2 # $Id: Encode.pm,v 1.89 2003/02/28 01:36:02 dankogai Exp $
6 our $VERSION = do { my @r = (q$Revision: 1.89 $ =~ /\d+/g); sprintf "%d."."%02d" x $#r, @r };
9 XSLoader::load(__PACKAGE__, $VERSION);
12 use base qw/Exporter/;
14 # Public, encouraged API is exported by default
17 decode decode_utf8 encode encode_utf8
18 encodings find_encoding
21 our @FB_FLAGS = qw(DIE_ON_ERR WARN_ON_ERR RETURN_ON_ERR LEAVE_SRC
22 PERLQQ HTMLCREF XMLCREF);
23 our @FB_CONSTS = qw(FB_DEFAULT FB_CROAK FB_QUIET FB_WARN
24 FB_PERLQQ FB_HTMLCREF FB_XMLCREF);
29 _utf8_off _utf8_on define_encoding from_to is_16bit is_8bit
30 is_utf8 perlio_ok resolve_alias utf8_downgrade utf8_upgrade
32 @FB_FLAGS, @FB_CONSTS,
37 all => [ @EXPORT, @EXPORT_OK ],
38 fallbacks => [ @FB_CONSTS ],
39 fallback_all => [ @FB_CONSTS, @FB_FLAGS ],
42 # Documentation moved after __END__ for speed - NI-S
44 our $ON_EBCDIC = (ord("A") == 193);
48 # Make a %Encoding package variable to allow a certain amount of cheating
51 require Encode::Config;
52 eval { require Encode::ConfigLocal };
58 if (@_ and $_[0] eq ":all"){
59 %enc = ( %Encoding, %ExtModule );
62 for my $mod (map {m/::/o ? $_ : "Encode::$_" } @_){
64 for my $enc (keys %ExtModule){
65 $ExtModule{$enc} eq $mod and $enc{$enc} = $mod;
70 sort { lc $a cmp lc $b }
71 grep {!/^(?:Internal|Unicode|Guess)$/o} keys %enc;
75 my $obj = ref($_[0]) ? $_[0] : find_encoding($_[0]);
76 $obj->can("perlio_ok") and return $obj->perlio_ok();
77 return 0; # safety net
84 $Encoding{$name} = $obj;
86 define_alias($lc => $obj) unless $lc eq $name;
89 define_alias($alias, $obj);
96 my ($class, $name, $skip_external) = @_;
98 ref($name) && $name->can('new_sequence') and return $name;
99 exists $Encoding{$name} and return $Encoding{$name};
101 exists $Encoding{$lc} and return $Encoding{$lc};
103 my $oc = $class->find_alias($name);
104 defined($oc) and return $oc;
105 $lc ne $name and $oc = $class->find_alias($lc);
106 defined($oc) and return $oc;
108 unless ($skip_external)
110 if (my $mod = $ExtModule{$name} || $ExtModule{$lc}){
111 $mod =~ s,::,/,g ; $mod .= '.pm';
112 eval{ require $mod; };
113 exists $Encoding{$name} and return $Encoding{$name};
121 my ($name, $skip_external) = @_;
122 return __PACKAGE__->getEncoding($name,$skip_external);
126 my $obj = find_encoding(shift);
127 defined $obj and return $obj->name;
133 my ($name, $string, $check) = @_;
134 return undef unless defined $string;
136 my $enc = find_encoding($name);
137 unless(defined $enc){
139 Carp::croak("Unknown encoding '$name'");
141 my $octets = $enc->encode($string,$check);
142 return undef if ($check && length($string));
148 my ($name,$octets,$check) = @_;
149 return undef unless defined $octets;
151 my $enc = find_encoding($name);
152 unless(defined $enc){
154 Carp::croak("Unknown encoding '$name'");
156 my $string = $enc->decode($octets,$check);
157 $_[1] = $octets if $check;
163 my ($string,$from,$to,$check) = @_;
164 return undef unless defined $string;
166 my $f = find_encoding($from);
169 Carp::croak("Unknown encoding '$from'");
171 my $t = find_encoding($to);
174 Carp::croak("Unknown encoding '$to'");
176 my $uni = $f->decode($string,$check);
177 return undef if ($check && length($string));
178 $string = $t->encode($uni,$check);
179 return undef if ($check && length($uni));
180 return defined($_[0] = $string) ? length($string) : undef ;
193 return undef unless utf8::decode($str);
197 predefine_encodings(1);
200 # This is to restore %Encoding if really needed;
203 sub predefine_encodings{
204 use Encode::Encoding;
