2 # $Id: Encode.pm,v 1.97 2003/07/08 21:52:14 dankogai Exp $
6 our $VERSION = do { my @r = (q$Revision: 1.97 $ =~ /\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 ;
201 return undef unless utf8::decode($str);
205 predefine_encodings(1);
208 # This is to restore %Encoding if really needed;
211 sub predefine_encodings{
212 use Encode::Encoding;
213 no warnings 'redefine';
216 # was in Encode::UTF_EBCDIC
217 package Encode::UTF_EBCDIC;
218 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
220 my ($obj,$str,$chk) = @_;
222 for (my $i = 0; $i < length($str); $i++) {
224 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
230 my ($obj,$str,$chk) = @_;
232 for (my $i = 0; $i < length($str); $i++) {
234 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
239 $Encode::Encoding{Unicode} =
240 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
242 package Encode::Internal;
243 push @Encode::Internal::ISA, 'Encode::Encoding';
245 my ($obj,$str,$chk) = @_;
251 $Encode::Encoding{Unicode} =
252 bless {Name => "Internal"} => "Encode::Internal";
256 # was in Encode::utf8
257 package Encode::utf8;
258 push @Encode::utf8::ISA, 'Encode::Encoding';
261 Encode::DEBUG and warn __PACKAGE__, " XS on";
262 *decode = \&decode_xs;
263 *encode = \&encode_xs;
265 Encode::DEBUG and warn __PACKAGE__, " XS off";
267 my ($obj,$octets,$chk) = @_;
268 my $str = Encode::decode_utf8($octets);
276 my ($obj,$string,$chk) = @_;
277 my $octets = Encode::encode_utf8($string);
282 *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk)
283 my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk
284 my ($rdst, $rsrc, $rpos) = \@_[1,2,3];
286 if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) {
287 $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm));
288 $$rpos = $npos + length($trm);
291 $$rdst .= substr($$rsrc, $pos);
292 $$rpos = length($$rsrc);
295 $Encode::Encoding{utf8} =
296 bless {Name => "utf8"} => "Encode::utf8";
306 Encode - character encodings
312 =head2 Table of Contents
314 Encode consists of a collection of modules whose details are too big
315 to fit in one document. This POD itself explains the top-level APIs
316 and general topics at a glance. For other topics and more details,
320 --------------------------------------------------------
321 Encode::Alias Alias definitions to encodings
322 Encode::Encoding Encode Implementation Base Class
323 Encode::Supported List of Supported Encodings
324 Encode::CN Simplified Chinese Encodings
325 Encode::JP Japanese Encodings
326 Encode::KR Korean Encodings
327 Encode::TW Traditional Chinese Encodings
328 --------------------------------------------------------
332 The C<Encode> module provides the interfaces between Perl's strings
333 and the rest of the system. Perl strings are sequences of
336 The repertoire of characters that Perl can represent is at least that
337 defined by the Unicode Consortium. On most platforms the ordinal
338 values of the characters (as returned by C<ord(ch)>) is the "Unicode
339 codepoint" for the character (the exceptions are those platforms where
340 the legacy encoding is some variant of EBCDIC rather than a super-set
341 of ASCII - see L<perlebcdic>).
343 Traditionally, computer data has been moved around in 8-bit chunks
344 often called "bytes". These chunks are also known as "octets" in
345 networking standards. Perl is widely used to manipulate data of many
346 types - not only strings of characters representing human or computer
347 languages but also "binary" data being the machine's representation of
348 numbers, pixels in an image - or just about anything.
350 When Perl is processing "binary data", the programmer wants Perl to
351 process "sequences of bytes". This is not a problem for Perl - as a
352 byte has 256 possible values, it easily fits in Perl's much larger
361 I<character>: a character in the range 0..(2**32-1) (or more).
362 (What Perl's strings are made of.)
366 I<byte>: a character in the range 0..255
367 (A special case of a Perl character.)
