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 };
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;
143 $string .= '' if ref $string; # stringify;
145 my $enc = find_encoding($name);
146 unless(defined $enc){
148 Carp::croak("Unknown encoding '$name'");
150 my $octets = $enc->encode($string,$check);
151 $_[1] = $string if $check;
157 my ($name,$octets,$check) = @_;
158 return undef unless defined $octets;
159 $octets .= '' if ref $octets;
161 my $enc = find_encoding($name);
162 unless(defined $enc){
164 Carp::croak("Unknown encoding '$name'");
166 my $string = $enc->decode($octets,$check);
167 $_[1] = $octets if $check;
173 my ($string,$from,$to,$check) = @_;
174 return undef unless defined $string;
176 my $f = find_encoding($from);
179 Carp::croak("Unknown encoding '$from'");
181 my $t = find_encoding($to);
184 Carp::croak("Unknown encoding '$to'");
186 my $uni = $f->decode($string,$check);
187 return undef if ($check && length($string));
188 $string = $t->encode($uni,$check);
189 return undef if ($check && length($uni));
190 return defined($_[0] = $string) ? length($string) : undef ;
202 my ($str, $check) = @_;
204 return decode("utf8", $str, $check);
206 return undef unless utf8::decode($str);
211 predefine_encodings(1);
214 # This is to restore %Encoding if really needed;
217 sub predefine_encodings{
218 use Encode::Encoding;
219 no warnings 'redefine';
222 # was in Encode::UTF_EBCDIC
223 package Encode::UTF_EBCDIC;
224 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
226 my ($obj,$str,$chk) = @_;
228 for (my $i = 0; $i < length($str); $i++) {
230 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
236 my ($obj,$str,$chk) = @_;
238 for (my $i = 0; $i < length($str); $i++) {
240 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
245 $Encode::Encoding{Unicode} =
246 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
248 package Encode::Internal;
249 push @Encode::Internal::ISA, 'Encode::Encoding';
251 my ($obj,$str,$chk) = @_;
257 $Encode::Encoding{Unicode} =
258 bless {Name => "Internal"} => "Encode::Internal";
262 # was in Encode::utf8
263 package Encode::utf8;
264 push @Encode::utf8::ISA, 'Encode::Encoding';
267 Encode::DEBUG and warn __PACKAGE__, " XS on";
268 *decode = \&decode_xs;
269 *encode = \&encode_xs;
271 Encode::DEBUG and warn __PACKAGE__, " XS off";
273 my ($obj,$octets,$chk) = @_;
274 my $str = Encode::decode_utf8($octets);
282 my ($obj,$string,$chk) = @_;
283 my $octets = Encode::encode_utf8($string);
288 *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk)
289 my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk
290 my ($rdst, $rsrc, $rpos) = \@_[1,2,3];
292 if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) {
293 $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm));
294 $$rpos = $npos + length($trm);
297 $$rdst .= substr($$rsrc, $pos);
298 $$rpos = length($$rsrc);
301 $Encode::Encoding{utf8} =
302 bless {Name => "utf8"} => "Encode::utf8";
312 Encode - character encodings
318 =head2 Table of Contents
320 Encode consists of a collection of modules whose details are too big
321 to fit in one document. This POD itself explains the top-level APIs
322 and general topics at a glance. For other topics and more details,
326 --------------------------------------------------------
327 Encode::Alias Alias definitions to encodings
328 Encode::Encoding Encode Implementation Base Class
329 Encode::Supported List of Supported Encodings
330 Encode::CN Simplified Chinese Encodings
331 Encode::JP Japanese Encodings
332 Encode::KR Korean Encodings
333 Encode::TW Traditional Chinese Encodings
334 --------------------------------------------------------
338 The C<Encode> module provides the interfaces between Perl's strings
339 and the rest of the system. Perl strings are sequences of
342 The repertoire of characters that Perl can represent is at least that
343 defined by the Unicode Consortium. On most platforms the ordinal
344 values of the characters (as returned by C<ord(ch)>) is the "Unicode
345 codepoint" for the character (the exceptions are those platforms where
346 the legacy encoding is some variant of EBCDIC rather than a super-set
347 of ASCII - see L<perlebcdic>).
