9 our @ISA = qw(Exporter DynaLoader);
11 # Public, encouraged API is exported by default
37 # Documentation moved after __END__ for speed - NI-S
41 # Make a %encoding package variable to allow a certain amount of cheating
43 my @alias; # ordered matching list
44 my %alias; # cached known aliases
46 # 0 1 2 3 4 5 6 7 8 9 10
47 our @latin2iso_num = ( 0, 1, 2, 3, 4, 9, 10, 13, 14, 15, 16 );
66 sort { $a->[1] cmp $b->[1] }
68 grep { $_ ne 'Internal' }
76 # print "# findAlias $_\n";
77 unless (exists $alias{$_})
79 for (my $i=0; $i < @alias; $i += 2)
81 my $alias = $alias[$i];
82 my $val = $alias[$i+1];
84 if (ref($alias) eq 'Regexp' && $_ =~ $alias)
88 elsif (ref($alias) eq 'CODE')
90 $new = &{$alias}($val)
92 elsif (lc($_) eq lc($alias))
98 next if $new eq $_; # avoid (direct) recursion on bugs
99 my $enc = (ref($new)) ? $new : find_encoding($new);
115 my ($alias,$name) = splice(@_,0,2);
116 push(@alias, $alias => $name);
120 # Allow variants of iso-8859-1 etc.
121 define_alias( qr/^iso[-_]?(\d+)[-_](\d+)$/i => '"iso-$1-$2"' );
123 # At least HP-UX has these.
124 define_alias( qr/^iso8859(\d+)$/i => '"iso-8859-$1"' );
127 define_alias( qr/^(?:hp-)?(arabic|greek|hebrew|kana|roman|thai|turkish)8$/i => '"${1}8"' );
129 # The Official name of ASCII.
130 define_alias( qr/^ANSI[-_]?X3\.4[-_]?1968$/i => '"ascii"' );
132 # This is a font issue, not an encoding issue.
133 # (The currency symbol of the Latin 1 upper half
134 # has been redefined as the euro symbol.)
135 define_alias( qr/^(.+)\@euro$/i => '"$1"' );
137 # Allow latin-1 style names as well
138 define_alias( qr/^(?:iso[-_]?)?latin[-_]?(\d+)$/i => '"iso-8859-$latin2iso_num[$1]"' );
140 # Allow winlatin1 style names as well
141 define_alias( qr/^win(latin[12]|cyrillic|baltic|greek|turkish|hebrew|arabic|baltic|vietnamese)$/i => '"cp$winlatin2cp{\u$1}"' );
143 # Common names for non-latin prefered MIME names
144 define_alias( 'ascii' => 'US-ascii',
145 'cyrillic' => 'iso-8859-5',
146 'arabic' => 'iso-8859-6',
147 'greek' => 'iso-8859-7',
148 'hebrew' => 'iso-8859-8',
149 'thai' => 'iso-8859-11',
150 'tis620' => 'iso-8859-11',
153 # At least AIX has IBM-NNN (surprisingly...) instead of cpNNN.
154 define_alias( qr/^ibm[-_]?(\d\d\d\d?)$/i => '"cp$1"');
156 # Standardize on the dashed versions.
157 define_alias( qr/^utf8$/i => 'utf-8' );
158 define_alias( qr/^koi8r$/i => 'koi8-r' );
159 define_alias( qr/^koi8u$/i => 'koi8-u' );
161 # TODO: HP-UX '8' encodings arabic8 greek8 hebrew8 kana8 thai8 turkish8
162 # TODO: HP-UX '15' encodings japanese15 korean15 roi15
163 # TODO: Cyrillic encoding ISO-IR-111 (useful?)
164 # TODO: Chinese encodings GB18030 GBK Big5-HSKCS EUC-TW
165 # TODO: Armenian encoding ARMSCII-8
166 # TODO: Hebrew encoding ISO-8859-8-1
167 # TODO: Thai encoding TCVN
168 # TODO: Korean encoding Johab
169 # TODO: Vietnamese encodings VPS
170 # TODO: Japanese encoding JIS (not the same as SJIS)
171 # TODO: Mac Asian+African encodings: Arabic Armenian Bengali Burmese
172 # ChineseSimp ChineseTrad Devanagari Ethiopic ExtArabic
173 # Farsi Georgian Gujarati Gurmukhi Hebrew Japanese
174 # Kannada Khmer Korean Laotian Malayalam Mongolian
175 # Oriya Sinhalese Symbol Tamil Telugu Tibetan Vietnamese
176 # TODO: what is the Japanese 'UJIS' encoding seen in some Linuxes?
