2 # $Id: Encode.pm,v 1.99 2003/12/29 02:47:16 dankogai Exp dankogai $
6 our $VERSION = do { my @r = (q$Revision: 1.99 $ =~ /\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 clone_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('renew') 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};
119 sub find_encoding($;$)
121 my ($name, $skip_external) = @_;
122 return __PACKAGE__->getEncoding($name,$skip_external);
125 sub resolve_alias($){
126 my $obj = find_encoding(shift);
127 defined $obj and return $obj->name;
131 sub clone_encoding($){
132 my $obj = find_encoding(shift);
134 eval { require Storable };
136 return Storable::dclone($obj);
141 my ($name, $string, $check) = @_;
142 return undef unless defined $string;
144 my $enc = find_encoding($name);
145 unless(defined $enc){
147 Carp::croak("Unknown encoding '$name'");
149 my $octets = $enc->encode($string,$check);
150 $_[1] = $string if $check;
156 my ($name,$octets,$check) = @_;
157 return undef unless defined $octets;
159 my $enc = find_encoding($name);
160 unless(defined $enc){
162 Carp::croak("Unknown encoding '$name'");
164 my $string = $enc->decode($octets,$check);
165 $_[1] = $octets if $check;
171 my ($string,$from,$to,$check) = @_;
172 return undef unless defined $string;
174 my $f = find_encoding($from);
177 Carp::croak("Unknown encoding '$from'");
179 my $t = find_encoding($to);
182 Carp::croak("Unknown encoding '$to'");
184 my $uni = $f->decode($string,$check);
185 return undef if ($check && length($string));
186 $string = $t->encode($uni,$check);
187 return undef if ($check && length($uni));
188 return defined($_[0] = $string) ? length($string) : undef ;
200 my ($str, $check) = @_;
202 return decode("utf8", $str, $check);
204 return undef unless utf8::decode($str);
209 predefine_encodings(1);
212 # This is to restore %Encoding if really needed;
215 sub predefine_encodings{
216 use Encode::Encoding;
217 no warnings 'redefine';
220 # was in Encode::UTF_EBCDIC
221 package Encode::UTF_EBCDIC;
222 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
224 my ($obj,$str,$chk) = @_;
226 for (my $i = 0; $i < length($str); $i++) {
228 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
234 my ($obj,$str,$chk) = @_;
236 for (my $i = 0; $i < length($str); $i++) {
238 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
243 $Encode::Encoding{Unicode} =
244 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
246 package Encode::Internal;
247 push @Encode::Internal::ISA, 'Encode::Encoding';
249 my ($obj,$str,$chk) = @_;
255 $Encode::Encoding{Unicode} =
256 bless {Name => "Internal"} => "Encode::Internal";
260 # was in Encode::utf8
261 package Encode::utf8;
262 push @Encode::utf8::ISA, 'Encode::Encoding';
265 Encode::DEBUG and warn __PACKAGE__, " XS on";
266 *decode = \&decode_xs;
267 *encode = \&encode_xs;
269 Encode::DEBUG and warn __PACKAGE__, " XS off";
271 my ($obj,$octets,$chk) = @_;
272 my $str = Encode::decode_utf8($octets);
280 my ($obj,$string,$chk) = @_;
281 my $octets = Encode::encode_utf8($string);
286 *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk)
287 my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk
288 my ($rdst, $rsrc, $rpos) = \@_[1,2,3];
290 if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) {
291 $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm));
292 $$rpos = $npos + length($trm);
295 $$rdst .= substr($$rsrc, $pos);
296 $$rpos = length($$rsrc);
299 $Encode::Encoding{utf8} =
300 bless {Name => "utf8"} => "Encode::utf8";
310 Encode - character encodings
316 =head2 Table of Contents
318 Encode consists of a collection of modules whose details are too big
319 to fit in one document. This POD itself explains the top-level APIs
320 and general topics at a glance. For other topics and more details,
324 --------------------------------------------------------
325 Encode::Alias Alias definitions to encodings
326 Encode::Encoding Encode Implementation Base Class
327 Encode::Supported List of Supported Encodings
328 Encode::CN Simplified Chinese Encodings
329 Encode::JP Japanese Encodings
330 Encode::KR Korean Encodings
331 Encode::TW Traditional Chinese Encodings
332 --------------------------------------------------------
336 The C<Encode> module provides the interfaces between Perl's strings
337 and the rest of the system. Perl strings are sequences of
340 The repertoire of characters that Perl can represent is at least that
341 defined by the Unicode Consortium. On most platforms the ordinal
342 values of the characters (as returned by C<ord(ch)>) is the "Unicode
343 codepoint" for the character (the exceptions are those platforms where
344 the legacy encoding is some variant of EBCDIC rather than a super-set
345 of ASCII - see L<perlebcdic>).
