
1949 for my $k (keys %$ret) {
1952 && Encode::_utf8_on($_) for $ret->{$k};
1956 defined && /[^\000-\177]/ && Encode::_utf8_on($_) for $ret;
1966 A large scalar that you know can only contain ASCII
1968 Scalars that contain only ASCII and are marked as UTF-8 are sometimes
1969 a drag to your program. If you recognize such a situation, just remove
1972 utf8::downgrade($val) if $] > 5.008;
1978 L<perlunitut>, L<perluniintro>, L<perluniprops>, L<Encode>, L<open>, L<utf8>, L<bytes>,
1979 L<perlretut>, L<perlvar/"${^UNICODE}">
1980 L<http://www.unicode.org/reports/tr44>).