2 # $Id: Encode.pm,v 2.29 2009/02/01 13:10:07 dankogai Exp $
7 our $VERSION = sprintf "%d.%02d", q$Revision: 2.29 $ =~ /(\d+)/g;
10 XSLoader::load( __PACKAGE__, $VERSION );
13 use base qw/Exporter/;
15 # Public, encouraged API is exported by default
18 decode decode_utf8 encode encode_utf8 str2bytes bytes2str
19 encodings find_encoding clone_encoding
22 DIE_ON_ERR WARN_ON_ERR RETURN_ON_ERR LEAVE_SRC
23 PERLQQ HTMLCREF XMLCREF STOP_AT_PARTIAL
26 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN
27 FB_PERLQQ FB_HTMLCREF FB_XMLCREF
31 _utf8_off _utf8_on define_encoding from_to is_16bit is_8bit
32 is_utf8 perlio_ok resolve_alias utf8_downgrade utf8_upgrade
34 @FB_FLAGS, @FB_CONSTS,
38 all => [ @EXPORT, @EXPORT_OK ],
39 default => [ @EXPORT ],
40 fallbacks => [ @FB_CONSTS ],
41 fallback_all => [ @FB_CONSTS, @FB_FLAGS ],
44 # Documentation moved after __END__ for speed - NI-S
46 our $ON_EBCDIC = ( ord("A") == 193 );
50 # Make a %Encoding package variable to allow a certain amount of cheating
53 require Encode::Config;
55 # https://bugzilla.redhat.com/show_bug.cgi?id=435505#c2
56 # to find why sig handers inside eval{} are disabled.
60 require Encode::ConfigLocal;
66 if ( @_ and $_[0] eq ":all" ) {
67 %enc = ( %Encoding, %ExtModule );
71 for my $mod ( map { m/::/o ? $_ : "Encode::$_" } @_ ) {
73 for my $enc ( keys %ExtModule ) {
74 $ExtModule{$enc} eq $mod and $enc{$enc} = $mod;
78 return sort { lc $a cmp lc $b }
79 grep { !/^(?:Internal|Unicode|Guess)$/o } keys %enc;
83 my $obj = ref( $_[0] ) ? $_[0] : find_encoding( $_[0] );
84 $obj->can("perlio_ok") and return $obj->perlio_ok();
85 return 0; # safety net
91 $Encoding{$name} = $obj;
93 define_alias( $lc => $obj ) unless $lc eq $name;
96 define_alias( $alias, $obj );
102 my ( $class, $name, $skip_external ) = @_;
104 ref($name) && $name->can('renew') and return $name;
105 exists $Encoding{$name} and return $Encoding{$name};
107 exists $Encoding{$lc} and return $Encoding{$lc};
109 my $oc = $class->find_alias($name);
110 defined($oc) and return $oc;
111 $lc ne $name and $oc = $class->find_alias($lc);
112 defined($oc) and return $oc;
114 unless ($skip_external) {
115 if ( my $mod = $ExtModule{$name} || $ExtModule{$lc} ) {
118 eval { require $mod; };
119 exists $Encoding{$name} and return $Encoding{$name};
125 sub find_encoding($;$) {
126 my ( $name, $skip_external ) = @_;
127 return __PACKAGE__->getEncoding( $name, $skip_external );
130 sub resolve_alias($) {
131 my $obj = find_encoding(shift);
132 defined $obj and return $obj->name;
136 sub clone_encoding($) {
137 my $obj = find_encoding(shift);
139 eval { require Storable };
141 return Storable::dclone($obj);
145 my ( $name, $string, $check ) = @_;
146 return undef unless defined $string;
147 $string .= '' if ref $string; # stringify;
149 my $enc = find_encoding($name);
150 unless ( defined $enc ) {
152 Carp::croak("Unknown encoding '$name'");
154 my $octets = $enc->encode( $string, $check );
155 $_[1] = $string if $check and !ref $check and !( $check & LEAVE_SRC() );
158 *str2bytes = \&encode;
161 my ( $name, $octets, $check ) = @_;
162 return undef unless defined $octets;
163 $octets .= '' if ref $octets;
165 my $enc = find_encoding($name);
166 unless ( defined $enc ) {
168 Carp::croak("Unknown encoding '$name'");
170 my $string = $enc->decode( $octets, $check );
171 $_[1] = $octets if $check and !ref $check and !( $check & LEAVE_SRC() );
174 *bytes2str = \&decode;
177 my ( $string, $from, $to, $check ) = @_;
178 return undef unless defined $string;
180 my $f = find_encoding($from);
181 unless ( defined $f ) {
183 Carp::croak("Unknown encoding '$from'");