347 Traditionally, computer data has been moved around in 8-bit chunks
348 often called "bytes". These chunks are also known as "octets" in
349 networking standards. Perl is widely used to manipulate data of many
350 types - not only strings of characters representing human or computer
351 languages but also "binary" data being the machine's representation of
352 numbers, pixels in an image - or just about anything.
354 When Perl is processing "binary data", the programmer wants Perl to
355 process "sequences of bytes". This is not a problem for Perl - as a
356 byte has 256 possible values, it easily fits in Perl's much larger
365 I<character>: a character in the range 0..(2**32-1) (or more).
366 (What Perl's strings are made of.)
370 I<byte>: a character in the range 0..255
371 (A special case of a Perl character.)
375 I<octet>: 8 bits of data, with ordinal values 0..255
376 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
380 =head1 PERL ENCODING API
384 =item $octets = encode(ENCODING, $string [, CHECK])
386 Encodes a string from Perl's internal form into I<ENCODING> and returns
387 a sequence of octets. ENCODING can be either a canonical name or
388 an alias. For encoding names and aliases, see L</"Defining Aliases">.
389 For CHECK, see L</"Handling Malformed Data">.
391 For example, to convert a string from Perl's internal format to
392 iso-8859-1 (also known as Latin1),
394 $octets = encode("iso-8859-1", $string);
396 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
397 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
398 for $octets is B<always> off. When you encode anything, utf8 flag of
399 the result is always off, even when it contains completely valid utf8
400 string. See L</"The UTF-8 flag"> below.
402 encode($valid_encoding, undef) is harmless but warns you for
403 C<Use of uninitialized value in subroutine entry>.
404 encode($valid_encoding, '') is harmless and warnless.
406 =item $string = decode(ENCODING, $octets [, CHECK])
408 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
409 internal form and returns the resulting string. As in encode(),
410 ENCODING can be either a canonical name or an alias. For encoding names
411 and aliases, see L</"Defining Aliases">. For CHECK, see
412 L</"Handling Malformed Data">.
414 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
416 $string = decode("iso-8859-1", $octets);
418 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
419 B<may not be equal to> $octets. Though they both contain the same data,
420 the utf8 flag for $string is on unless $octets entirely consists of
421 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
424 decode($valid_encoding, undef) is harmless but warns you for
425 C<Use of uninitialized value in subroutine entry>.
426 decode($valid_encoding, '') is harmless and warnless.
428 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
430 Converts B<in-place> data between two encodings. The data in $octets
431 must be encoded as octets and not as characters in Perl's internal
432 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 encoding:
434 from_to($octets, "iso-8859-1", "cp1250");
436 and to convert it back:
438 from_to($octets, "cp1250", "iso-8859-1");
440 Note that because the conversion happens in place, the data to be
441 converted cannot be a string constant; it must be a scalar variable.
443 from_to() returns the length of the converted string in octets on success, undef
446 B<CAVEAT>: The following operations look the same but are not quite so;
448 from_to($data, "iso-8859-1", "utf8"); #1
449 $data = decode("iso-8859-1", $data); #2
451 Both #1 and #2 make $data consist of a completely valid UTF-8 string
452 but only #2 turns utf8 flag on. #1 is equivalent to
454 $data = encode("utf8", decode("iso-8859-1", $data));
456 See L</"The UTF-8 flag"> below.
458 =item $octets = encode_utf8($string);
460 Equivalent to C<$octets = encode("utf8", $string);> The characters
461 that comprise $string are encoded in Perl's internal format and the
462 result is returned as a sequence of octets. All possible
463 characters have a UTF-8 representation so this function cannot fail.