178 # Map white space and _ to '-'
179 define_alias( qr/^(\S+)[\s_]+(.*)$/i => '"$1-$2"' );
185 $encoding{$name} = $obj;
187 define_alias($lc => $obj) unless $lc eq $name;
191 define_alias($alias,$obj);
198 my ($class,$name) = @_;
200 if (ref($name) && $name->can('new_sequence'))
205 if (exists $encoding{$name})
207 return $encoding{$name};
209 if (exists $encoding{$lc})
211 return $encoding{$lc};
214 my $oc = $class->findAlias($name);
215 return $oc if defined $oc;
216 return $class->findAlias($lc) if $lc ne $name;
224 return __PACKAGE__->getEncoding($name);
229 my ($name,$string,$check) = @_;
230 my $enc = find_encoding($name);
231 croak("Unknown encoding '$name'") unless defined $enc;
232 my $octets = $enc->encode($string,$check);
233 return undef if ($check && length($string));
239 my ($name,$octets,$check) = @_;
240 my $enc = find_encoding($name);
241 croak("Unknown encoding '$name'") unless defined $enc;
242 my $string = $enc->decode($octets,$check);
243 $_[1] = $octets if $check;
249 my ($string,$from,$to,$check) = @_;
250 my $f = find_encoding($from);
251 croak("Unknown encoding '$from'") unless defined $f;
252 my $t = find_encoding($to);
253 croak("Unknown encoding '$to'") unless defined $t;
254 my $uni = $f->decode($string,$check);
255 return undef if ($check && length($string));
256 $string = $t->encode($uni,$check);
257 return undef if ($check && length($uni));
258 return length($_[0] = $string);
271 return undef unless utf8::decode($str);
275 package Encode::Encoding;
276 # Base class for classes which implement encodings
281 my $canonical = shift;
282 $obj = bless { Name => $canonical },$obj unless ref $obj;
283 # warn "$canonical => $obj\n";
284 Encode::define_encoding($obj, $canonical, @_);
287 sub name { shift->{'Name'} }
289 # Temporary legacy methods
290 sub toUnicode { shift->decode(@_) }
291 sub fromUnicode { shift->encode(@_) }
293 sub new_sequence { return $_[0] }
296 use base 'Encode::Encoding';
298 package Encode::Internal;
299 use base 'Encode::Encoding';
301 # Dummy package that provides the encode interface but leaves data
302 # as UTF-X encoded. It is here so that from_to() works.
304 __PACKAGE__->Define('Internal');
306 Encode::define_alias( 'Unicode' => 'Internal' ) if ord('A') == 65;
310 my ($obj,$str,$chk) = @_;
318 package Encoding::Unicode;
319 use base 'Encode::Encoding';
321 __PACKAGE__->Define('Unicode') unless ord('A') == 65;
325 my ($obj,$str,$chk) = @_;
327 for (my $i = 0; $i < length($str); $i++)
329 $res .= chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
337 my ($obj,$str,$chk) = @_;
339 for (my $i = 0; $i < length($str); $i++)
341 $res .= chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
348 package Encode::utf8;
349 use base 'Encode::Encoding';
350 # package to allow long-hand
351 # $octets = encode( utf8 => $string );
354 __PACKAGE__->Define(qw(UTF-8 utf8));
358 my ($obj,$octets,$chk) = @_;
359 my $str = Encode::decode_utf8($octets);
370 my ($obj,$string,$chk) = @_;
371 my $octets = Encode::encode_utf8($string);
376 package Encode::iso10646_1;
377 use base 'Encode::Encoding';
378 # Encoding is 16-bit network order Unicode (no surogates)
379 # Used for X font encodings
381 __PACKAGE__->Define(qw(UCS-2 iso-10646-1));
385 my ($obj,$str,$chk) = @_;
389 my $code = unpack('n',substr($str,0,2,'')) & 0xffff;
392 $_[1] = $str if $chk;
399 my ($obj,$uni,$chk) = @_;
403 my $ch = substr($uni,0,1,'');
410 $str .= pack('n',$x);
412 $_[1] = $uni if $chk;
416 package Encode::ucs_2le;
417 use base 'Encode::Encoding';
419 __PACKAGE__->Define(qw(UCS-2le UCS-2LE ucs-2le));
423 my ($obj,$str,$chk) = @_;
427 my $code = unpack('v',substr($str,0,2,'')) & 0xffff;
430 $_[1] = $str if $chk;
437 my ($obj,$uni,$chk) = @_;
441 my $ch = substr($uni,0,1,'');
448 $str .= pack('v',$x);
450 $_[1] = $uni if $chk;
454 # switch back to Encode package in case we ever add AutoLoader
463 Encode - character encodings
471 The C<Encode> module provides the interfaces between Perl's strings
472 and the rest of the system. Perl strings are sequences of B<characters>.