371 I<octet>: 8 bits of data, with ordinal values 0..255
372 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
376 =head1 PERL ENCODING API
380 =item $octets = encode(ENCODING, $string [, CHECK])
382 Encodes a string from Perl's internal form into I<ENCODING> and returns
383 a sequence of octets. ENCODING can be either a canonical name or
384 an alias. For encoding names and aliases, see L</"Defining Aliases">.
385 For CHECK, see L</"Handling Malformed Data">.
387 For example, to convert a string from Perl's internal format to
388 iso-8859-1 (also known as Latin1),
390 $octets = encode("iso-8859-1", $string);
392 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
393 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
394 for $octets is B<always> off. When you encode anything, utf8 flag of
395 the result is always off, even when it contains completely valid utf8
396 string. See L</"The UTF-8 flag"> below.
398 encode($valid_encoding, undef) is harmless but warns you for
399 C<Use of uninitialized value in subroutine entry>.
400 encode($valid_encoding, '') is harmless and warnless.
402 =item $string = decode(ENCODING, $octets [, CHECK])
404 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
405 internal form and returns the resulting string. As in encode(),
406 ENCODING can be either a canonical name or an alias. For encoding names
407 and aliases, see L</"Defining Aliases">. For CHECK, see
408 L</"Handling Malformed Data">.
410 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
412 $string = decode("iso-8859-1", $octets);
414 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
415 B<may not be equal to> $octets. Though they both contain the same data,
416 the utf8 flag for $string is on unless $octets entirely consists of
417 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
420 decode($valid_encoding, undef) is harmless but warns you for
421 C<Use of uninitialized value in subroutine entry>.
422 decode($valid_encoding, '') is harmless and warnless.
424 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
426 Converts B<in-place> data between two encodings. The data in $octets
427 must be encoded as octets and not as characters in Perl's internal
428 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 encoding:
430 from_to($octets, "iso-8859-1", "cp1250");
432 and to convert it back:
434 from_to($octets, "cp1250", "iso-8859-1");
436 Note that because the conversion happens in place, the data to be
437 converted cannot be a string constant; it must be a scalar variable.
439 from_to() returns the length of the converted string in octets on success, undef
442 B<CAVEAT>: The following operations look the same but are not quite so;
444 from_to($data, "iso-8859-1", "utf8"); #1
445 $data = decode("iso-8859-1", $data); #2
447 Both #1 and #2 make $data consist of a completely valid UTF-8 string
448 but only #2 turns utf8 flag on. #1 is equivalent to
450 $data = encode("utf8", decode("iso-8859-1", $data));
452 See L</"The UTF-8 flag"> below.
454 =item $octets = encode_utf8($string);
456 Equivalent to C<$octets = encode("utf8", $string);> The characters
457 that comprise $string are encoded in Perl's internal format and the
458 result is returned as a sequence of octets. All possible
459 characters have a UTF-8 representation so this function cannot fail.
462 =item $string = decode_utf8($octets [, CHECK]);
464 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
465 The sequence of octets represented by
466 $octets is decoded from UTF-8 into a sequence of logical
467 characters. Not all sequences of octets form valid UTF-8 encodings, so
468 it is possible for this call to fail. For CHECK, see
469 L</"Handling Malformed Data">.
473 =head2 Listing available encodings
476 @list = Encode->encodings();
478 Returns a list of the canonical names of the available encodings that
479 are loaded. To get a list of all available encodings including the
480 ones that are not loaded yet, say
482 @all_encodings = Encode->encodings(":all");
484 Or you can give the name of a specific module.
486 @with_jp = Encode->encodings("Encode::JP");
488 When "::" is not in the name, "Encode::" is assumed.
490 @ebcdic = Encode->encodings("EBCDIC");
492 To find out in detail which encodings are supported by this package,
493 see L<Encode::Supported>.