349 Traditionally, computer data has been moved around in 8-bit chunks
350 often called "bytes". These chunks are also known as "octets" in
351 networking standards. Perl is widely used to manipulate data of many
352 types - not only strings of characters representing human or computer
353 languages but also "binary" data being the machine's representation of
354 numbers, pixels in an image - or just about anything.
356 When Perl is processing "binary data", the programmer wants Perl to
357 process "sequences of bytes". This is not a problem for Perl - as a
358 byte has 256 possible values, it easily fits in Perl's much larger
367 I<character>: a character in the range 0..(2**32-1) (or more).
368 (What Perl's strings are made of.)
372 I<byte>: a character in the range 0..255
373 (A special case of a Perl character.)
377 I<octet>: 8 bits of data, with ordinal values 0..255
378 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
382 =head1 PERL ENCODING API
386 =item $octets = encode(ENCODING, $string [, CHECK])
388 Encodes a string from Perl's internal form into I<ENCODING> and returns
389 a sequence of octets. ENCODING can be either a canonical name or
390 an alias. For encoding names and aliases, see L</"Defining Aliases">.
391 For CHECK, see L</"Handling Malformed Data">.
393 For example, to convert a string from Perl's internal format to
394 iso-8859-1 (also known as Latin1),
396 $octets = encode("iso-8859-1", $string);
398 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
399 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
400 for $octets is B<always> off. When you encode anything, utf8 flag of
401 the result is always off, even when it contains completely valid utf8
402 string. See L</"The UTF-8 flag"> below.
404 If the $string is C<undef> or a reference then C<undef> is returned.
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 If the $string is C<undef> or a reference then C<undef> is returned.
426 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
428 Converts B<in-place> data between two encodings. The data in $octets
429 must be encoded as octets and not as characters in Perl's internal
430 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250
433 from_to($octets, "iso-8859-1", "cp1250");
435 and to convert it back:
437 from_to($octets, "cp1250", "iso-8859-1");
439 Note that because the conversion happens in place, the data to be
440 converted cannot be a string constant; it must be a scalar variable.
442 from_to() returns the length of the converted string in octets on
443 success, I<undef> on error.
445 B<CAVEAT>: The following operations look the same but are not quite so;
447 from_to($data, "iso-8859-1", "utf8"); #1
448 $data = decode("iso-8859-1", $data); #2
450 Both #1 and #2 make $data consist of a completely valid UTF-8 string
451 but only #2 turns utf8 flag on. #1 is equivalent to
453 $data = encode("utf8", decode("iso-8859-1", $data));
455 See L</"The UTF-8 flag"> below.
457 =item $octets = encode_utf8($string);
459 Equivalent to C<$octets = encode("utf8", $string);> The characters
460 that comprise $string are encoded in Perl's internal format and the
461 result is returned as a sequence of octets. All possible
462 characters have a UTF-8 representation so this function cannot fail.
465 =item $string = decode_utf8($octets [, CHECK]);
467 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
468 The sequence of octets represented by
469 $octets is decoded from UTF-8 into a sequence of logical
470 characters. Not all sequences of octets form valid UTF-8 encodings, so
471 it is possible for this call to fail. For CHECK, see
472 L</"Handling Malformed Data">.
476 =head2 Listing available encodings
479 @list = Encode->encodings();
481 Returns a list of the canonical names of the available encodings that
482 are loaded. To get a list of all available encodings including the
483 ones that are not loaded yet, say
485 @all_encodings = Encode->encodings(":all");
487 Or you can give the name of a specific module.
489 @with_jp = Encode->encodings("Encode::JP");
491 When "::" is not in the name, "Encode::" is assumed.
493 @ebcdic = Encode->encodings("EBCDIC");
495 To find out in detail which encodings are supported by this package,
496 see L<Encode::Supported>.