205 no warnings 'redefine';
208 # was in Encode::UTF_EBCDIC
209 package Encode::UTF_EBCDIC;
210 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
212 my ($obj,$str,$chk) = @_;
214 for (my $i = 0; $i < length($str); $i++) {
216 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
222 my ($obj,$str,$chk) = @_;
224 for (my $i = 0; $i < length($str); $i++) {
226 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
231 $Encode::Encoding{Unicode} =
232 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
234 package Encode::Internal;
235 push @Encode::Internal::ISA, 'Encode::Encoding';
237 my ($obj,$str,$chk) = @_;
243 $Encode::Encoding{Unicode} =
244 bless {Name => "Internal"} => "Encode::Internal";
248 # was in Encode::utf8
249 package Encode::utf8;
250 push @Encode::utf8::ISA, 'Encode::Encoding';
253 $DEBUG and warn __PACKAGE__, " XS on";
254 *decode = \&decode_xs;
255 *encode = \&encode_xs;
257 $DEBUG and warn __PACKAGE__, " XS off";
259 my ($obj,$octets,$chk) = @_;
260 my $str = Encode::decode_utf8($octets);
268 my ($obj,$string,$chk) = @_;
269 my $octets = Encode::encode_utf8($string);
274 *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk)
275 my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk
276 my ($rdst, $rsrc, $rpos) = \@_[1,2,3];
278 if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) {
279 $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm));
280 $$rpos = $npos + length($trm);
283 $$rdst .= substr($$rsrc, $pos);
284 $$rpos = length($$rsrc);
287 $Encode::Encoding{utf8} =
288 bless {Name => "utf8"} => "Encode::utf8";
298 Encode - character encodings
304 =head2 Table of Contents
306 Encode consists of a collection of modules whose details are too big
307 to fit in one document. This POD itself explains the top-level APIs
308 and general topics at a glance. For other topics and more details,
312 --------------------------------------------------------
313 Encode::Alias Alias definitions to encodings
314 Encode::Encoding Encode Implementation Base Class
315 Encode::Supported List of Supported Encodings
316 Encode::CN Simplified Chinese Encodings
317 Encode::JP Japanese Encodings
318 Encode::KR Korean Encodings
319 Encode::TW Traditional Chinese Encodings
320 --------------------------------------------------------
324 The C<Encode> module provides the interfaces between Perl's strings
325 and the rest of the system. Perl strings are sequences of
328 The repertoire of characters that Perl can represent is at least that
329 defined by the Unicode Consortium. On most platforms the ordinal
330 values of the characters (as returned by C<ord(ch)>) is the "Unicode
331 codepoint" for the character (the exceptions are those platforms where
332 the legacy encoding is some variant of EBCDIC rather than a super-set
333 of ASCII - see L<perlebcdic>).
335 Traditionally, computer data has been moved around in 8-bit chunks
336 often called "bytes". These chunks are also known as "octets" in
337 networking standards. Perl is widely used to manipulate data of many
338 types - not only strings of characters representing human or computer
339 languages but also "binary" data being the machine's representation of
340 numbers, pixels in an image - or just about anything.
342 When Perl is processing "binary data", the programmer wants Perl to
343 process "sequences of bytes". This is not a problem for Perl - as a
344 byte has 256 possible values, it easily fits in Perl's much larger
353 I<character>: a character in the range 0..(2**32-1) (or more).
354 (What Perl's strings are made of.)
