2 # $Id: Encode.pm,v 1.80 2002/10/21 20:38:45 dankogai Exp $
6 our $VERSION = do { my @r = (q$Revision: 1.80 $ =~ /\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';
246 # encode and decode methods now in Encode.xs
247 $Encode::Encoding{utf8} =
248 bless {Name => "utf8"} => "Encode::utf8";
258 Encode - character encodings
264 =head2 Table of Contents
266 Encode consists of a collection of modules whose details are too big
267 to fit in one document. This POD itself explains the top-level APIs
268 and general topics at a glance. For other topics and more details,
272 --------------------------------------------------------
273 Encode::Alias Alias definitions to encodings
274 Encode::Encoding Encode Implementation Base Class
275 Encode::Supported List of Supported Encodings
276 Encode::CN Simplified Chinese Encodings
277 Encode::JP Japanese Encodings
278 Encode::KR Korean Encodings
279 Encode::TW Traditional Chinese Encodings
280 --------------------------------------------------------
284 The C<Encode> module provides the interfaces between Perl's strings
285 and the rest of the system. Perl strings are sequences of
288 The repertoire of characters that Perl can represent is at least that
289 defined by the Unicode Consortium. On most platforms the ordinal
290 values of the characters (as returned by C<ord(ch)>) is the "Unicode
291 codepoint" for the character (the exceptions are those platforms where
292 the legacy encoding is some variant of EBCDIC rather than a super-set
293 of ASCII - see L<perlebcdic>).
295 Traditionally, computer data has been moved around in 8-bit chunks
296 often called "bytes". These chunks are also known as "octets" in
297 networking standards. Perl is widely used to manipulate data of many
298 types - not only strings of characters representing human or computer
299 languages but also "binary" data being the machine's representation of
300 numbers, pixels in an image - or just about anything.
302 When Perl is processing "binary data", the programmer wants Perl to
303 process "sequences of bytes". This is not a problem for Perl - as a
304 byte has 256 possible values, it easily fits in Perl's much larger
313 I<character>: a character in the range 0..(2**32-1) (or more).
314 (What Perl's strings are made of.)
318 I<byte>: a character in the range 0..255
319 (A special case of a Perl character.)
323 I<octet>: 8 bits of data, with ordinal values 0..255
324 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
328 =head1 PERL ENCODING API
332 =item $octets = encode(ENCODING, $string [, CHECK])
334 Encodes a string from Perl's internal form into I<ENCODING> and returns
335 a sequence of octets. ENCODING can be either a canonical name or
336 an alias. For encoding names and aliases, see L</"Defining Aliases">.
337 For CHECK, see L</"Handling Malformed Data">.
339 For example, to convert a string from Perl's internal format to
340 iso-8859-1 (also known as Latin1),
342 $octets = encode("iso-8859-1", $string);
344 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
345 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
346 for $octets is B<always> off. When you encode anything, utf8 flag of
347 the result is always off, even when it contains completely valid utf8
348 string. See L</"The UTF-8 flag"> below.
350 encode($valid_encoding, undef) is harmless but warns you for
351 C<Use of uninitialized value in subroutine entry>.
352 encode($valid_encoding, '') is harmless and warnless.
354 =item $string = decode(ENCODING, $octets [, CHECK])
356 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
357 internal form and returns the resulting string. As in encode(),
358 ENCODING can be either a canonical name or an alias. For encoding names
359 and aliases, see L</"Defining Aliases">. For CHECK, see
360 L</"Handling Malformed Data">.
362 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
364 $string = decode("iso-8859-1", $octets);
366 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
367 B<may not be equal to> $octets. Though they both contain the same data,
368 the utf8 flag for $string is on unless $octets entirely consists of
369 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
372 decode($valid_encoding, undef) is harmless but warns you for
373 C<Use of uninitialized value in subroutine entry>.
374 decode($valid_encoding, '') is harmless and warnless.
376 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
378 Converts B<in-place> data between two encodings. The data in $octets
379 must be encoded as octets and not as characters in Perl's internal
380 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 encoding:
382 from_to($octets, "iso-8859-1", "cp1250");
384 and to convert it back:
386 from_to($octets, "cp1250", "iso-8859-1");
388 Note that because the conversion happens in place, the data to be
389 converted cannot be a string constant; it must be a scalar variable.