495 =head2 Defining Aliases
497 To add a new alias to a given encoding, use:
501 define_alias(newName => ENCODING);
503 After that, newName can be used as an alias for ENCODING.
504 ENCODING may be either the name of an encoding or an
507 But before you do so, make sure the alias is nonexistent with
508 C<resolve_alias()>, which returns the canonical name thereof.
511 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
512 Encode::resolve_alias("iso-8859-12") # false; nonexistent
513 Encode::resolve_alias($name) eq $name # true if $name is canonical
515 resolve_alias() does not need C<use Encode::Alias>; it can be
516 exported via C<use Encode qw(resolve_alias)>.
518 See L<Encode::Alias> for details.
520 =head1 Encoding via PerlIO
522 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
523 and encode directly via a filehandle. The following two examples
524 are totally identical in their functionality.
527 open my $in, "<:encoding(shiftjis)", $infile or die;
528 open my $out, ">:encoding(euc-jp)", $outfile or die;
529 while(<$in>){ print $out $_; }
532 open my $in, "<", $infile or die;
533 open my $out, ">", $outfile or die;
535 from_to($_, "shiftjis", "euc-jp", 1);
539 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
540 if your encoding is supported by PerlIO by calling the C<perlio_ok>
543 Encode::perlio_ok("hz"); # False
544 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
546 use Encode qw(perlio_ok); # exported upon request
549 Fortunately, all encodings that come with Encode core are PerlIO-savvy
550 except for hz and ISO-2022-kr. For gory details, see L<Encode::Encoding> and L<Encode::PerlIO>.
552 =head1 Handling Malformed Data
554 The I<CHECK> argument is used as follows. When you omit it,
555 the behaviour is the same as if you had passed a value of 0 for
560 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
562 If I<CHECK> is 0, (en|de)code will put a I<substitution character>
563 in place of a malformed character. For UCM-based encodings,
564 E<lt>subcharE<gt> will be used. For Unicode, the code point C<0xFFFD> is used.
565 If the data is supposed to be UTF-8, an optional lexical warning
566 (category utf8) is given.
568 =item I<CHECK> = Encode::FB_CROAK ( == 1)
570 If I<CHECK> is 1, methods will die on error immediately with an error
571 message. Therefore, when I<CHECK> is set to 1, you should trap the
572 fatal error with eval{} unless you really want to let it die on error.
574 =item I<CHECK> = Encode::FB_QUIET
576 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
577 return the portion of the data that has been processed so far when
578 an error occurs. The data argument will be overwritten with
579 everything after that point (that is, the unprocessed part of data).
580 This is handy when you have to call decode repeatedly in the case
581 where your source data may contain partial multi-byte character
582 sequences, for example because you are reading with a fixed-width
583 buffer. Here is some sample code that does exactly this:
585 my $data = ''; my $utf8 = '';
586 while(defined(read $fh, $buffer, 256)){
587 # buffer may end in a partial character so we append
589 $utf8 .= decode($encoding, $data, Encode::FB_QUIET);
590 # $data now contains the unprocessed partial character
593 =item I<CHECK> = Encode::FB_WARN
595 This is the same as above, except that it warns on error. Handy when
596 you are debugging the mode above.
598 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
600 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
602 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
604 For encodings that are implemented by Encode::XS, CHECK ==
605 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
607 When you decode, C<\xI<HH>> will be inserted for a malformed character,
608 where I<HH> is the hex representation of the octet that could not be
609 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
610 where I<HHHH> is the Unicode ID of the character that cannot be found
611 in the character repertoire of the encoding.
613 HTML/XML character reference modes are about the same, in place of
614 C<\x{I<HHHH>}>, HTML uses C<&#I<NNNN>>; where I<NNNN> is a decimal digit and
615 XML uses C<&#xI<HHHH>>; where I<HHHH> is the hexadecimal digit.
619 These modes are actually set via a bitmask. Here is how the FB_XX
620 constants are laid out. You can import the FB_XX constants via
621 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
622 constants via C<use Encode qw(:fallback_all)>.
