2 # $Id: Encode.pm,v 1.83 2002/11/18 17:28:29 dankogai Exp $
6 our $VERSION = do { my @r = (q$Revision: 1.83 $ =~ /\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) = @_;
135 my $enc = find_encoding($name);
136 unless(defined $enc){
138 Carp::croak("Unknown encoding '$name'");
140 my $octets = $enc->encode($string,$check);
141 return undef if ($check && length($string));
147 my ($name,$octets,$check) = @_;
149 my $enc = find_encoding($name);
150 unless(defined $enc){
152 Carp::croak("Unknown encoding '$name'");
154 my $string = $enc->decode($octets,$check);
155 $_[1] = $octets if $check;
161 my ($string,$from,$to,$check) = @_;
163 my $f = find_encoding($from);
166 Carp::croak("Unknown encoding '$from'");
168 my $t = find_encoding($to);
171 Carp::croak("Unknown encoding '$to'");
173 my $uni = $f->decode($string,$check);
174 return undef if ($check && length($string));
175 $string = $t->encode($uni,$check);
176 return undef if ($check && length($uni));
177 return defined($_[0] = $string) ? length($string) : undef ;
190 return undef unless utf8::decode($str);
194 predefine_encodings(1);
197 # This is to restore %Encoding if really needed;
200 sub predefine_encodings{
201 use Encode::Encoding;
202 no warnings 'redefine';
205 # was in Encode::UTF_EBCDIC
206 package Encode::UTF_EBCDIC;
207 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
209 my ($obj,$str,$chk) = @_;
211 for (my $i = 0; $i < length($str); $i++) {
213 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
219 my ($obj,$str,$chk) = @_;
221 for (my $i = 0; $i < length($str); $i++) {
223 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
228 $Encode::Encoding{Unicode} =
229 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
231 package Encode::Internal;
232 push @Encode::Internal::ISA, 'Encode::Encoding';
234 my ($obj,$str,$chk) = @_;
240 $Encode::Encoding{Unicode} =
241 bless {Name => "Internal"} => "Encode::Internal";
245 # was in Encode::utf8
246 package Encode::utf8;
247 push @Encode::utf8::ISA, 'Encode::Encoding';
250 $DEBUG and warn __PACKAGE__, " XS on";
251 *decode = \&decode_xs;
252 *encode = \&encode_xs;
254 $DEBUG and warn __PACKAGE__, " XS off";
256 my ($obj,$octets,$chk) = @_;
257 my $str = Encode::decode_utf8($octets);
265 my ($obj,$string,$chk) = @_;
266 my $octets = Encode::encode_utf8($string);
271 $Encode::Encoding{utf8} =
272 bless {Name => "utf8"} => "Encode::utf8";
282 Encode - character encodings
288 =head2 Table of Contents
290 Encode consists of a collection of modules whose details are too big
291 to fit in one document. This POD itself explains the top-level APIs
292 and general topics at a glance. For other topics and more details,
296 --------------------------------------------------------
297 Encode::Alias Alias definitions to encodings
298 Encode::Encoding Encode Implementation Base Class
299 Encode::Supported List of Supported Encodings
300 Encode::CN Simplified Chinese Encodings
301 Encode::JP Japanese Encodings
302 Encode::KR Korean Encodings
303 Encode::TW Traditional Chinese Encodings
304 --------------------------------------------------------
308 The C<Encode> module provides the interfaces between Perl's strings
309 and the rest of the system. Perl strings are sequences of
312 The repertoire of characters that Perl can represent is at least that
313 defined by the Unicode Consortium. On most platforms the ordinal
314 values of the characters (as returned by C<ord(ch)>) is the "Unicode
315 codepoint" for the character (the exceptions are those platforms where
316 the legacy encoding is some variant of EBCDIC rather than a super-set
317 of ASCII - see L<perlebcdic>).
319 Traditionally, computer data has been moved around in 8-bit chunks
320 often called "bytes". These chunks are also known as "octets" in
321 networking standards. Perl is widely used to manipulate data of many
322 types - not only strings of characters representing human or computer
323 languages but also "binary" data being the machine's representation of
324 numbers, pixels in an image - or just about anything.
