2 # $Id: Encode.pm,v 2.9 2004/12/03 19:16:40 dankogai Exp $
6 # our $VERSION = do { my @r = (q$Revision: 2.9 $ =~ /\d+/g); sprintf "%d."."%02d" x $#r, @r };
7 our $VERSION = '2.0902_01';
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
19 encodings find_encoding clone_encoding
22 our @FB_FLAGS = qw(DIE_ON_ERR WARN_ON_ERR RETURN_ON_ERR LEAVE_SRC
23 PERLQQ HTMLCREF XMLCREF);
24 our @FB_CONSTS = qw(FB_DEFAULT FB_CROAK FB_QUIET FB_WARN
25 FB_PERLQQ FB_HTMLCREF FB_XMLCREF);
30 _utf8_off _utf8_on define_encoding from_to is_16bit is_8bit
31 is_utf8 perlio_ok resolve_alias utf8_downgrade utf8_upgrade
33 @FB_FLAGS, @FB_CONSTS,
38 all => [ @EXPORT, @EXPORT_OK ],
39 fallbacks => [ @FB_CONSTS ],
40 fallback_all => [ @FB_CONSTS, @FB_FLAGS ],
43 # Documentation moved after __END__ for speed - NI-S
45 our $ON_EBCDIC = (ord("A") == 193);
49 # Make a %Encoding package variable to allow a certain amount of cheating
52 require Encode::Config;
53 eval { require Encode::ConfigLocal };
59 if (@_ and $_[0] eq ":all"){
60 %enc = ( %Encoding, %ExtModule );
63 for my $mod (map {m/::/o ? $_ : "Encode::$_" } @_){
65 for my $enc (keys %ExtModule){
66 $ExtModule{$enc} eq $mod and $enc{$enc} = $mod;
71 sort { lc $a cmp lc $b }
72 grep {!/^(?:Internal|Unicode|Guess)$/o} keys %enc;
76 my $obj = ref($_[0]) ? $_[0] : find_encoding($_[0]);
77 $obj->can("perlio_ok") and return $obj->perlio_ok();
78 return 0; # safety net
85 $Encoding{$name} = $obj;
87 define_alias($lc => $obj) unless $lc eq $name;
90 define_alias($alias, $obj);
97 my ($class, $name, $skip_external) = @_;
99 ref($name) && $name->can('renew') and return $name;
100 exists $Encoding{$name} and return $Encoding{$name};
102 exists $Encoding{$lc} and return $Encoding{$lc};
104 my $oc = $class->find_alias($name);
105 defined($oc) and return $oc;
106 $lc ne $name and $oc = $class->find_alias($lc);
107 defined($oc) and return $oc;
109 unless ($skip_external)
111 if (my $mod = $ExtModule{$name} || $ExtModule{$lc}){
112 $mod =~ s,::,/,g ; $mod .= '.pm';
113 eval{ require $mod; };
114 exists $Encoding{$name} and return $Encoding{$name};
120 sub find_encoding($;$)
122 my ($name, $skip_external) = @_;
123 return __PACKAGE__->getEncoding($name,$skip_external);
126 sub resolve_alias($){
127 my $obj = find_encoding(shift);
128 defined $obj and return $obj->name;
132 sub clone_encoding($){
133 my $obj = find_encoding(shift);
135 eval { require Storable };
137 return Storable::dclone($obj);
142 my ($name, $string, $check) = @_;
143 return undef unless defined $string;
144 $string .= '' if ref $string; # stringify;
146 my $enc = find_encoding($name);
147 unless(defined $enc){
149 Carp::croak("Unknown encoding '$name'");
151 my $octets = $enc->encode($string,$check);
152 $_[1] = $string if $check and !($check & LEAVE_SRC());
158 my ($name,$octets,$check) = @_;
159 return undef unless defined $octets;
160 $octets .= '' if ref $octets;
162 my $enc = find_encoding($name);
163 unless(defined $enc){
165 Carp::croak("Unknown encoding '$name'");
167 my $string = $enc->decode($octets,$check);
168 $_[1] = $octets if $check and !($check & LEAVE_SRC());
174 my ($string,$from,$to,$check) = @_;
175 return undef unless defined $string;
177 my $f = find_encoding($from);
180 Carp::croak("Unknown encoding '$from'");
182 my $t = find_encoding($to);
185 Carp::croak("Unknown encoding '$to'");
187 my $uni = $f->decode($string,$check);
188 return undef if ($check && length($string));
189 $string = $t->encode($uni,$check);
190 return undef if ($check && length($uni));
191 return defined($_[0] = $string) ? length($string) : undef ;
203 my ($str, $check) = @_;
205 return decode("utf8", $str, $check);
207 return undef unless utf8::decode($str);
212 predefine_encodings(1);
215 # This is to restore %Encoding if really needed;
218 sub predefine_encodings{
219 use Encode::Encoding;
220 no warnings 'redefine';
223 # was in Encode::UTF_EBCDIC
224 package Encode::UTF_EBCDIC;
225 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
227 my ($obj,$str,$chk) = @_;
