2 # $Id: Encode.pm,v 1.70 2002/05/06 10:27:11 dankogai Exp dankogai $
6 our $VERSION = do { my @r = (q$Revision: 1.70 $ =~ /\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($str);
194 predefine_encodings();
197 # This is to restore %Encoding if really needed;
200 sub predefine_encodings{
201 use Encode::Encoding;
203 # was in Encode::UTF_EBCDIC
204 package Encode::UTF_EBCDIC;
205 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
207 my ($obj,$str,$chk) = @_;
209 for (my $i = 0; $i < length($str); $i++) {
211 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
217 my ($obj,$str,$chk) = @_;
219 for (my $i = 0; $i < length($str); $i++) {
221 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
226 $Encode::Encoding{Unicode} =
227 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
229 package Encode::Internal;
230 push @Encode::Internal::ISA, 'Encode::Encoding';
232 my ($obj,$str,$chk) = @_;
238 $Encode::Encoding{Unicode} =
239 bless {Name => "Internal"} => "Encode::Internal";
243 # was in Encode::utf8
244 package Encode::utf8;
245 push @Encode::utf8::ISA, 'Encode::Encoding';
247 my ($obj,$octets,$chk) = @_;
248 my $str = Encode::decode_utf8($octets);
256 my ($obj,$string,$chk) = @_;
257 my $octets = Encode::encode_utf8($string);
261 $Encode::Encoding{utf8} =
262 bless {Name => "utf8"} => "Encode::utf8";
272 Encode - character encodings
278 =head2 Table of Contents
280 Encode consists of a collection of modules whose details are too big
281 to fit in one document. This POD itself explains the top-level APIs
282 and general topics at a glance. For other topics and more details,
286 --------------------------------------------------------
287 Encode::Alias Alias definitions to encodings
288 Encode::Encoding Encode Implementation Base Class
289 Encode::Supported List of Supported Encodings
290 Encode::CN Simplified Chinese Encodings
291 Encode::JP Japanese Encodings
292 Encode::KR Korean Encodings
293 Encode::TW Traditional Chinese Encodings
294 --------------------------------------------------------
298 The C<Encode> module provides the interfaces between Perl's strings
299 and the rest of the system. Perl strings are sequences of
302 The repertoire of characters that Perl can represent is at least that
303 defined by the Unicode Consortium. On most platforms the ordinal
304 values of the characters (as returned by C<ord(ch)>) is the "Unicode
305 codepoint" for the character (the exceptions are those platforms where
306 the legacy encoding is some variant of EBCDIC rather than a super-set
307 of ASCII - see L<perlebcdic>).
309 Traditionally, computer data has been moved around in 8-bit chunks
310 often called "bytes". These chunks are also known as "octets" in
311 networking standards. Perl is widely used to manipulate data of many
312 types - not only strings of characters representing human or computer
313 languages but also "binary" data being the machine's representation of
314 numbers, pixels in an image - or just about anything.
316 When Perl is processing "binary data", the programmer wants Perl to
317 process "sequences of bytes". This is not a problem for Perl - as a
318 byte has 256 possible values, it easily fits in Perl's much larger
327 I<character>: a character in the range 0..(2**32-1) (or more).
328 (What Perl's strings are made of.)
332 I<byte>: a character in the range 0..255
333 (A special case of a Perl character.)
337 I<octet>: 8 bits of data, with ordinal values 0..255
338 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
342 =head1 PERL ENCODING API
346 =item $octets = encode(ENCODING, $string [, CHECK])
348 Encodes a string from Perl's internal form into I<ENCODING> and returns
349 a sequence of octets. ENCODING can be either a canonical name or
350 an alias. For encoding names and aliases, see L</"Defining Aliases">.
351 For CHECK, see L</"Handling Malformed Data">.
353 For example, to convert a string from Perl's internal format to
354 iso-8859-1 (also known as Latin1),
356 $octets = encode("iso-8859-1", $string);
358 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
359 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
360 for $octets is B<always> off. When you encode anything, utf8 flag of
361 the result is always off, even when it contains completely valid utf8
362 string. See L</"The UTF-8 flag"> below.
364 encode($valid_encoding, undef) is harmless but warns you for
365 C<Use of uninitialized value in subroutine entry>.
366 encode($valid_encoding, '') is harmless and warnless.
368 =item $string = decode(ENCODING, $octets [, CHECK])
370 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
371 internal form and returns the resulting string. As in encode(),
372 ENCODING can be either a canonical name or an alias. For encoding names
373 and aliases, see L</"Defining Aliases">. For CHECK, see
374 L</"Handling Malformed Data">.
376 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
378 $string = decode("iso-8859-1", $octets);
380 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
381 B<may not be equal to> $octets. Though they both contain the same data,
382 the utf8 flag for $string is on unless $octets entirely consists of
383 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
386 decode($valid_encoding, undef) is harmless but warns you for
387 C<Use of uninitialized value in subroutine entry>.
388 decode($valid_encoding, '') is harmless and warnless.