185 my $t = find_encoding($to);
186 unless ( defined $t ) {
188 Carp::croak("Unknown encoding '$to'");
190 my $uni = $f->decode($string);
191 $_[0] = $string = $t->encode( $uni, $check );
192 return undef if ( $check && length($uni) );
193 return defined( $_[0] ) ? length($string) : undef;
202 sub decode_utf8($;$) {
203 my ( $str, $check ) = @_;
204 return $str if is_utf8($str);
206 return decode( "utf8", $str, $check );
209 return decode( "utf8", $str );
214 predefine_encodings(1);
217 # This is to restore %Encoding if really needed;
220 sub predefine_encodings {
221 require Encode::Encoding;
222 no warnings 'redefine';
226 # was in Encode::UTF_EBCDIC
227 package Encode::UTF_EBCDIC;
228 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
230 my ( $obj, $str, $chk ) = @_;
232 for ( my $i = 0 ; $i < length($str) ; $i++ ) {
235 utf8::unicode_to_native( ord( substr( $str, $i, 1 ) ) )
242 my ( $obj, $str, $chk ) = @_;
244 for ( my $i = 0 ; $i < length($str) ; $i++ ) {
247 utf8::native_to_unicode( ord( substr( $str, $i, 1 ) ) )
253 $Encode::Encoding{Unicode} =
254 bless { Name => "UTF_EBCDIC" } => "Encode::UTF_EBCDIC";
258 package Encode::Internal;
259 push @Encode::Internal::ISA, 'Encode::Encoding';
261 my ( $obj, $str, $chk ) = @_;
267 $Encode::Encoding{Unicode} =
268 bless { Name => "Internal" } => "Encode::Internal";
273 # was in Encode::utf8
274 package Encode::utf8;
275 push @Encode::utf8::ISA, 'Encode::Encoding';
279 Encode::DEBUG and warn __PACKAGE__, " XS on";
280 *decode = \&decode_xs;
281 *encode = \&encode_xs;
284 Encode::DEBUG and warn __PACKAGE__, " XS off";
286 my ( $obj, $octets, $chk ) = @_;
287 my $str = Encode::decode_utf8($octets);
288 if ( defined $str ) {
295 my ( $obj, $string, $chk ) = @_;
296 my $octets = Encode::encode_utf8($string);
301 *cat_decode = sub { # ($obj, $dst, $src, $pos, $trm, $chk)
302 # currently ignores $chk
303 my ( $obj, undef, undef, $pos, $trm ) = @_;
304 my ( $rdst, $rsrc, $rpos ) = \@_[ 1, 2, 3 ];
306 if ( ( my $npos = index( $$rsrc, $trm, $pos ) ) >= 0 ) {
308 substr( $$rsrc, $pos, $npos - $pos + length($trm) );
309 $$rpos = $npos + length($trm);
312 $$rdst .= substr( $$rsrc, $pos );
313 $$rpos = length($$rsrc);
316 $Encode::Encoding{utf8} =
317 bless { Name => "utf8" } => "Encode::utf8";
318 $Encode::Encoding{"utf-8-strict"} =
319 bless { Name => "utf-8-strict", strict_utf8 => 1 } =>
330 Encode - character encodings
336 =head2 Table of Contents
338 Encode consists of a collection of modules whose details are too big
339 to fit in one document. This POD itself explains the top-level APIs
340 and general topics at a glance. For other topics and more details,
344 --------------------------------------------------------
345 Encode::Alias Alias definitions to encodings
346 Encode::Encoding Encode Implementation Base Class
347 Encode::Supported List of Supported Encodings
348 Encode::CN Simplified Chinese Encodings
349 Encode::JP Japanese Encodings
350 Encode::KR Korean Encodings
351 Encode::TW Traditional Chinese Encodings
352 --------------------------------------------------------
356 The C<Encode> module provides the interfaces between Perl's strings
357 and the rest of the system. Perl strings are sequences of
360 The repertoire of characters that Perl can represent is at least that
361 defined by the Unicode Consortium. On most platforms the ordinal
362 values of the characters (as returned by C<ord(ch)>) is the "Unicode
363 codepoint" for the character (the exceptions are those platforms where
364 the legacy encoding is some variant of EBCDIC rather than a super-set
365 of ASCII - see L<perlebcdic>).