466 =item $string = decode_utf8($octets [, CHECK]);
468 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
469 The sequence of octets represented by
470 $octets is decoded from UTF-8 into a sequence of logical
471 characters. Not all sequences of octets form valid UTF-8 encodings, so
472 it is possible for this call to fail. For CHECK, see
473 L</"Handling Malformed Data">.
477 =head2 Listing available encodings
480 @list = Encode->encodings();
482 Returns a list of the canonical names of the available encodings that
483 are loaded. To get a list of all available encodings including the
484 ones that are not loaded yet, say
486 @all_encodings = Encode->encodings(":all");
488 Or you can give the name of a specific module.
490 @with_jp = Encode->encodings("Encode::JP");
492 When "::" is not in the name, "Encode::" is assumed.
494 @ebcdic = Encode->encodings("EBCDIC");
496 To find out in detail which encodings are supported by this package,
497 see L<Encode::Supported>.
499 =head2 Defining Aliases
501 To add a new alias to a given encoding, use:
505 define_alias(newName => ENCODING);
507 After that, newName can be used as an alias for ENCODING.
508 ENCODING may be either the name of an encoding or an
511 But before you do so, make sure the alias is nonexistent with
512 C<resolve_alias()>, which returns the canonical name thereof.
515 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
516 Encode::resolve_alias("iso-8859-12") # false; nonexistent
517 Encode::resolve_alias($name) eq $name # true if $name is canonical
519 resolve_alias() does not need C<use Encode::Alias>; it can be
520 exported via C<use Encode qw(resolve_alias)>.
522 See L<Encode::Alias> for details.
524 =head1 Encoding via PerlIO
526 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
527 and encode directly via a filehandle. The following two examples
528 are totally identical in their functionality.
531 open my $in, "<:encoding(shiftjis)", $infile or die;
532 open my $out, ">:encoding(euc-jp)", $outfile or die;
533 while(<$in>){ print $out $_; }
536 open my $in, "<", $infile or die;
537 open my $out, ">", $outfile or die;
539 from_to($_, "shiftjis", "euc-jp", 1);
543 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
544 if your encoding is supported by PerlIO by calling the C<perlio_ok>
547 Encode::perlio_ok("hz"); # False
548 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
550 use Encode qw(perlio_ok); # exported upon request
553 Fortunately, all encodings that come with Encode core are PerlIO-savvy
554 except for hz and ISO-2022-kr. For gory details, see L<Encode::Encoding> and L<Encode::PerlIO>.
556 =head1 Handling Malformed Data
558 The I<CHECK> argument is used as follows. When you omit it,
559 the behaviour is the same as if you had passed a value of 0 for
564 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
566 If I<CHECK> is 0, (en|de)code will put a I<substitution character>
567 in place of a malformed character. For UCM-based encodings,
568 E<lt>subcharE<gt> will be used. For Unicode, the code point C<0xFFFD> is used.
569 If the data is supposed to be UTF-8, an optional lexical warning
570 (category utf8) is given.
572 =item I<CHECK> = Encode::FB_CROAK ( == 1)
574 If I<CHECK> is 1, methods will die on error immediately with an error
575 message. Therefore, when I<CHECK> is set to 1, you should trap the
576 fatal error with eval{} unless you really want to let it die on error.
578 =item I<CHECK> = Encode::FB_QUIET
580 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
581 return the portion of the data that has been processed so far when
582 an error occurs. The data argument will be overwritten with
583 everything after that point (that is, the unprocessed part of data).
584 This is handy when you have to call decode repeatedly in the case
585 where your source data may contain partial multi-byte character
586 sequences, for example because you are reading with a fixed-width
587 buffer. Here is some sample code that does exactly this:
589 my $data = ''; my $utf8 = '';
590 while(defined(read $fh, $buffer, 256)){
591 # buffer may end in a partial character so we append
593 $utf8 .= decode($encoding, $data, Encode::FB_QUIET);
594 # $data now contains the unprocessed partial character
597 =item I<CHECK> = Encode::FB_WARN
599 This is the same as above, except that it warns on error. Handy when
600 you are debugging the mode above.
602 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
604 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
606 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
608 For encodings that are implemented by Encode::XS, CHECK ==
609 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
611 When you decode, C<\xI<HH>> will be inserted for a malformed character,
612 where I<HH> is the hex representation of the octet that could not be
613 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
614 where I<HHHH> is the Unicode ID of the character that cannot be found
615 in the character repertoire of the encoding.