390 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
392 Converts B<in-place> data between two encodings. The data in $octets
393 must be encoded as octets and not as characters in Perl's internal
394 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 encoding:
396 from_to($octets, "iso-8859-1", "cp1250");
398 and to convert it back:
400 from_to($octets, "cp1250", "iso-8859-1");
402 Note that because the conversion happens in place, the data to be
403 converted cannot be a string constant; it must be a scalar variable.
405 from_to() returns the length of the converted string in octets on success, undef
408 B<CAVEAT>: The following operations look the same but are not quite so;
410 from_to($data, "iso-8859-1", "utf8"); #1
411 $data = decode("iso-8859-1", $data); #2
413 Both #1 and #2 make $data consist of a completely valid UTF-8 string
414 but only #2 turns utf8 flag on. #1 is equivalent to
416 $data = encode("utf8", decode("iso-8859-1", $data));
418 See L</"The UTF-8 flag"> below.
420 =item $octets = encode_utf8($string);
422 Equivalent to C<$octets = encode("utf8", $string);> The characters
423 that comprise $string are encoded in Perl's internal format and the
424 result is returned as a sequence of octets. All possible
425 characters have a UTF-8 representation so this function cannot fail.
428 =item $string = decode_utf8($octets [, CHECK]);
430 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
431 The sequence of octets represented by
432 $octets is decoded from UTF-8 into a sequence of logical
433 characters. Not all sequences of octets form valid UTF-8 encodings, so
434 it is possible for this call to fail. For CHECK, see
435 L</"Handling Malformed Data">.
439 =head2 Listing available encodings
442 @list = Encode->encodings();
444 Returns a list of the canonical names of the available encodings that
445 are loaded. To get a list of all available encodings including the
446 ones that are not loaded yet, say
448 @all_encodings = Encode->encodings(":all");
450 Or you can give the name of a specific module.
452 @with_jp = Encode->encodings("Encode::JP");
454 When "::" is not in the name, "Encode::" is assumed.
456 @ebcdic = Encode->encodings("EBCDIC");
458 To find out in detail which encodings are supported by this package,
459 see L<Encode::Supported>.
461 =head2 Defining Aliases
463 To add a new alias to a given encoding, use:
467 define_alias(newName => ENCODING);
469 After that, newName can be used as an alias for ENCODING.
470 ENCODING may be either the name of an encoding or an
473 But before you do so, make sure the alias is nonexistent with
474 C<resolve_alias()>, which returns the canonical name thereof.
477 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
478 Encode::resolve_alias("iso-8859-12") # false; nonexistent
479 Encode::resolve_alias($name) eq $name # true if $name is canonical
481 resolve_alias() does not need C<use Encode::Alias>; it can be
482 exported via C<use Encode qw(resolve_alias)>.
484 See L<Encode::Alias> for details.
486 =head1 Encoding via PerlIO
488 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
489 and encode directly via a filehandle. The following two examples
490 are totally identical in their functionality.
493 open my $in, "<:encoding(shiftjis)", $infile or die;
494 open my $out, ">:encoding(euc-jp)", $outfile or die;
495 while(<$in>){ print $out $_; }
498 open my $in, "<", $infile or die;
499 open my $out, ">", $outfile or die;
501 from_to($_, "shiftjis", "euc-jp", 1);
505 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
506 if your encoding is supported by PerlIO by calling the C<perlio_ok>
509 Encode::perlio_ok("hz"); # False
510 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
512 use Encode qw(perlio_ok); # exported upon request
515 Fortunately, all encodings that come with Encode core are PerlIO-savvy
516 except for hz and ISO-2022-kr. For gory details, see L<Encode::Encoding> and L<Encode::PerlIO>.
518 =head1 Handling Malformed Data
522 The I<CHECK> argument is used as follows. When you omit it,
523 the behaviour is the same as if you had passed a value of 0 for
526 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
528 If I<CHECK> is 0, (en|de)code will put a I<substitution character>
529 in place of a malformed character. For UCM-based encodings,
530 E<lt>subcharE<gt> will be used. For Unicode, the code point C<0xFFFD> is used.
531 If the data is supposed to be UTF-8, an optional lexical warning
532 (category utf8) is given.
534 =item I<CHECK> = Encode::FB_CROAK ( == 1)
536 If I<CHECK> is 1, methods will die on error immediately with an error
537 message. Therefore, when I<CHECK> is set to 1, you should trap the
538 fatal error with eval{} unless you really want to let it die on error.
540 =item I<CHECK> = Encode::FB_QUIET
542 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
543 return the portion of the data that has been processed so far when
544 an error occurs. The data argument will be overwritten with
545 everything after that point (that is, the unprocessed part of data).