367 Traditionally, computer data has been moved around in 8-bit chunks
368 often called "bytes". These chunks are also known as "octets" in
369 networking standards. Perl is widely used to manipulate data of many
370 types - not only strings of characters representing human or computer
371 languages but also "binary" data being the machine's representation of
372 numbers, pixels in an image - or just about anything.
374 When Perl is processing "binary data", the programmer wants Perl to
375 process "sequences of bytes". This is not a problem for Perl - as a
376 byte has 256 possible values, it easily fits in Perl's much larger
385 I<character>: a character in the range 0..(2**32-1) (or more).
386 (What Perl's strings are made of.)
390 I<byte>: a character in the range 0..255
391 (A special case of a Perl character.)
395 I<octet>: 8 bits of data, with ordinal values 0..255
396 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
400 =head1 PERL ENCODING API
404 =item $octets = encode(ENCODING, $string [, CHECK])
406 Encodes a string from Perl's internal form into I<ENCODING> and returns
407 a sequence of octets. ENCODING can be either a canonical name or
408 an alias. For encoding names and aliases, see L</"Defining Aliases">.
409 For CHECK, see L</"Handling Malformed Data">.
411 For example, to convert a string from Perl's internal format to
412 iso-8859-1 (also known as Latin1),
414 $octets = encode("iso-8859-1", $string);
416 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then
417 $octets B<may not be equal to> $string. Though they both contain the
418 same data, the UTF8 flag for $octets is B<always> off. When you
419 encode anything, UTF8 flag of the result is always off, even when it
420 contains completely valid utf8 string. See L</"The UTF8 flag"> below.
422 If the $string is C<undef> then C<undef> is returned.
424 =item $string = decode(ENCODING, $octets [, CHECK])
426 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
427 internal form and returns the resulting string. As in encode(),
428 ENCODING can be either a canonical name or an alias. For encoding names
429 and aliases, see L</"Defining Aliases">. For CHECK, see
430 L</"Handling Malformed Data">.
432 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
434 $string = decode("iso-8859-1", $octets);
436 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
437 B<may not be equal to> $octets. Though they both contain the same data,
438 the UTF8 flag for $string is on unless $octets entirely consists of
439 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF8 flag">
442 If the $string is C<undef> then C<undef> is returned.
444 =item [$obj =] find_encoding(ENCODING)
446 Returns the I<encoding object> corresponding to ENCODING. Returns
447 undef if no matching ENCODING is find.
449 This object is what actually does the actual (en|de)coding.
451 $utf8 = decode($name, $bytes);
456 $obj = find_encoding($name);
457 croak qq(encoding "$name" not found) unless ref $obj;
461 with more error checking.
463 Therefore you can save time by reusing this object as follows;
465 my $enc = find_encoding("iso-8859-1");
467 my $utf8 = $enc->decode($_);
468 # and do someting with $utf8;
471 Besides C<< ->decode >> and C<< ->encode >>, other methods are
472 available as well. For instance, C<< -> name >> returns the canonical
473 name of the encoding object.
475 find_encoding("latin1")->name; # iso-8859-1
477 See L<Encode::Encoding> for details.
479 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
481 Converts B<in-place> data between two encodings. The data in $octets
482 must be encoded as octets and not as characters in Perl's internal
483 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250
486 from_to($octets, "iso-8859-1", "cp1250");
488 and to convert it back:
490 from_to($octets, "cp1250", "iso-8859-1");
492 Note that because the conversion happens in place, the data to be
493 converted cannot be a string constant; it must be a scalar variable.
495 from_to() returns the length of the converted string in octets on
496 success, I<undef> on error.
498 B<CAVEAT>: The following operations look the same but are not quite so;
500 from_to($data, "iso-8859-1", "utf8"); #1
501 $data = decode("iso-8859-1", $data); #2
503 Both #1 and #2 make $data consist of a completely valid UTF-8 string
504 but only #2 turns UTF8 flag on. #1 is equivalent to
506 $data = encode("utf8", decode("iso-8859-1", $data));
508 See L</"The UTF8 flag"> below.
512 from_to($octets, $from, $to, $check);
516 $octets = encode($to, decode($from, $octets), $check);
518 Yes, it does not respect the $check during decoding. It is
519 deliberately done that way. If you need minute control, C<decode>
520 then C<encode> as follows;
522 $octets = encode($to, decode($from, $octets, $check_from), $check_to);
524 =item $octets = encode_utf8($string);
526 Equivalent to C<$octets = encode("utf8", $string);> The characters
527 that comprise $string are encoded in Perl's internal format and the
528 result is returned as a sequence of octets. All possible
529 characters have a UTF-8 representation so this function cannot fail.