617 HTML/XML character reference modes are about the same, in place of
618 C<\x{I<HHHH>}>, HTML uses C<&#I<NNNN>>; where I<NNNN> is a decimal digit and
619 XML uses C<&#xI<HHHH>>; where I<HHHH> is the hexadecimal digit.
623 These modes are actually set via a bitmask. Here is how the FB_XX
624 constants are laid out. You can import the FB_XX constants via
625 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
626 constants via C<use Encode qw(:fallback_all)>.
628 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
631 RETURN_ON_ERR 0x0004 X X
639 =head2 Unimplemented fallback schemes
641 In the future, you will be able to use a code reference to a callback
642 function for the value of I<CHECK> but its API is still undecided.
644 The fallback scheme does not work on EBCDIC platforms.
646 =head1 Defining Encodings
648 To define a new encoding, use:
650 use Encode qw(define_encoding);
651 define_encoding($object, 'canonicalName' [, alias...]);
653 I<canonicalName> will be associated with I<$object>. The object
654 should provide the interface described in L<Encode::Encoding>.
655 If more than two arguments are provided then additional
656 arguments are taken as aliases for I<$object>.
658 See L<Encode::Encoding> for more details.
660 =head1 The UTF-8 flag
662 Before the introduction of utf8 support in perl, The C<eq> operator
663 just compared the strings represented by two scalars. Beginning with
664 perl 5.8, C<eq> compares two strings with simultaneous consideration
665 of I<the utf8 flag>. To explain why we made it so, I will quote page
666 402 of C<Programming Perl, 3rd ed.>
672 Old byte-oriented programs should not spontaneously break on the old
673 byte-oriented data they used to work on.
677 Old byte-oriented programs should magically start working on the new
678 character-oriented data when appropriate.
682 Programs should run just as fast in the new character-oriented mode
683 as in the old byte-oriented mode.
687 Perl should remain one language, rather than forking into a
688 byte-oriented Perl and a character-oriented Perl.
692 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
693 was born and many features documented in the book remained
694 unimplemented for a long time. Perl 5.8 corrected this and the introduction
695 of the UTF-8 flag is one of them. You can think of this perl notion as of a
696 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
699 Here is how Encode takes care of the utf8 flag.
705 When you encode, the resulting utf8 flag is always off.
709 When you decode, the resulting utf8 flag is on unless you can
710 unambiguously represent data. Here is the definition of
713 After C<$utf8 = decode('foo', $octet);>,
715 When $octet is... The utf8 flag in $utf8 is
716 ---------------------------------------------
717 In ASCII only (or EBCDIC only) OFF
719 In any other Encoding ON
720 ---------------------------------------------
722 As you see, there is one exception, In ASCII. That way you can assue
723 Goal #1. And with Encode Goal #2 is assumed but you still have to be
724 careful in such cases mentioned in B<CAVEAT> paragraphs.
726 This utf8 flag is not visible in perl scripts, exactly for the same
727 reason you cannot (or you I<don't have to>) see if a scalar contains a
728 string, integer, or floating point number. But you can still peek
729 and poke these if you will. See the section below.
733 =head2 Messing with Perl's Internals
735 The following API uses parts of Perl's internals in the current
736 implementation. As such, they are efficient but may change.
740 =item is_utf8(STRING [, CHECK])
742 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
743 If CHECK is true, also checks the data in STRING for being well-formed
744 UTF-8. Returns true if successful, false otherwise.
746 As of perl 5.8.1, L<utf8> also has utf8::is_utif8().
748 =item _utf8_on(STRING)
750 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
751 B<not> checked for being well-formed UTF-8. Do not use unless you
752 B<know> that the STRING is well-formed UTF-8. Returns the previous
753 state of the UTF-8 flag (so please don't treat the return value as
754 indicating success or failure), or C<undef> if STRING is not a string.
756 =item _utf8_off(STRING)
758 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
759 Returns the previous state of the UTF-8 flag (so please don't treat the
760 return value as indicating success or failure), or C<undef> if STRING is
768 L<Encode::Supported>,
775 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
779 This project was originated by Nick Ing-Simmons and later maintained
780 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
781 list of people involved. For any questions, use
782 E<lt>perl-unicode@perl.orgE<gt> so we can all share.