546 This is handy when you have to call decode repeatedly in the case
547 where your source data may contain partial multi-byte character
548 sequences, for example because you are reading with a fixed-width
549 buffer. Here is some sample code that does exactly this:
551 my $data = ''; my $utf8 = '';
552 while(defined(read $fh, $buffer, 256)){
553 # buffer may end in a partial character so we append
555 $utf8 .= decode($encoding, $data, ENCODE::FB_QUIET);
556 # $data now contains the unprocessed partial character
559 =item I<CHECK> = Encode::FB_WARN
561 This is the same as above, except that it warns on error. Handy when
562 you are debugging the mode above.
564 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
566 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
568 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
570 For encodings that are implemented by Encode::XS, CHECK ==
571 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
573 When you decode, C<\xI<HH>> will be inserted for a malformed character,
574 where I<HH> is the hex representation of the octet that could not be
575 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
576 where I<HHHH> is the Unicode ID of the character that cannot be found
577 in the character repertoire of the encoding.
579 HTML/XML character reference modes are about the same, in place of
580 C<\x{I<HHHH>}>, HTML uses C<&#I<NNNN>>; where I<NNNN> is a decimal digit and
581 XML uses C<&#xI<HHHH>>; where I<HHHH> is the hexadecimal digit.
585 These modes are actually set via a bitmask. Here is how the FB_XX
586 constants are laid out. You can import the FB_XX constants via
587 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
588 constants via C<use Encode qw(:fallback_all)>.
590 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
593 RETURN_ON_ERR 0x0004 X X
599 =head2 Unimplemented fallback schemes
601 In the future, you will be able to use a code reference to a callback
602 function for the value of I<CHECK> but its API is still undecided.
604 The fallback scheme does not work on EBCDIC platforms.
606 =head1 Defining Encodings
608 To define a new encoding, use:
610 use Encode qw(define_encoding);
611 define_encoding($object, 'canonicalName' [, alias...]);
613 I<canonicalName> will be associated with I<$object>. The object
614 should provide the interface described in L<Encode::Encoding>.
615 If more than two arguments are provided then additional
616 arguments are taken as aliases for I<$object>.
618 See L<Encode::Encoding> for more details.
620 =head1 The UTF-8 flag
622 Before the introduction of utf8 support in perl, The C<eq> operator
623 just compared the strings represented by two scalars. Beginning with
624 perl 5.8, C<eq> compares two strings with simultaneous consideration
625 of I<the utf8 flag>. To explain why we made it so, I will quote page
626 402 of C<Programming Perl, 3rd ed.>
632 Old byte-oriented programs should not spontaneously break on the old
633 byte-oriented data they used to work on.
637 Old byte-oriented programs should magically start working on the new
638 character-oriented data when appropriate.
642 Programs should run just as fast in the new character-oriented mode
643 as in the old byte-oriented mode.
647 Perl should remain one language, rather than forking into a
648 byte-oriented Perl and a character-oriented Perl.
652 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
653 was born and many features documented in the book remained
654 unimplemented for a long time. Perl 5.8 corrected this and the introduction
655 of the UTF-8 flag is one of them. You can think of this perl notion as of a
656 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
659 Here is how Encode takes care of the utf8 flag.
665 When you encode, the resulting utf8 flag is always off.
669 When you decode, the resulting utf8 flag is on unless you can
670 unambiguously represent data. Here is the definition of
673 After C<$utf8 = decode('foo', $octet);>,
675 When $octet is... The utf8 flag in $utf8 is
676 ---------------------------------------------
677 In ASCII only (or EBCDIC only) OFF
679 In any other Encoding ON
680 ---------------------------------------------
682 As you see, there is one exception, In ASCII. That way you can assue
683 Goal #1. And with Encode Goal #2 is assumed but you still have to be
684 careful in such cases mentioned in B<CAVEAT> paragraphs.
686 This utf8 flag is not visible in perl scripts, exactly for the same
687 reason you cannot (or you I<don't have to>) see if a scalar contains a
688 string, integer, or floating point number. But you can still peek
689 and poke these if you will. See the section below.
693 =head2 Messing with Perl's Internals
695 The following API uses parts of Perl's internals in the current
696 implementation. As such, they are efficient but may change.
700 =item is_utf8(STRING [, CHECK])
702 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
703 If CHECK is true, also checks the data in STRING for being well-formed
704 UTF-8. Returns true if successful, false otherwise.
706 =item _utf8_on(STRING)
708 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
709 B<not> checked for being well-formed UTF-8. Do not use unless you
710 B<know> that the STRING is well-formed UTF-8. Returns the previous
711 state of the UTF-8 flag (so please don't treat the return value as
712 indicating success or failure), or C<undef> if STRING is not a string.
714 =item _utf8_off(STRING)
716 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
717 Returns the previous state of the UTF-8 flag (so please don't treat the
718 return value as indicating success or failure), or C<undef> if STRING is
726 L<Encode::Supported>,
733 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
737 This project was originated by Nick Ing-Simmons and later maintained
738 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
739 list of people involved. For any questions, use
740 E<lt>perl-unicode@perl.orgE<gt> so we can all share.