474 The repertoire of characters that Perl can represent is at least that
475 defined by the Unicode Consortium. On most platforms the ordinal
476 values of the characters (as returned by C<ord(ch)>) is the "Unicode
477 codepoint" for the character (the exceptions are those platforms where
478 the legacy encoding is some variant of EBCDIC rather than a super-set
479 of ASCII - see L<perlebcdic>).
481 Traditionaly computer data has been moved around in 8-bit chunks
482 often called "bytes". These chunks are also known as "octets" in
483 networking standards. Perl is widely used to manipulate data of
484 many types - not only strings of characters representing human or
485 computer languages but also "binary" data being the machines representation
486 of numbers, pixels in an image - or just about anything.
488 When Perl is processing "binary data" the programmer wants Perl to process
489 "sequences of bytes". This is not a problem for Perl - as a byte has 256
490 possible values it easily fits in Perl's much larger "logical character".
498 I<character>: a character in the range 0..(2**32-1) (or more).
499 (What Perl's strings are made of.)
503 I<byte>: a character in the range 0..255
504 (A special case of a Perl character.)
508 I<octet>: 8 bits of data, with ordinal values 0..255
509 (Term for bytes passed to or from a non-Perl context, e.g. disk file.)
513 The marker [INTERNAL] marks Internal Implementation Details, in
514 general meant only for those who think they know what they are doing,
515 and such details may change in future releases.
519 =head2 Characteristics of an Encoding
521 An encoding has a "repertoire" of characters that it can represent,
522 and for each representable character there is at least one sequence of
523 octets that represents it.
525 =head2 Types of Encodings
527 Encodings can be divided into the following types:
531 =item * Fixed length 8-bit (or less) encodings.
533 Each character is a single octet so may have a repertoire of up to
534 256 characters. ASCII and iso-8859-* are typical examples.
536 =item * Fixed length 16-bit encodings
538 Each character is two octets so may have a repertoire of up to
539 65 536 characters. Unicode's UCS-2 is an example. Also used for
540 encodings for East Asian languages.
542 =item * Fixed length 32-bit encodings.
544 Not really very "encoded" encodings. The Unicode code points
545 are just represented as 4-octet integers. None the less because
546 different architectures use different representations of integers
547 (so called "endian") there at least two disctinct encodings.
549 =item * Multi-byte encodings
551 The number of octets needed to represent a character varies.
552 UTF-8 is a particularly complex but regular case of a multi-byte
553 encoding. Several East Asian countries use a multi-byte encoding
554 where 1-octet is used to cover western roman characters and Asian
555 characters get 2-octets.
556 (UTF-16 is strictly a multi-byte encoding taking either 2 or 4 octets
557 to represent a Unicode code point.)
559 =item * "Escape" encodings.
561 These encodings embed "escape sequences" into the octet sequence
562 which describe how the following octets are to be interpreted.
563 The iso-2022-* family is typical. Following the escape sequence
564 octets are encoded by an "embedded" encoding (which will be one
565 of the above types) until another escape sequence switches to
566 a different "embedded" encoding.
568 These schemes are very flexible and can handle mixed languages but are
569 very complex to process (and have state). No escape encodings are
570 implemented for Perl yet.