326 When Perl is processing "binary data", the programmer wants Perl to
327 process "sequences of bytes". This is not a problem for Perl - as a
328 byte has 256 possible values, it easily fits in Perl's much larger
337 I<character>: a character in the range 0..(2**32-1) (or more).
338 (What Perl's strings are made of.)
342 I<byte>: a character in the range 0..255
343 (A special case of a Perl character.)
347 I<octet>: 8 bits of data, with ordinal values 0..255
348 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
352 =head1 PERL ENCODING API
356 =item $octets = encode(ENCODING, $string [, CHECK])
358 Encodes a string from Perl's internal form into I<ENCODING> and returns
359 a sequence of octets. ENCODING can be either a canonical name or
360 an alias. For encoding names and aliases, see L</"Defining Aliases">.
361 For CHECK, see L</"Handling Malformed Data">.
363 For example, to convert a string from Perl's internal format to
364 iso-8859-1 (also known as Latin1),
366 $octets = encode("iso-8859-1", $string);
368 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
369 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
370 for $octets is B<always> off. When you encode anything, utf8 flag of
371 the result is always off, even when it contains completely valid utf8
372 string. See L</"The UTF-8 flag"> below.
374 encode($valid_encoding, undef) is harmless but warns you for
375 C<Use of uninitialized value in subroutine entry>.
376 encode($valid_encoding, '') is harmless and warnless.
378 =item $string = decode(ENCODING, $octets [, CHECK])
380 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
381 internal form and returns the resulting string. As in encode(),
382 ENCODING can be either a canonical name or an alias. For encoding names
383 and aliases, see L</"Defining Aliases">. For CHECK, see
384 L</"Handling Malformed Data">.
386 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
388 $string = decode("iso-8859-1", $octets);
390 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
391 B<may not be equal to> $octets. Though they both contain the same data,
392 the utf8 flag for $string is on unless $octets entirely consists of
393 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
396 decode($valid_encoding, undef) is harmless but warns you for
397 C<Use of uninitialized value in subroutine entry>.
398 decode($valid_encoding, '') is harmless and warnless.
400 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
402 Converts B<in-place> data between two encodings. The data in $octets
403 must be encoded as octets and not as characters in Perl's internal
404 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 encoding:
406 from_to($octets, "iso-8859-1", "cp1250");
408 and to convert it back:
410 from_to($octets, "cp1250", "iso-8859-1");
412 Note that because the conversion happens in place, the data to be
413 converted cannot be a string constant; it must be a scalar variable.
415 from_to() returns the length of the converted string in octets on success, undef
418 B<CAVEAT>: The following operations look the same but are not quite so;
420 from_to($data, "iso-8859-1", "utf8"); #1
421 $data = decode("iso-8859-1", $data); #2
423 Both #1 and #2 make $data consist of a completely valid UTF-8 string
424 but only #2 turns utf8 flag on. #1 is equivalent to
426 $data = encode("utf8", decode("iso-8859-1", $data));
428 See L</"The UTF-8 flag"> below.
430 =item $octets = encode_utf8($string);
432 Equivalent to C<$octets = encode("utf8", $string);> The characters
433 that comprise $string are encoded in Perl's internal format and the
434 result is returned as a sequence of octets. All possible
435 characters have a UTF-8 representation so this function cannot fail.
438 =item $string = decode_utf8($octets [, CHECK]);
440 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
441 The sequence of octets represented by
442 $octets is decoded from UTF-8 into a sequence of logical
443 characters. Not all sequences of octets form valid UTF-8 encodings, so
444 it is possible for this call to fail. For CHECK, see
445 L</"Handling Malformed Data">.
449 =head2 Listing available encodings
452 @list = Encode->encodings();
454 Returns a list of the canonical names of the available encodings that
455 are loaded. To get a list of all available encodings including the
456 ones that are not loaded yet, say
458 @all_encodings = Encode->encodings(":all");
460 Or you can give the name of a specific module.
462 @with_jp = Encode->encodings("Encode::JP");
464 When "::" is not in the name, "Encode::" is assumed.
466 @ebcdic = Encode->encodings("EBCDIC");
468 To find out in detail which encodings are supported by this package,
469 see L<Encode::Supported>.