624 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
627 RETURN_ON_ERR 0x0004 X X
635 =head2 Unimplemented fallback schemes
637 In the future, you will be able to use a code reference to a callback
638 function for the value of I<CHECK> but its API is still undecided.
640 The fallback scheme does not work on EBCDIC platforms.
642 =head1 Defining Encodings
644 To define a new encoding, use:
646 use Encode qw(define_encoding);
647 define_encoding($object, 'canonicalName' [, alias...]);
649 I<canonicalName> will be associated with I<$object>. The object
650 should provide the interface described in L<Encode::Encoding>.
651 If more than two arguments are provided then additional
652 arguments are taken as aliases for I<$object>.
654 See L<Encode::Encoding> for more details.
656 =head1 The UTF-8 flag
658 Before the introduction of utf8 support in perl, The C<eq> operator
659 just compared the strings represented by two scalars. Beginning with
660 perl 5.8, C<eq> compares two strings with simultaneous consideration
661 of I<the utf8 flag>. To explain why we made it so, I will quote page
662 402 of C<Programming Perl, 3rd ed.>
668 Old byte-oriented programs should not spontaneously break on the old
669 byte-oriented data they used to work on.
673 Old byte-oriented programs should magically start working on the new
674 character-oriented data when appropriate.
678 Programs should run just as fast in the new character-oriented mode
679 as in the old byte-oriented mode.
683 Perl should remain one language, rather than forking into a
684 byte-oriented Perl and a character-oriented Perl.
688 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
689 was born and many features documented in the book remained
690 unimplemented for a long time. Perl 5.8 corrected this and the introduction
691 of the UTF-8 flag is one of them. You can think of this perl notion as of a
692 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
695 Here is how Encode takes care of the utf8 flag.
701 When you encode, the resulting utf8 flag is always off.
705 When you decode, the resulting utf8 flag is on unless you can
706 unambiguously represent data. Here is the definition of
709 After C<$utf8 = decode('foo', $octet);>,
711 When $octet is... The utf8 flag in $utf8 is
712 ---------------------------------------------
713 In ASCII only (or EBCDIC only) OFF
715 In any other Encoding ON
716 ---------------------------------------------
718 As you see, there is one exception, In ASCII. That way you can assue
719 Goal #1. And with Encode Goal #2 is assumed but you still have to be
720 careful in such cases mentioned in B<CAVEAT> paragraphs.
722 This utf8 flag is not visible in perl scripts, exactly for the same
723 reason you cannot (or you I<don't have to>) see if a scalar contains a
724 string, integer, or floating point number. But you can still peek
725 and poke these if you will. See the section below.
729 =head2 Messing with Perl's Internals
731 The following API uses parts of Perl's internals in the current
732 implementation. As such, they are efficient but may change.
736 =item is_utf8(STRING [, CHECK])
738 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
739 If CHECK is true, also checks the data in STRING for being well-formed
740 UTF-8. Returns true if successful, false otherwise.
742 As of perl 5.8.1, L<utf8> also has utf8::is_utif8().
744 =item _utf8_on(STRING)
746 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
747 B<not> checked for being well-formed UTF-8. Do not use unless you
748 B<know> that the STRING is well-formed UTF-8. Returns the previous
749 state of the UTF-8 flag (so please don't treat the return value as
750 indicating success or failure), or C<undef> if STRING is not a string.
752 =item _utf8_off(STRING)
754 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
755 Returns the previous state of the UTF-8 flag (so please don't treat the
756 return value as indicating success or failure), or C<undef> if STRING is
764 L<Encode::Supported>,
771 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
775 This project was originated by Nick Ing-Simmons and later maintained
776 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
777 list of people involved. For any questions, use
778 E<lt>perl-unicode@perl.orgE<gt> so we can all share.