498 =head2 Defining Aliases
500 To add a new alias to a given encoding, use:
504 define_alias(newName => ENCODING);
506 After that, newName can be used as an alias for ENCODING.
507 ENCODING may be either the name of an encoding or an
510 But before you do so, make sure the alias is nonexistent with
511 C<resolve_alias()>, which returns the canonical name thereof.
514 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
515 Encode::resolve_alias("iso-8859-12") # false; nonexistent
516 Encode::resolve_alias($name) eq $name # true if $name is canonical
518 resolve_alias() does not need C<use Encode::Alias>; it can be
519 exported via C<use Encode qw(resolve_alias)>.
521 See L<Encode::Alias> for details.
523 =head1 Encoding via PerlIO
525 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
526 and encode directly via a filehandle. The following two examples
527 are totally identical in their functionality.
530 open my $in, "<:encoding(shiftjis)", $infile or die;
531 open my $out, ">:encoding(euc-jp)", $outfile or die;
532 while(<$in>){ print $out $_; }
535 open my $in, "<", $infile or die;
536 open my $out, ">", $outfile or die;
538 from_to($_, "shiftjis", "euc-jp", 1);
542 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
543 if your encoding is supported by PerlIO by calling the C<perlio_ok>
546 Encode::perlio_ok("hz"); # False
547 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
549 use Encode qw(perlio_ok); # exported upon request
552 Fortunately, all encodings that come with Encode core are PerlIO-savvy
553 except for hz and ISO-2022-kr. For gory details, see
554 L<Encode::Encoding> and L<Encode::PerlIO>.
556 =head1 Handling Malformed Data
558 The optional I<CHECK> argument is used as follows. When you omit it,
559 Encode::FB_DEFAULT ( == 0 ) is assumed.
563 =item B<NOTE:> Not all encoding suppport this feature
565 Some encodings ignore I<CHECK> argument. For example,
566 L<Encode::Unicode> ignores I<CHECK> and it always croaks on error.
570 Now here is the list of I<CHECK> values available
574 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
576 If I<CHECK> is 0, (en|de)code will put a I<substitution character> in
577 place of a malformed character. When you encode, E<lt>subcharE<gt>
578 will be used. When you decode the code point C<0xFFFD> is used. If
579 the data is supposed to be UTF-8, an optional lexical warning
580 (category utf8) is given.
582 =item I<CHECK> = Encode::FB_CROAK ( == 1)
584 If I<CHECK> is 1, methods will die on error immediately with an error
585 message. Therefore, when I<CHECK> is set to 1, you should trap the
586 error with eval{} unless you really want to let it die.
588 =item I<CHECK> = Encode::FB_QUIET
590 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
591 return the portion of the data that has been processed so far when an
592 error occurs. The data argument will be overwritten with everything
593 after that point (that is, the unprocessed part of data). This is
594 handy when you have to call decode repeatedly in the case where your
595 source data may contain partial multi-byte character sequences,
596 (i.e. you are reading with a fixed-width buffer). Here is a sample
597 code that does exactly this:
599 my $buffer = ''; my $string = '';
600 while(read $fh, $buffer, 256, length($buffer)){
601 $string .= decode($encoding, $buffer, Encode::FB_QUIET);
602 # $buffer now contains the unprocessed partial character
605 =item I<CHECK> = Encode::FB_WARN
607 This is the same as above, except that it warns on error. Handy when
608 you are debugging the mode above.
610 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
612 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
614 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
616 For encodings that are implemented by Encode::XS, CHECK ==
617 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
619 When you decode, C<\xI<HH>> will be inserted for a malformed character,
620 where I<HH> is the hex representation of the octet that could not be
621 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
622 where I<HHHH> is the Unicode ID of the character that cannot be found
623 in the character repertoire of the encoding.