532 =item $string = decode_utf8($octets [, CHECK]);
534 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
535 The sequence of octets represented by
536 $octets is decoded from UTF-8 into a sequence of logical
537 characters. Not all sequences of octets form valid UTF-8 encodings, so
538 it is possible for this call to fail. For CHECK, see
539 L</"Handling Malformed Data">.
543 =head2 Listing available encodings
546 @list = Encode->encodings();
548 Returns a list of the canonical names of the available encodings that
549 are loaded. To get a list of all available encodings including the
550 ones that are not loaded yet, say
552 @all_encodings = Encode->encodings(":all");
554 Or you can give the name of a specific module.
556 @with_jp = Encode->encodings("Encode::JP");
558 When "::" is not in the name, "Encode::" is assumed.
560 @ebcdic = Encode->encodings("EBCDIC");
562 To find out in detail which encodings are supported by this package,
563 see L<Encode::Supported>.
565 =head2 Defining Aliases
567 To add a new alias to a given encoding, use:
571 define_alias(newName => ENCODING);
573 After that, newName can be used as an alias for ENCODING.
574 ENCODING may be either the name of an encoding or an
577 But before you do so, make sure the alias is nonexistent with
578 C<resolve_alias()>, which returns the canonical name thereof.
581 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
582 Encode::resolve_alias("iso-8859-12") # false; nonexistent
583 Encode::resolve_alias($name) eq $name # true if $name is canonical
585 resolve_alias() does not need C<use Encode::Alias>; it can be
586 exported via C<use Encode qw(resolve_alias)>.
588 See L<Encode::Alias> for details.
590 =head2 Finding IANA Character Set Registry names
592 The canonical name of a given encoding does not necessarily agree with
593 IANA IANA Character Set Registry, commonly seen as C<< Content-Type:
594 text/plain; charset=I<whatever> >>. For most cases canonical names
595 work but sometimes it does not (notably 'utf-8-strict').
597 Therefore as of Encode version 2.21, a new method C<mime_name()> is added.
600 my $enc = find_encoding('UTF-8');
601 warn $enc->name; # utf-8-strict
602 warn $enc->mime_name; # UTF-8
604 See also: L<Encode::Encoding>
606 =head1 Encoding via PerlIO
608 If your perl supports I<PerlIO> (which is the default), you can use a
609 PerlIO layer to decode and encode directly via a filehandle. The
610 following two examples are totally identical in their functionality.
613 open my $in, "<:encoding(shiftjis)", $infile or die;
614 open my $out, ">:encoding(euc-jp)", $outfile or die;
615 while(<$in>){ print $out $_; }
618 open my $in, "<", $infile or die;
619 open my $out, ">", $outfile or die;
621 from_to($_, "shiftjis", "euc-jp", 1);
625 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
626 if your encoding is supported by PerlIO by calling the C<perlio_ok>
629 Encode::perlio_ok("hz"); # False
630 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
632 use Encode qw(perlio_ok); # exported upon request
635 Fortunately, all encodings that come with Encode core are PerlIO-savvy
636 except for hz and ISO-2022-kr. For gory details, see
637 L<Encode::Encoding> and L<Encode::PerlIO>.
639 =head1 Handling Malformed Data
641 The optional I<CHECK> argument tells Encode what to do when it
642 encounters malformed data. Without CHECK, Encode::FB_DEFAULT ( == 0 )
645 As of version 2.12 Encode supports coderef values for CHECK. See below.
649 =item B<NOTE:> Not all encoding support this feature
651 Some encodings ignore I<CHECK> argument. For example,
652 L<Encode::Unicode> ignores I<CHECK> and it always croaks on error.
656 Now here is the list of I<CHECK> values available
660 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
662 If I<CHECK> is 0, (en|de)code will put a I<substitution character> in
663 place of a malformed character. When you encode, E<lt>subcharE<gt>
664 will be used. When you decode the code point C<0xFFFD> is used. If
665 the data is supposed to be UTF-8, an optional lexical warning
666 (category utf8) is given.
668 =item I<CHECK> = Encode::FB_CROAK ( == 1)
670 If I<CHECK> is 1, methods will die on error immediately with an error
671 message. Therefore, when I<CHECK> is set to 1, you should trap the
672 error with eval{} unless you really want to let it die.