229 for (my $i = 0; $i < length($str); $i++) {
231 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
237 my ($obj,$str,$chk) = @_;
239 for (my $i = 0; $i < length($str); $i++) {
241 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
246 $Encode::Encoding{Unicode} =
247 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
249 package Encode::Internal;
250 push @Encode::Internal::ISA, 'Encode::Encoding';
252 my ($obj,$str,$chk) = @_;
258 $Encode::Encoding{Unicode} =
259 bless {Name => "Internal"} => "Encode::Internal";
263 # was in Encode::utf8
264 package Encode::utf8;
265 push @Encode::utf8::ISA, 'Encode::Encoding';
268 Encode::DEBUG and warn __PACKAGE__, " XS on";
269 *decode = \&decode_xs;
270 *encode = \&encode_xs;
272 Encode::DEBUG and warn __PACKAGE__, " XS off";
274 my ($obj,$octets,$chk) = @_;
275 my $str = Encode::decode_utf8($octets);
283 my ($obj,$string,$chk) = @_;
284 my $octets = Encode::encode_utf8($string);
289 *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk)
290 my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk
291 my ($rdst, $rsrc, $rpos) = \@_[1,2,3];
293 if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) {
294 $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm));
295 $$rpos = $npos + length($trm);
298 $$rdst .= substr($$rsrc, $pos);
299 $$rpos = length($$rsrc);
302 $Encode::Encoding{utf8} =
303 bless {Name => "utf8"} => "Encode::utf8";
304 $Encode::Encoding{"utf-8-strict"} =
305 bless {Name => "utf-8-strict", strict_utf8 => 1 } => "Encode::utf8";
315 Encode - character encodings
321 =head2 Table of Contents
323 Encode consists of a collection of modules whose details are too big
324 to fit in one document. This POD itself explains the top-level APIs
325 and general topics at a glance. For other topics and more details,
329 --------------------------------------------------------
330 Encode::Alias Alias definitions to encodings
331 Encode::Encoding Encode Implementation Base Class
332 Encode::Supported List of Supported Encodings
333 Encode::CN Simplified Chinese Encodings
334 Encode::JP Japanese Encodings
335 Encode::KR Korean Encodings
336 Encode::TW Traditional Chinese Encodings
337 --------------------------------------------------------
341 The C<Encode> module provides the interfaces between Perl's strings
342 and the rest of the system. Perl strings are sequences of
345 The repertoire of characters that Perl can represent is at least that
346 defined by the Unicode Consortium. On most platforms the ordinal
347 values of the characters (as returned by C<ord(ch)>) is the "Unicode
348 codepoint" for the character (the exceptions are those platforms where
349 the legacy encoding is some variant of EBCDIC rather than a super-set
350 of ASCII - see L<perlebcdic>).
352 Traditionally, computer data has been moved around in 8-bit chunks
353 often called "bytes". These chunks are also known as "octets" in
354 networking standards. Perl is widely used to manipulate data of many
355 types - not only strings of characters representing human or computer
356 languages but also "binary" data being the machine's representation of
357 numbers, pixels in an image - or just about anything.
359 When Perl is processing "binary data", the programmer wants Perl to
360 process "sequences of bytes". This is not a problem for Perl - as a
361 byte has 256 possible values, it easily fits in Perl's much larger
370 I<character>: a character in the range 0..(2**32-1) (or more).
371 (What Perl's strings are made of.)
375 I<byte>: a character in the range 0..255
376 (A special case of a Perl character.)
380 I<octet>: 8 bits of data, with ordinal values 0..255
381 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
385 =head1 PERL ENCODING API
389 =item $octets = encode(ENCODING, $string [, CHECK])
391 Encodes a string from Perl's internal form into I<ENCODING> and returns
392 a sequence of octets. ENCODING can be either a canonical name or
393 an alias. For encoding names and aliases, see L</"Defining Aliases">.