391 from_to() returns the length of the converted string in octets on success, undef
394 B<CAVEAT>: The following operations look the same but are not quite so;
396 from_to($data, "iso-8859-1", "utf8"); #1
397 $data = decode("iso-8859-1", $data); #2
399 Both #1 and #2 make $data consist of a completely valid UTF-8 string
400 but only #2 turns utf8 flag on. #1 is equivalent to
402 $data = encode("utf8", decode("iso-8859-1", $data));
404 See L</"The UTF-8 flag"> below.
406 =item $octets = encode_utf8($string);
408 Equivalent to C<$octets = encode("utf8", $string);> The characters
409 that comprise $string are encoded in Perl's internal format and the
410 result is returned as a sequence of octets. All possible
411 characters have a UTF-8 representation so this function cannot fail.
414 =item $string = decode_utf8($octets [, CHECK]);
416 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
417 The sequence of octets represented by
418 $octets is decoded from UTF-8 into a sequence of logical
419 characters. Not all sequences of octets form valid UTF-8 encodings, so
420 it is possible for this call to fail. For CHECK, see
421 L</"Handling Malformed Data">.
425 =head2 Listing available encodings
428 @list = Encode->encodings();
430 Returns a list of the canonical names of the available encodings that
431 are loaded. To get a list of all available encodings including the
432 ones that are not loaded yet, say
434 @all_encodings = Encode->encodings(":all");
436 Or you can give the name of a specific module.
438 @with_jp = Encode->encodings("Encode::JP");
440 When "::" is not in the name, "Encode::" is assumed.
442 @ebcdic = Encode->encodings("EBCDIC");
444 To find out in detail which encodings are supported by this package,
445 see L<Encode::Supported>.
447 =head2 Defining Aliases
449 To add a new alias to a given encoding, use:
453 define_alias(newName => ENCODING);
455 After that, newName can be used as an alias for ENCODING.
456 ENCODING may be either the name of an encoding or an
459 But before you do so, make sure the alias is nonexistent with
460 C<resolve_alias()>, which returns the canonical name thereof.
463 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
464 Encode::resolve_alias("iso-8859-12") # false; nonexistent
465 Encode::resolve_alias($name) eq $name # true if $name is canonical
467 resolve_alias() does not need C<use Encode::Alias>; it can be
468 exported via C<use Encode qw(resolve_alias)>.
470 See L<Encode::Alias> for details.
472 =head1 Encoding via PerlIO
474 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
475 and encode directly via a filehandle. The following two examples
476 are totally identical in their functionality.
479 open my $in, "<:encoding(shiftjis)", $infile or die;
480 open my $out, ">:encoding(euc-jp)", $outfile or die;
481 while(<$in>){ print $out $_; }
484 open my $in, "<", $infile or die;
485 open my $out, ">", $outfile or die;
487 from_to($_, "shiftjis", "euc-jp", 1);
491 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
492 if your encoding is supported by PerlIO by calling the C<perlio_ok>
495 Encode::perlio_ok("hz"); # False
496 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
498 use Encode qw(perlio_ok); # exported upon request
501 Fortunately, all encodings that come with Encode core are PerlIO-savvy
502 except for hz and ISO-2022-kr. For gory details, see L<Encode::Encoding> and L<Encode::PerlIO>.
504 =head1 Handling Malformed Data
508 The I<CHECK> argument is used as follows. When you omit it,
509 the behaviour is the same as if you had passed a value of 0 for
512 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
514 If I<CHECK> is 0, (en|de)code will put a I<substitution character>
515 in place of a malformed character. For UCM-based encodings,
516 E<lt>subcharE<gt> will be used. For Unicode, the code point C<0xFFFD> is used.
517 If the data is supposed to be UTF-8, an optional lexical warning
518 (category utf8) is given.
520 =item I<CHECK> = Encode::FB_CROAK ( == 1)
522 If I<CHECK> is 1, methods will die on error immediately with an error
523 message. Therefore, when I<CHECK> is set to 1, you should trap the
524 fatal error with eval{} unless you really want to let it die on error.
526 =item I<CHECK> = Encode::FB_QUIET
528 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
529 return the portion of the data that has been processed so far when
530 an error occurs. The data argument will be overwritten with
531 everything after that point (that is, the unprocessed part of data).
532 This is handy when you have to call decode repeatedly in the case
533 where your source data may contain partial multi-byte character
534 sequences, for example because you are reading with a fixed-width
535 buffer. Here is some sample code that does exactly this:
537 my $data = ''; my $utf8 = '';
538 while(defined(read $fh, $buffer, 256)){
539 # buffer may end in a partial character so we append
541 $utf8 .= decode($encoding, $data, Encode::FB_QUIET);
542 # $data now contains the unprocessed partial character
545 =item I<CHECK> = Encode::FB_WARN
547 This is the same as above, except that it warns on error. Handy when
548 you are debugging the mode above.