674 =item I<CHECK> = Encode::FB_QUIET
676 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
677 return the portion of the data that has been processed so far when an
678 error occurs. The data argument will be overwritten with everything
679 after that point (that is, the unprocessed part of data). This is
680 handy when you have to call decode repeatedly in the case where your
681 source data may contain partial multi-byte character sequences,
682 (i.e. you are reading with a fixed-width buffer). Here is a sample
683 code that does exactly this:
685 my $buffer = ''; my $string = '';
686 while(read $fh, $buffer, 256, length($buffer)){
687 $string .= decode($encoding, $buffer, Encode::FB_QUIET);
688 # $buffer now contains the unprocessed partial character
691 =item I<CHECK> = Encode::FB_WARN
693 This is the same as above, except that it warns on error. Handy when
694 you are debugging the mode above.
696 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
698 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
700 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
702 For encodings that are implemented by Encode::XS, CHECK ==
703 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
705 When you decode, C<\xI<HH>> will be inserted for a malformed character,
706 where I<HH> is the hex representation of the octet that could not be
707 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
708 where I<HHHH> is the Unicode ID of the character that cannot be found
709 in the character repertoire of the encoding.
711 HTML/XML character reference modes are about the same, in place of
712 C<\x{I<HHHH>}>, HTML uses C<&#I<NNN>;> where I<NNN> is a decimal number and
713 XML uses C<&#xI<HHHH>;> where I<HHHH> is the hexadecimal number.
715 In Encode 2.10 or later, C<LEAVE_SRC> is also implied.
719 These modes are actually set via a bitmask. Here is how the FB_XX
720 constants are laid out. You can import the FB_XX constants via
721 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
722 constants via C<use Encode qw(:fallback_all)>.
724 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
727 RETURN_ON_ERR 0x0004 X X
737 =item Encode::LEAVE_SRC
739 If the C<Encode::LEAVE_SRC> bit is not set, but I<CHECK> is, then the second
740 argument to C<encode()> or C<decode()> may be assigned to by the functions. If
741 you're not interested in this, then bitwise-or the bitmask with it.
745 =head2 coderef for CHECK
747 As of Encode 2.12 CHECK can also be a code reference which takes the
748 ord value of unmapped caharacter as an argument and returns a string
749 that represents the fallback character. For instance,
751 $ascii = encode("ascii", $utf8, sub{ sprintf "<U+%04X>", shift });
753 Acts like FB_PERLQQ but E<lt>U+I<XXXX>E<gt> is used instead of
756 =head1 Defining Encodings
758 To define a new encoding, use:
760 use Encode qw(define_encoding);
761 define_encoding($object, 'canonicalName' [, alias...]);
763 I<canonicalName> will be associated with I<$object>. The object
764 should provide the interface described in L<Encode::Encoding>.
765 If more than two arguments are provided then additional
766 arguments are taken as aliases for I<$object>.
768 See L<Encode::Encoding> for more details.
772 Before the introduction of Unicode support in perl, The C<eq> operator
773 just compared the strings represented by two scalars. Beginning with
774 perl 5.8, C<eq> compares two strings with simultaneous consideration of
775 I<the UTF8 flag>. To explain why we made it so, I will quote page 402 of
776 C<Programming Perl, 3rd ed.>
782 Old byte-oriented programs should not spontaneously break on the old
783 byte-oriented data they used to work on.
787 Old byte-oriented programs should magically start working on the new
788 character-oriented data when appropriate.
792 Programs should run just as fast in the new character-oriented mode
793 as in the old byte-oriented mode.
797 Perl should remain one language, rather than forking into a
798 byte-oriented Perl and a character-oriented Perl.
802 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
803 was born and many features documented in the book remained
804 unimplemented for a long time. Perl 5.8 corrected this and the introduction
805 of the UTF8 flag is one of them. You can think of this perl notion as of a
806 byte-oriented mode (UTF8 flag off) and a character-oriented mode (UTF8
809 Here is how Encode takes care of the UTF8 flag.
815 When you encode, the resulting UTF8 flag is always off.
819 When you decode, the resulting UTF8 flag is on unless you can
820 unambiguously represent data. Here is the definition of
823 After C<$utf8 = decode('foo', $octet);>,
825 When $octet is... The UTF8 flag in $utf8 is
826 ---------------------------------------------
827 In ASCII only (or EBCDIC only) OFF
829 In any other Encoding ON
830 ---------------------------------------------
832 As you see, there is one exception, In ASCII. That way you can assume
833 Goal #1. And with Encode Goal #2 is assumed but you still have to be
834 careful in such cases mentioned in B<CAVEAT> paragraphs.