394 For CHECK, see L</"Handling Malformed Data">.
396 For example, to convert a string from Perl's internal format to
397 iso-8859-1 (also known as Latin1),
399 $octets = encode("iso-8859-1", $string);
401 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
402 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
403 for $octets is B<always> off. When you encode anything, utf8 flag of
404 the result is always off, even when it contains completely valid utf8
405 string. See L</"The UTF-8 flag"> below.
407 If the $string is C<undef> then C<undef> is returned.
409 =item $string = decode(ENCODING, $octets [, CHECK])
411 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
412 internal form and returns the resulting string. As in encode(),
413 ENCODING can be either a canonical name or an alias. For encoding names
414 and aliases, see L</"Defining Aliases">. For CHECK, see
415 L</"Handling Malformed Data">.
417 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
419 $string = decode("iso-8859-1", $octets);
421 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
422 B<may not be equal to> $octets. Though they both contain the same data,
423 the utf8 flag for $string is on unless $octets entirely consists of
424 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
427 If the $string is C<undef> then C<undef> is returned.
429 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
431 Converts B<in-place> data between two encodings. The data in $octets
432 must be encoded as octets and not as characters in Perl's internal
433 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250
436 from_to($octets, "iso-8859-1", "cp1250");
438 and to convert it back:
440 from_to($octets, "cp1250", "iso-8859-1");
442 Note that because the conversion happens in place, the data to be
443 converted cannot be a string constant; it must be a scalar variable.
445 from_to() returns the length of the converted string in octets on
446 success, I<undef> on error.
448 B<CAVEAT>: The following operations look the same but are not quite so;
450 from_to($data, "iso-8859-1", "utf8"); #1
451 $data = decode("iso-8859-1", $data); #2
453 Both #1 and #2 make $data consist of a completely valid UTF-8 string
454 but only #2 turns utf8 flag on. #1 is equivalent to
456 $data = encode("utf8", decode("iso-8859-1", $data));
458 See L</"The UTF-8 flag"> below.
460 =item $octets = encode_utf8($string);
462 Equivalent to C<$octets = encode("utf8", $string);> The characters
463 that comprise $string are encoded in Perl's internal format and the
464 result is returned as a sequence of octets. All possible
465 characters have a UTF-8 representation so this function cannot fail.
468 =item $string = decode_utf8($octets [, CHECK]);
470 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
471 The sequence of octets represented by
472 $octets is decoded from UTF-8 into a sequence of logical
473 characters. Not all sequences of octets form valid UTF-8 encodings, so
474 it is possible for this call to fail. For CHECK, see
475 L</"Handling Malformed Data">.
479 =head2 Listing available encodings
482 @list = Encode->encodings();
484 Returns a list of the canonical names of the available encodings that
485 are loaded. To get a list of all available encodings including the
486 ones that are not loaded yet, say
488 @all_encodings = Encode->encodings(":all");
490 Or you can give the name of a specific module.
492 @with_jp = Encode->encodings("Encode::JP");
494 When "::" is not in the name, "Encode::" is assumed.
496 @ebcdic = Encode->encodings("EBCDIC");
498 To find out in detail which encodings are supported by this package,
499 see L<Encode::Supported>.
501 =head2 Defining Aliases
503 To add a new alias to a given encoding, use:
507 define_alias(newName => ENCODING);
509 After that, newName can be used as an alias for ENCODING.
510 ENCODING may be either the name of an encoding or an
513 But before you do so, make sure the alias is nonexistent with
514 C<resolve_alias()>, which returns the canonical name thereof.
517 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
518 Encode::resolve_alias("iso-8859-12") # false; nonexistent
519 Encode::resolve_alias($name) eq $name # true if $name is canonical
521 resolve_alias() does not need C<use Encode::Alias>; it can be
522 exported via C<use Encode qw(resolve_alias)>.
524 See L<Encode::Alias> for details.
526 =head1 Encoding via PerlIO
528 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
529 and encode directly via a filehandle. The following two examples
530 are totally identical in their functionality.