358 I<byte>: a character in the range 0..255
359 (A special case of a Perl character.)
363 I<octet>: 8 bits of data, with ordinal values 0..255
364 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
368 =head1 PERL ENCODING API
372 =item $octets = encode(ENCODING, $string [, CHECK])
374 Encodes a string from Perl's internal form into I<ENCODING> and returns
375 a sequence of octets. ENCODING can be either a canonical name or
376 an alias. For encoding names and aliases, see L</"Defining Aliases">.
377 For CHECK, see L</"Handling Malformed Data">.
379 For example, to convert a string from Perl's internal format to
380 iso-8859-1 (also known as Latin1),
382 $octets = encode("iso-8859-1", $string);
384 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
385 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
386 for $octets is B<always> off. When you encode anything, utf8 flag of
387 the result is always off, even when it contains completely valid utf8
388 string. See L</"The UTF-8 flag"> below.
390 encode($valid_encoding, undef) is harmless but warns you for
391 C<Use of uninitialized value in subroutine entry>.
392 encode($valid_encoding, '') is harmless and warnless.
394 =item $string = decode(ENCODING, $octets [, CHECK])
396 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
397 internal form and returns the resulting string. As in encode(),
398 ENCODING can be either a canonical name or an alias. For encoding names
399 and aliases, see L</"Defining Aliases">. For CHECK, see
400 L</"Handling Malformed Data">.
402 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
404 $string = decode("iso-8859-1", $octets);
406 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
407 B<may not be equal to> $octets. Though they both contain the same data,
408 the utf8 flag for $string is on unless $octets entirely consists of
409 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
412 decode($valid_encoding, undef) is harmless but warns you for
413 C<Use of uninitialized value in subroutine entry>.
414 decode($valid_encoding, '') is harmless and warnless.
416 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
418 Converts B<in-place> data between two encodings. The data in $octets
419 must be encoded as octets and not as characters in Perl's internal
420 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 encoding:
422 from_to($octets, "iso-8859-1", "cp1250");
424 and to convert it back:
426 from_to($octets, "cp1250", "iso-8859-1");
428 Note that because the conversion happens in place, the data to be
429 converted cannot be a string constant; it must be a scalar variable.
431 from_to() returns the length of the converted string in octets on success, undef
434 B<CAVEAT>: The following operations look the same but are not quite so;
436 from_to($data, "iso-8859-1", "utf8"); #1
437 $data = decode("iso-8859-1", $data); #2
439 Both #1 and #2 make $data consist of a completely valid UTF-8 string
440 but only #2 turns utf8 flag on. #1 is equivalent to
442 $data = encode("utf8", decode("iso-8859-1", $data));
444 See L</"The UTF-8 flag"> below.
446 =item $octets = encode_utf8($string);
448 Equivalent to C<$octets = encode("utf8", $string);> The characters
449 that comprise $string are encoded in Perl's internal format and the
450 result is returned as a sequence of octets. All possible
451 characters have a UTF-8 representation so this function cannot fail.
454 =item $string = decode_utf8($octets [, CHECK]);
456 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
457 The sequence of octets represented by
458 $octets is decoded from UTF-8 into a sequence of logical
459 characters. Not all sequences of octets form valid UTF-8 encodings, so
460 it is possible for this call to fail. For CHECK, see
461 L</"Handling Malformed Data">.
465 =head2 Listing available encodings
468 @list = Encode->encodings();
470 Returns a list of the canonical names of the available encodings that
471 are loaded. To get a list of all available encodings including the
472 ones that are not loaded yet, say
474 @all_encodings = Encode->encodings(":all");
476 Or you can give the name of a specific module.
478 @with_jp = Encode->encodings("Encode::JP");
480 When "::" is not in the name, "Encode::" is assumed.
482 @ebcdic = Encode->encodings("EBCDIC");
484 To find out in detail which encodings are supported by this package,
485 see L<Encode::Supported>.