836 This UTF8 flag is not visible in perl scripts, exactly for the same
837 reason you cannot (or you I<don't have to>) see if a scalar contains a
838 string, integer, or floating point number. But you can still peek
839 and poke these if you will. See the section below.
843 =head2 Messing with Perl's Internals
845 The following API uses parts of Perl's internals in the current
846 implementation. As such, they are efficient but may change.
850 =item is_utf8(STRING [, CHECK])
852 [INTERNAL] Tests whether the UTF8 flag is turned on in the STRING.
853 If CHECK is true, also checks the data in STRING for being well-formed
854 UTF-8. Returns true if successful, false otherwise.
856 As of perl 5.8.1, L<utf8> also has utf8::is_utf8().
858 =item _utf8_on(STRING)
860 [INTERNAL] Turns on the UTF8 flag in STRING. The data in STRING is
861 B<not> checked for being well-formed UTF-8. Do not use unless you
862 B<know> that the STRING is well-formed UTF-8. Returns the previous
863 state of the UTF8 flag (so please don't treat the return value as
864 indicating success or failure), or C<undef> if STRING is not a string.
866 This function does not work on tainted values.
868 =item _utf8_off(STRING)
870 [INTERNAL] Turns off the UTF8 flag in STRING. Do not use frivolously.
871 Returns the previous state of the UTF8 flag (so please don't treat the
872 return value as indicating success or failure), or C<undef> if STRING is
875 This function does not work on tainted values.
879 =head1 UTF-8 vs. utf8 vs. UTF8
881 ....We now view strings not as sequences of bytes, but as sequences
882 of numbers in the range 0 .. 2**32-1 (or in the case of 64-bit
883 computers, 0 .. 2**64-1) -- Programming Perl, 3rd ed.
885 That has been the perl's notion of UTF-8 but official UTF-8 is more
886 strict; Its ranges is much narrower (0 .. 10FFFF), some sequences are
887 not allowed (i.e. Those used in the surrogate pair, 0xFFFE, et al).
889 Now that is overruled by Larry Wall himself.
891 From: Larry Wall <larry@wall.org>
892 Date: December 04, 2004 11:51:58 JST
893 To: perl-unicode@perl.org
894 Subject: Re: Make Encode.pm support the real UTF-8
895 Message-Id: <20041204025158.GA28754@wall.org>
897 On Fri, Dec 03, 2004 at 10:12:12PM +0000, Tim Bunce wrote:
898 : I've no problem with 'utf8' being perl's unrestricted uft8 encoding,
899 : but "UTF-8" is the name of the standard and should give the
900 : corresponding behaviour.
902 For what it's worth, that's how I've always kept them straight in my
905 Also for what it's worth, Perl 6 will mostly default to strict but
906 make it easy to switch back to lax.
910 Do you copy? As of Perl 5.8.7, B<UTF-8> means strict, official UTF-8
911 while B<utf8> means liberal, lax, version thereof. And Encode version
912 2.10 or later thus groks the difference between C<UTF-8> and C"utf8".
914 encode("utf8", "\x{FFFF_FFFF}", 1); # okay
915 encode("UTF-8", "\x{FFFF_FFFF}", 1); # croaks
917 C<UTF-8> in Encode is actually a canonical name for C<utf-8-strict>.
918 Yes, the hyphen between "UTF" and "8" is important. Without it Encode
921 find_encoding("UTF-8")->name # is 'utf-8-strict'
922 find_encoding("utf-8")->name # ditto. names are case insensitive
923 find_encoding("utf_8")->name # ditto. "_" are treated as "-"
924 find_encoding("UTF8")->name # is 'utf8'.
926 The UTF8 flag is internally called UTF8, without a hyphen. It indicates
927 whether a string is internally encoded as utf8, also without a hypen.
932 L<Encode::Supported>,
937 L<perlunicode>, L<perluniintro>, L<perlunifaq>, L<perlunitut>
939 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
943 This project was originated by Nick Ing-Simmons and later maintained
944 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
945 list of people involved. For any questions, use
946 E<lt>perl-unicode@perl.orgE<gt> so we can all share.
948 While Dan Kogai retains the copyright as a maintainer, the credit
949 should go to all those involoved. See AUTHORS for those submitted
954 Copyright 2002-2006 Dan Kogai E<lt>dankogai@dan.co.jpE<gt>
956 This library is free software; you can redistribute it and/or modify
957 it under the same terms as Perl itself.