625 HTML/XML character reference modes are about the same, in place of
626 C<\x{I<HHHH>}>, HTML uses C<&#I<NNN>;> where I<NNN> is a decimal number and
627 XML uses C<&#xI<HHHH>;> where I<HHHH> is the hexadecimal number.
631 These modes are actually set via a bitmask. Here is how the FB_XX
632 constants are laid out. You can import the FB_XX constants via
633 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
634 constants via C<use Encode qw(:fallback_all)>.
636 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
639 RETURN_ON_ERR 0x0004 X X
647 =head2 Unimplemented fallback schemes
649 In the future, you will be able to use a code reference to a callback
650 function for the value of I<CHECK> but its API is still undecided.
652 The fallback scheme does not work on EBCDIC platforms.
654 =head1 Defining Encodings
656 To define a new encoding, use:
658 use Encode qw(define_encoding);
659 define_encoding($object, 'canonicalName' [, alias...]);
661 I<canonicalName> will be associated with I<$object>. The object
662 should provide the interface described in L<Encode::Encoding>.
663 If more than two arguments are provided then additional
664 arguments are taken as aliases for I<$object>.
666 See L<Encode::Encoding> for more details.
668 =head1 The UTF-8 flag
670 Before the introduction of utf8 support in perl, The C<eq> operator
671 just compared the strings represented by two scalars. Beginning with
672 perl 5.8, C<eq> compares two strings with simultaneous consideration
673 of I<the utf8 flag>. To explain why we made it so, I will quote page
674 402 of C<Programming Perl, 3rd ed.>
680 Old byte-oriented programs should not spontaneously break on the old
681 byte-oriented data they used to work on.
685 Old byte-oriented programs should magically start working on the new
686 character-oriented data when appropriate.
690 Programs should run just as fast in the new character-oriented mode
691 as in the old byte-oriented mode.
695 Perl should remain one language, rather than forking into a
696 byte-oriented Perl and a character-oriented Perl.
700 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
701 was born and many features documented in the book remained
702 unimplemented for a long time. Perl 5.8 corrected this and the introduction
703 of the UTF-8 flag is one of them. You can think of this perl notion as of a
704 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
707 Here is how Encode takes care of the utf8 flag.
713 When you encode, the resulting utf8 flag is always off.
717 When you decode, the resulting utf8 flag is on unless you can
718 unambiguously represent data. Here is the definition of
721 After C<$utf8 = decode('foo', $octet);>,
723 When $octet is... The utf8 flag in $utf8 is
724 ---------------------------------------------
725 In ASCII only (or EBCDIC only) OFF
727 In any other Encoding ON
728 ---------------------------------------------
730 As you see, there is one exception, In ASCII. That way you can assue
731 Goal #1. And with Encode Goal #2 is assumed but you still have to be
732 careful in such cases mentioned in B<CAVEAT> paragraphs.
734 This utf8 flag is not visible in perl scripts, exactly for the same
735 reason you cannot (or you I<don't have to>) see if a scalar contains a
736 string, integer, or floating point number. But you can still peek
737 and poke these if you will. See the section below.
741 =head2 Messing with Perl's Internals
743 The following API uses parts of Perl's internals in the current
744 implementation. As such, they are efficient but may change.
748 =item is_utf8(STRING [, CHECK])
750 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
751 If CHECK is true, also checks the data in STRING for being well-formed
752 UTF-8. Returns true if successful, false otherwise.
754 As of perl 5.8.1, L<utf8> also has utf8::is_utf8().
756 =item _utf8_on(STRING)
758 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
759 B<not> checked for being well-formed UTF-8. Do not use unless you
760 B<know> that the STRING is well-formed UTF-8. Returns the previous
761 state of the UTF-8 flag (so please don't treat the return value as
762 indicating success or failure), or C<undef> if STRING is not a string.
764 =item _utf8_off(STRING)
766 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
767 Returns the previous state of the UTF-8 flag (so please don't treat the
768 return value as indicating success or failure), or C<undef> if STRING is
776 L<Encode::Supported>,
783 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
787 This project was originated by Nick Ing-Simmons and later maintained
788 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
789 list of people involved. For any questions, use
790 E<lt>perl-unicode@perl.orgE<gt> so we can all share.