471 =head2 Defining Aliases
473 To add a new alias to a given encoding, use:
477 define_alias(newName => ENCODING);
479 After that, newName can be used as an alias for ENCODING.
480 ENCODING may be either the name of an encoding or an
483 But before you do so, make sure the alias is nonexistent with
484 C<resolve_alias()>, which returns the canonical name thereof.
487 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
488 Encode::resolve_alias("iso-8859-12") # false; nonexistent
489 Encode::resolve_alias($name) eq $name # true if $name is canonical
491 resolve_alias() does not need C<use Encode::Alias>; it can be
492 exported via C<use Encode qw(resolve_alias)>.
494 See L<Encode::Alias> for details.
496 =head1 Encoding via PerlIO
498 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
499 and encode directly via a filehandle. The following two examples
500 are totally identical in their functionality.
503 open my $in, "<:encoding(shiftjis)", $infile or die;
504 open my $out, ">:encoding(euc-jp)", $outfile or die;
505 while(<$in>){ print $out $_; }
508 open my $in, "<", $infile or die;
509 open my $out, ">", $outfile or die;
511 from_to($_, "shiftjis", "euc-jp", 1);
515 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
516 if your encoding is supported by PerlIO by calling the C<perlio_ok>
519 Encode::perlio_ok("hz"); # False
520 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
522 use Encode qw(perlio_ok); # exported upon request
525 Fortunately, all encodings that come with Encode core are PerlIO-savvy
526 except for hz and ISO-2022-kr. For gory details, see L<Encode::Encoding> and L<Encode::PerlIO>.
528 =head1 Handling Malformed Data
532 The I<CHECK> argument is used as follows. When you omit it,
533 the behaviour is the same as if you had passed a value of 0 for
536 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
538 If I<CHECK> is 0, (en|de)code will put a I<substitution character>
539 in place of a malformed character. For UCM-based encodings,
540 E<lt>subcharE<gt> will be used. For Unicode, the code point C<0xFFFD> is used.
541 If the data is supposed to be UTF-8, an optional lexical warning
542 (category utf8) is given.
544 =item I<CHECK> = Encode::FB_CROAK ( == 1)
546 If I<CHECK> is 1, methods will die on error immediately with an error
547 message. Therefore, when I<CHECK> is set to 1, you should trap the
548 fatal error with eval{} unless you really want to let it die on error.
550 =item I<CHECK> = Encode::FB_QUIET
552 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
553 return the portion of the data that has been processed so far when
554 an error occurs. The data argument will be overwritten with
555 everything after that point (that is, the unprocessed part of data).
556 This is handy when you have to call decode repeatedly in the case
557 where your source data may contain partial multi-byte character
558 sequences, for example because you are reading with a fixed-width
559 buffer. Here is some sample code that does exactly this:
561 my $data = ''; my $utf8 = '';
562 while(defined(read $fh, $buffer, 256)){
563 # buffer may end in a partial character so we append
565 $utf8 .= decode($encoding, $data, Encode::FB_QUIET);
566 # $data now contains the unprocessed partial character
569 =item I<CHECK> = Encode::FB_WARN
571 This is the same as above, except that it warns on error. Handy when
572 you are debugging the mode above.
574 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
576 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
578 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
580 For encodings that are implemented by Encode::XS, CHECK ==
581 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
583 When you decode, C<\xI<HH>> will be inserted for a malformed character,
584 where I<HH> is the hex representation of the octet that could not be
585 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
586 where I<HHHH> is the Unicode ID of the character that cannot be found
587 in the character repertoire of the encoding.