533 open my $in, "<:encoding(shiftjis)", $infile or die;
534 open my $out, ">:encoding(euc-jp)", $outfile or die;
535 while(<$in>){ print $out $_; }
538 open my $in, "<", $infile or die;
539 open my $out, ">", $outfile or die;
541 from_to($_, "shiftjis", "euc-jp", 1);
545 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
546 if your encoding is supported by PerlIO by calling the C<perlio_ok>
549 Encode::perlio_ok("hz"); # False
550 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
552 use Encode qw(perlio_ok); # exported upon request
555 Fortunately, all encodings that come with Encode core are PerlIO-savvy
556 except for hz and ISO-2022-kr. For gory details, see
557 L<Encode::Encoding> and L<Encode::PerlIO>.
559 =head1 Handling Malformed Data
561 The optional I<CHECK> argument is used as follows. When you omit it,
562 Encode::FB_DEFAULT ( == 0 ) is assumed.
566 =item B<NOTE:> Not all encoding suppport this feature
568 Some encodings ignore I<CHECK> argument. For example,
569 L<Encode::Unicode> ignores I<CHECK> and it always croaks on error.
573 Now here is the list of I<CHECK> values available
577 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
579 If I<CHECK> is 0, (en|de)code will put a I<substitution character> in
580 place of a malformed character. When you encode, E<lt>subcharE<gt>
581 will be used. When you decode the code point C<0xFFFD> is used. If
582 the data is supposed to be UTF-8, an optional lexical warning
583 (category utf8) is given.
585 =item I<CHECK> = Encode::FB_CROAK ( == 1)
587 If I<CHECK> is 1, methods will die on error immediately with an error
588 message. Therefore, when I<CHECK> is set to 1, you should trap the
589 error with eval{} unless you really want to let it die.
591 =item I<CHECK> = Encode::FB_QUIET
593 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
594 return the portion of the data that has been processed so far when an
595 error occurs. The data argument will be overwritten with everything
596 after that point (that is, the unprocessed part of data). This is
597 handy when you have to call decode repeatedly in the case where your
598 source data may contain partial multi-byte character sequences,
599 (i.e. you are reading with a fixed-width buffer). Here is a sample
600 code that does exactly this:
602 my $buffer = ''; my $string = '';
603 while(read $fh, $buffer, 256, length($buffer)){
604 $string .= decode($encoding, $buffer, Encode::FB_QUIET);
605 # $buffer now contains the unprocessed partial character
608 =item I<CHECK> = Encode::FB_WARN
610 This is the same as above, except that it warns on error. Handy when
611 you are debugging the mode above.
613 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
615 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
617 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
619 For encodings that are implemented by Encode::XS, CHECK ==
620 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
622 When you decode, C<\xI<HH>> will be inserted for a malformed character,
623 where I<HH> is the hex representation of the octet that could not be
624 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
625 where I<HHHH> is the Unicode ID of the character that cannot be found
626 in the character repertoire of the encoding.
628 HTML/XML character reference modes are about the same, in place of
629 C<\x{I<HHHH>}>, HTML uses C<&#I<NNN>;> where I<NNN> is a decimal number and
630 XML uses C<&#xI<HHHH>;> where I<HHHH> is the hexadecimal number.
632 In Encode 2.10 or later, C<LEAVE_SRC> is also implied.
636 These modes are actually set via a bitmask. Here is how the FB_XX
637 constants are laid out. You can import the FB_XX constants via
638 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
639 constants via C<use Encode qw(:fallback_all)>.
641 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
644 RETURN_ON_ERR 0x0004 X X
652 =head2 Unimplemented fallback schemes
654 In the future, you will be able to use a code reference to a callback
655 function for the value of I<CHECK> but its API is still undecided.
657 The fallback scheme does not work on EBCDIC platforms.
659 =head1 Defining Encodings
661 To define a new encoding, use:
663 use Encode qw(define_encoding);
664 define_encoding($object, 'canonicalName' [, alias...]);
666 I<canonicalName> will be associated with I<$object>. The object
667 should provide the interface described in L<Encode::Encoding>.
668 If more than two arguments are provided then additional
669 arguments are taken as aliases for I<$object>.
671 See L<Encode::Encoding> for more details.
673 =head1 The UTF-8 flag
675 Before the introduction of utf8 support in perl, The C<eq> operator
676 just compared the strings represented by two scalars. Beginning with
677 perl 5.8, C<eq> compares two strings with simultaneous consideration
678 of I<the utf8 flag>. To explain why we made it so, I will quote page
679 402 of C<Programming Perl, 3rd ed.>
685 Old byte-oriented programs should not spontaneously break on the old
686 byte-oriented data they used to work on.
690 Old byte-oriented programs should magically start working on the new
691 character-oriented data when appropriate.
695 Programs should run just as fast in the new character-oriented mode
696 as in the old byte-oriented mode.
700 Perl should remain one language, rather than forking into a
701 byte-oriented Perl and a character-oriented Perl.