550 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
552 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
554 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
556 For encodings that are implemented by Encode::XS, CHECK ==
557 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
559 When you decode, C<\xI<HH>> will be inserted for a malformed character,
560 where I<HH> is the hex representation of the octet that could not be
561 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
562 where I<HHHH> is the Unicode ID of the character that cannot be found
563 in the character repertoire of the encoding.
565 HTML/XML character reference modes are about the same, in place of
566 C<\x{I<HHHH>}>, HTML uses C<&#I<NNNN>>; where I<NNNN> is a decimal digit and
567 XML uses C<&#xI<HHHH>>; where I<HHHH> is the hexadecimal digit.
571 These modes are actually set via a bitmask. Here is how the FB_XX
572 constants are laid out. You can import the FB_XX constants via
573 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
574 constants via C<use Encode qw(:fallback_all)>.
576 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
579 RETURN_ON_ERR 0x0004 X X
585 =head2 Unimplemented fallback schemes
587 In the future, you will be able to use a code reference to a callback
588 function for the value of I<CHECK> but its API is still undecided.
590 The fallback scheme does not work on EBCDIC platforms.
592 =head1 Defining Encodings
594 To define a new encoding, use:
596 use Encode qw(define_encoding);
597 define_encoding($object, 'canonicalName' [, alias...]);
599 I<canonicalName> will be associated with I<$object>. The object
600 should provide the interface described in L<Encode::Encoding>.
601 If more than two arguments are provided then additional
602 arguments are taken as aliases for I<$object>.
604 See L<Encode::Encoding> for more details.
606 =head1 The UTF-8 flag
608 Before the introduction of utf8 support in perl, The C<eq> operator
609 just compared the strings represented by two scalars. Beginning with
610 perl 5.8, C<eq> compares two strings with simultaneous consideration
611 of I<the utf8 flag>. To explain why we made it so, I will quote page
612 402 of C<Programming Perl, 3rd ed.>
618 Old byte-oriented programs should not spontaneously break on the old
619 byte-oriented data they used to work on.
623 Old byte-oriented programs should magically start working on the new
624 character-oriented data when appropriate.
628 Programs should run just as fast in the new character-oriented mode
629 as in the old byte-oriented mode.
633 Perl should remain one language, rather than forking into a
634 byte-oriented Perl and a character-oriented Perl.
638 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
639 was born and many features documented in the book remained
640 unimplemented for a long time. Perl 5.8 corrected this and the introduction
641 of the UTF-8 flag is one of them. You can think of this perl notion as of a
642 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
645 Here is how Encode takes care of the utf8 flag.
651 When you encode, the resulting utf8 flag is always off.
655 When you decode, the resulting utf8 flag is on unless you can
656 unambiguously represent data. Here is the definition of
659 After C<$utf8 = decode('foo', $octet);>,
661 When $octet is... The utf8 flag in $utf8 is
662 ---------------------------------------------
663 In ASCII only (or EBCDIC only) OFF
665 In any other Encoding ON
666 ---------------------------------------------
668 As you see, there is one exception, In ASCII. That way you can assue
669 Goal #1. And with Encode Goal #2 is assumed but you still have to be
670 careful in such cases mentioned in B<CAVEAT> paragraphs.
672 This utf8 flag is not visible in perl scripts, exactly for the same
673 reason you cannot (or you I<don't have to>) see if a scalar contains a
674 string, integer, or floating point number. But you can still peek
675 and poke these if you will. See the section below.
679 =head2 Messing with Perl's Internals
681 The following API uses parts of Perl's internals in the current
682 implementation. As such, they are efficient but may change.
686 =item is_utf8(STRING [, CHECK])
688 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
689 If CHECK is true, also checks the data in STRING for being well-formed
690 UTF-8. Returns true if successful, false otherwise.
692 =item _utf8_on(STRING)
694 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
695 B<not> checked for being well-formed UTF-8. Do not use unless you
696 B<know> that the STRING is well-formed UTF-8. Returns the previous
697 state of the UTF-8 flag (so please don't treat the return value as
698 indicating success or failure), or C<undef> if STRING is not a string.
700 =item _utf8_off(STRING)
702 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
703 Returns the previous state of the UTF-8 flag (so please don't treat the
704 return value as indicating success or failure), or C<undef> if STRING is
712 L<Encode::Supported>,
719 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
723 This project was originated by Nick Ing-Simmons and later maintained
724 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
725 list of people involved. For any questions, use
726 E<lt>perl-unicode@perl.orgE<gt> so we can all share.