487 =head2 Defining Aliases
489 To add a new alias to a given encoding, use:
493 define_alias(newName => ENCODING);
495 After that, newName can be used as an alias for ENCODING.
496 ENCODING may be either the name of an encoding or an
499 But before you do so, make sure the alias is nonexistent with
500 C<resolve_alias()>, which returns the canonical name thereof.
503 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
504 Encode::resolve_alias("iso-8859-12") # false; nonexistent
505 Encode::resolve_alias($name) eq $name # true if $name is canonical
507 resolve_alias() does not need C<use Encode::Alias>; it can be
508 exported via C<use Encode qw(resolve_alias)>.
510 See L<Encode::Alias> for details.
512 =head1 Encoding via PerlIO
514 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
515 and encode directly via a filehandle. The following two examples
516 are totally identical in their functionality.
519 open my $in, "<:encoding(shiftjis)", $infile or die;
520 open my $out, ">:encoding(euc-jp)", $outfile or die;
521 while(<$in>){ print $out $_; }
524 open my $in, "<", $infile or die;
525 open my $out, ">", $outfile or die;
527 from_to($_, "shiftjis", "euc-jp", 1);
531 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
532 if your encoding is supported by PerlIO by calling the C<perlio_ok>
535 Encode::perlio_ok("hz"); # False
536 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
538 use Encode qw(perlio_ok); # exported upon request
541 Fortunately, all encodings that come with Encode core are PerlIO-savvy
542 except for hz and ISO-2022-kr. For gory details, see L<Encode::Encoding> and L<Encode::PerlIO>.
544 =head1 Handling Malformed Data
546 The I<CHECK> argument is used as follows. When you omit it,
547 the behaviour is the same as if you had passed a value of 0 for
552 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
554 If I<CHECK> is 0, (en|de)code will put a I<substitution character>
555 in place of a malformed character. For UCM-based encodings,
556 E<lt>subcharE<gt> will be used. For Unicode, the code point C<0xFFFD> is used.
557 If the data is supposed to be UTF-8, an optional lexical warning
558 (category utf8) is given.
560 =item I<CHECK> = Encode::FB_CROAK ( == 1)
562 If I<CHECK> is 1, methods will die on error immediately with an error
563 message. Therefore, when I<CHECK> is set to 1, you should trap the
564 fatal error with eval{} unless you really want to let it die on error.
566 =item I<CHECK> = Encode::FB_QUIET
568 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
569 return the portion of the data that has been processed so far when
570 an error occurs. The data argument will be overwritten with
571 everything after that point (that is, the unprocessed part of data).
572 This is handy when you have to call decode repeatedly in the case
573 where your source data may contain partial multi-byte character
574 sequences, for example because you are reading with a fixed-width
575 buffer. Here is some sample code that does exactly this:
577 my $data = ''; my $utf8 = '';
578 while(defined(read $fh, $buffer, 256)){
579 # buffer may end in a partial character so we append
581 $utf8 .= decode($encoding, $data, Encode::FB_QUIET);
582 # $data now contains the unprocessed partial character
585 =item I<CHECK> = Encode::FB_WARN
587 This is the same as above, except that it warns on error. Handy when
588 you are debugging the mode above.
590 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
592 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
594 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
596 For encodings that are implemented by Encode::XS, CHECK ==
597 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
599 When you decode, C<\xI<HH>> will be inserted for a malformed character,
600 where I<HH> is the hex representation of the octet that could not be
601 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
602 where I<HHHH> is the Unicode ID of the character that cannot be found
603 in the character repertoire of the encoding.
605 HTML/XML character reference modes are about the same, in place of
606 C<\x{I<HHHH>}>, HTML uses C<&#I<NNNN>>; where I<NNNN> is a decimal digit and
607 XML uses C<&#xI<HHHH>>; where I<HHHH> is the hexadecimal digit.
611 These modes are actually set via a bitmask. Here is how the FB_XX
612 constants are laid out. You can import the FB_XX constants via
613 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
614 constants via C<use Encode qw(:fallback_all)>.