705 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
706 was born and many features documented in the book remained
707 unimplemented for a long time. Perl 5.8 corrected this and the introduction
708 of the UTF-8 flag is one of them. You can think of this perl notion as of a
709 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
712 Here is how Encode takes care of the utf8 flag.
718 When you encode, the resulting utf8 flag is always off.
722 When you decode, the resulting utf8 flag is on unless you can
723 unambiguously represent data. Here is the definition of
726 After C<$utf8 = decode('foo', $octet);>,
728 When $octet is... The utf8 flag in $utf8 is
729 ---------------------------------------------
730 In ASCII only (or EBCDIC only) OFF
732 In any other Encoding ON
733 ---------------------------------------------
735 As you see, there is one exception, In ASCII. That way you can assue
736 Goal #1. And with Encode Goal #2 is assumed but you still have to be
737 careful in such cases mentioned in B<CAVEAT> paragraphs.
739 This utf8 flag is not visible in perl scripts, exactly for the same
740 reason you cannot (or you I<don't have to>) see if a scalar contains a
741 string, integer, or floating point number. But you can still peek
742 and poke these if you will. See the section below.
746 =head2 Messing with Perl's Internals
748 The following API uses parts of Perl's internals in the current
749 implementation. As such, they are efficient but may change.
753 =item is_utf8(STRING [, CHECK])
755 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
756 If CHECK is true, also checks the data in STRING for being well-formed
757 UTF-8. Returns true if successful, false otherwise.
759 As of perl 5.8.1, L<utf8> also has utf8::is_utf8().
761 =item _utf8_on(STRING)
763 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
764 B<not> checked for being well-formed UTF-8. Do not use unless you
765 B<know> that the STRING is well-formed UTF-8. Returns the previous
766 state of the UTF-8 flag (so please don't treat the return value as
767 indicating success or failure), or C<undef> if STRING is not a string.
769 =item _utf8_off(STRING)
771 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
772 Returns the previous state of the UTF-8 flag (so please don't treat the
773 return value as indicating success or failure), or C<undef> if STRING is
778 =head1 UTF-8 vs. utf8
780 ....We now view strings not as sequences of bytes, but as sequences
781 of numbers in the range 0 .. 2**32-1 (or in the case of 64-bit
782 computers, 0 .. 2**64-1) -- Programming Perl, 3rd ed.
784 That has been the perl's notion of UTF-8 but official UTF-8 is more
785 strict; Its ranges is much narrower (0 .. 10FFFF), some sequences are
786 not allowed (i.e. Those used in the surrogate pair, 0xFFFE, et al).
788 Now that is overruled by Larry Wall himself.
790 From: Larry Wall <larry@wall.org>
791 Date: December 04, 2004 11:51:58 JST
792 To: perl-unicode@perl.org
793 Subject: Re: Make Encode.pm support the real UTF-8
794 Message-Id: <20041204025158.GA28754@wall.org>
796 On Fri, Dec 03, 2004 at 10:12:12PM +0000, Tim Bunce wrote:
797 : I've no problem with 'utf8' being perl's unrestricted uft8 encoding,
798 : but "UTF-8" is the name of the standard and should give the
799 : corresponding behaviour.
801 For what it's worth, that's how I've always kept them straight in my
804 Also for what it's worth, Perl 6 will mostly default to strict but
805 make it easy to switch back to lax.
809 Do you copy? As of Perl 5.8.7, B<UTF-8> means strict, official UTF-8
810 while B<utf8> means liberal, lax, version thereof. And Encode version
811 2.10 or later thus groks the difference between C<UTF-8> and C"utf8".
813 encode("utf8", "\x{FFFF_FFFF}", 1); # okay
814 encode("UTF-8", "\x{FFFF_FFFF}", 1); # croaks
816 C<UTF-8> in Encode is actually a canonical name for C<utf-8-strict>.
817 Yes, the hyphen between "UTF" and "8" is important. Without it Encode
820 find_encoding("UTF-8")->name # is 'utf-8-strict'
821 find_encoding("utf-8")->name # ditto. names are case insensitive
822 find_encoding("utf8")->name # ditto. "_" are treated as "-"
823 find_encoding("UTF8")->name # is 'utf8'.
829 L<Encode::Supported>,
836 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
840 This project was originated by Nick Ing-Simmons and later maintained
841 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
842 list of people involved. For any questions, use
843 E<lt>perl-unicode@perl.orgE<gt> so we can all share.