589 HTML/XML character reference modes are about the same, in place of
590 C<\x{I<HHHH>}>, HTML uses C<&#I<NNNN>>; where I<NNNN> is a decimal digit and
591 XML uses C<&#xI<HHHH>>; where I<HHHH> is the hexadecimal digit.
595 These modes are actually set via a bitmask. Here is how the FB_XX
596 constants are laid out. You can import the FB_XX constants via
597 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
598 constants via C<use Encode qw(:fallback_all)>.
600 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
603 RETURN_ON_ERR 0x0004 X X
609 =head2 Unimplemented fallback schemes
611 In the future, you will be able to use a code reference to a callback
612 function for the value of I<CHECK> but its API is still undecided.
614 The fallback scheme does not work on EBCDIC platforms.
616 =head1 Defining Encodings
618 To define a new encoding, use:
620 use Encode qw(define_encoding);
621 define_encoding($object, 'canonicalName' [, alias...]);
623 I<canonicalName> will be associated with I<$object>. The object
624 should provide the interface described in L<Encode::Encoding>.
625 If more than two arguments are provided then additional
626 arguments are taken as aliases for I<$object>.
628 See L<Encode::Encoding> for more details.
630 =head1 The UTF-8 flag
632 Before the introduction of utf8 support in perl, The C<eq> operator
633 just compared the strings represented by two scalars. Beginning with
634 perl 5.8, C<eq> compares two strings with simultaneous consideration
635 of I<the utf8 flag>. To explain why we made it so, I will quote page
636 402 of C<Programming Perl, 3rd ed.>
642 Old byte-oriented programs should not spontaneously break on the old
643 byte-oriented data they used to work on.
647 Old byte-oriented programs should magically start working on the new
648 character-oriented data when appropriate.
652 Programs should run just as fast in the new character-oriented mode
653 as in the old byte-oriented mode.
657 Perl should remain one language, rather than forking into a
658 byte-oriented Perl and a character-oriented Perl.
662 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
663 was born and many features documented in the book remained
664 unimplemented for a long time. Perl 5.8 corrected this and the introduction
665 of the UTF-8 flag is one of them. You can think of this perl notion as of a
666 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
669 Here is how Encode takes care of the utf8 flag.
675 When you encode, the resulting utf8 flag is always off.
679 When you decode, the resulting utf8 flag is on unless you can
680 unambiguously represent data. Here is the definition of
683 After C<$utf8 = decode('foo', $octet);>,
685 When $octet is... The utf8 flag in $utf8 is
686 ---------------------------------------------
687 In ASCII only (or EBCDIC only) OFF
689 In any other Encoding ON
690 ---------------------------------------------
692 As you see, there is one exception, In ASCII. That way you can assue
693 Goal #1. And with Encode Goal #2 is assumed but you still have to be
694 careful in such cases mentioned in B<CAVEAT> paragraphs.
696 This utf8 flag is not visible in perl scripts, exactly for the same
697 reason you cannot (or you I<don't have to>) see if a scalar contains a
698 string, integer, or floating point number. But you can still peek
699 and poke these if you will. See the section below.
703 =head2 Messing with Perl's Internals
705 The following API uses parts of Perl's internals in the current
706 implementation. As such, they are efficient but may change.
710 =item is_utf8(STRING [, CHECK])
712 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
713 If CHECK is true, also checks the data in STRING for being well-formed
714 UTF-8. Returns true if successful, false otherwise.
716 =item _utf8_on(STRING)
718 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
719 B<not> checked for being well-formed UTF-8. Do not use unless you
720 B<know> that the STRING is well-formed UTF-8. Returns the previous
721 state of the UTF-8 flag (so please don't treat the return value as
722 indicating success or failure), or C<undef> if STRING is not a string.
724 =item _utf8_off(STRING)
726 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
727 Returns the previous state of the UTF-8 flag (so please don't treat the
728 return value as indicating success or failure), or C<undef> if STRING is
736 L<Encode::Supported>,
743 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
747 This project was originated by Nick Ing-Simmons and later maintained
748 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
749 list of people involved. For any questions, use
750 E<lt>perl-unicode@perl.orgE<gt> so we can all share.