616 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
619 RETURN_ON_ERR 0x0004 X X
627 =head2 Unimplemented fallback schemes
629 In the future, you will be able to use a code reference to a callback
630 function for the value of I<CHECK> but its API is still undecided.
632 The fallback scheme does not work on EBCDIC platforms.
634 =head1 Defining Encodings
636 To define a new encoding, use:
638 use Encode qw(define_encoding);
639 define_encoding($object, 'canonicalName' [, alias...]);
641 I<canonicalName> will be associated with I<$object>. The object
642 should provide the interface described in L<Encode::Encoding>.
643 If more than two arguments are provided then additional
644 arguments are taken as aliases for I<$object>.
646 See L<Encode::Encoding> for more details.
648 =head1 The UTF-8 flag
650 Before the introduction of utf8 support in perl, The C<eq> operator
651 just compared the strings represented by two scalars. Beginning with
652 perl 5.8, C<eq> compares two strings with simultaneous consideration
653 of I<the utf8 flag>. To explain why we made it so, I will quote page
654 402 of C<Programming Perl, 3rd ed.>
660 Old byte-oriented programs should not spontaneously break on the old
661 byte-oriented data they used to work on.
665 Old byte-oriented programs should magically start working on the new
666 character-oriented data when appropriate.
670 Programs should run just as fast in the new character-oriented mode
671 as in the old byte-oriented mode.
675 Perl should remain one language, rather than forking into a
676 byte-oriented Perl and a character-oriented Perl.
680 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
681 was born and many features documented in the book remained
682 unimplemented for a long time. Perl 5.8 corrected this and the introduction
683 of the UTF-8 flag is one of them. You can think of this perl notion as of a
684 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
687 Here is how Encode takes care of the utf8 flag.
693 When you encode, the resulting utf8 flag is always off.
697 When you decode, the resulting utf8 flag is on unless you can
698 unambiguously represent data. Here is the definition of
701 After C<$utf8 = decode('foo', $octet);>,
703 When $octet is... The utf8 flag in $utf8 is
704 ---------------------------------------------
705 In ASCII only (or EBCDIC only) OFF
707 In any other Encoding ON
708 ---------------------------------------------
710 As you see, there is one exception, In ASCII. That way you can assue
711 Goal #1. And with Encode Goal #2 is assumed but you still have to be
712 careful in such cases mentioned in B<CAVEAT> paragraphs.
714 This utf8 flag is not visible in perl scripts, exactly for the same
715 reason you cannot (or you I<don't have to>) see if a scalar contains a
716 string, integer, or floating point number. But you can still peek
717 and poke these if you will. See the section below.
721 =head2 Messing with Perl's Internals
723 The following API uses parts of Perl's internals in the current
724 implementation. As such, they are efficient but may change.
728 =item is_utf8(STRING [, CHECK])
730 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
731 If CHECK is true, also checks the data in STRING for being well-formed
732 UTF-8. Returns true if successful, false otherwise.
734 =item _utf8_on(STRING)
736 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
737 B<not> checked for being well-formed UTF-8. Do not use unless you
738 B<know> that the STRING is well-formed UTF-8. Returns the previous
739 state of the UTF-8 flag (so please don't treat the return value as
740 indicating success or failure), or C<undef> if STRING is not a string.
742 =item _utf8_off(STRING)
744 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
745 Returns the previous state of the UTF-8 flag (so please don't treat the
746 return value as indicating success or failure), or C<undef> if STRING is
754 L<Encode::Supported>,
761 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
765 This project was originated by Nick Ing-Simmons and later maintained
766 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
767 list of people involved. For any questions, use
768 E<lt>perl-unicode@perl.orgE<gt> so we can all share.