574 =head2 Specifying Encodings
576 Encodings can be specified to the API described below in two ways:
582 Encoding names are strings with characters taken from a restricted
583 repertoire. See L</"Encoding Names">.
585 =item 2. As an object
587 Encoding objects are returned by C<find_encoding($name)>.
591 =head2 Encoding Names
593 Encoding names are case insensitive. White space in names is ignored.
594 In addition an encoding may have aliases. Each encoding has one
595 "canonical" name. The "canonical" name is chosen from the names of
596 the encoding by picking the first in the following sequence:
600 =item * The MIME name as defined in IETF RFCs.
602 =item * The name in the IANA registry.
604 =item * The name used by the organization that defined it.
608 Because of all the alias issues, and because in the general case
609 encodings have state C<Encode> uses the encoding object internally
610 once an operation is in progress.
612 As of Perl 5.8.0, at least the following encodings are recognized
613 (the => marks aliases):
627 The ISO 8859 and KOI:
629 ISO 8859-1 ISO 8859-6 ISO 8859-11 KOI8-F
630 ISO 8859-2 ISO 8859-7 (12 doesn't exist) KOI8-R
631 ISO 8859-3 ISO 8859-8 ISO 8859-13 KOI8-U
632 ISO 8859-4 ISO 8859-9 ISO 8859-14
633 ISO 8859-5 ISO 8859-10 ISO 8859-15
636 Latin1 => 8859-1 Latin6 => 8859-10
637 Latin2 => 8859-2 Latin7 => 8859-13
638 Latin3 => 8859-3 Latin8 => 8859-14
639 Latin4 => 8859-4 Latin9 => 8859-15
640 Latin5 => 8859-9 Latin10 => 8859-16
649 The CJKV: Chinese, Japanese, Korean, Vietnamese:
651 ISO 2022 ISO 2022 JP-1 JIS 0201 GB 1988 Big5 EUC-CN
652 ISO 2022 CN ISO 2022 JP-2 JIS 0208 GB 2312 HZ EUC-JP
653 ISO 2022 JP ISO 2022 KR JIS 0210 GB 12345 CNS 11643 EUC-JP-0212
659 CP37 CP852 CP861 CP866 CP949 CP1251 CP1256
660 CP424 CP855 CP862 CP869 CP950 CP1252 CP1257
661 CP737 CP856 CP863 CP874 CP1006 CP1253 CP1258
662 CP775 CP857 CP864 CP932 CP1047 CP1254
663 CP850 CP860 CP865 CP936 CP1250 CP1255
667 WinCyrillic => CP1251
669 WinTurkiskh => CP1254
673 WinVietnamese => CP1258
675 (All the CPI<NNN...> are available also as IBMI<NNN...>.)
679 MacCentralEuropean MacJapanese
681 MacCyrillic MacRumanian
684 MacIcelandic MacTurkish
695 =head1 PERL ENCODING API
697 =head2 Generic Encoding Interface
703 $bytes = encode(ENCODING, $string[, CHECK])
705 Encodes string from Perl's internal form into I<ENCODING> and returns
706 a sequence of octets. For CHECK see L</"Handling Malformed Data">.
708 For example to convert (internally UTF-8 encoded) Unicode data
711 $octets = encode("utf8", $unicode);
715 $string = decode(ENCODING, $bytes[, CHECK])
717 Decode sequence of octets assumed to be in I<ENCODING> into Perl's
718 internal form and returns the resulting string. For CHECK see
719 L</"Handling Malformed Data">.
721 For example to convert ISO 8859-1 data to UTF-8:
723 $utf8 = decode("latin1", $latin1);
727 from_to($string, FROM_ENCODING, TO_ENCODING[, CHECK])
729 Convert B<in-place> the data between two encodings. How did the data
730 in $string originally get to be in FROM_ENCODING? Either using
731 encode() or through PerlIO: See L</"Encoding and IO">. For CHECK
732 see L</"Handling Malformed Data">.
734 For example to convert ISO 8859-1 data to UTF-8:
736 from_to($data, "iso-8859-1", "utf-8");
738 and to convert it back:
740 from_to($data, "utf-8", "iso-8859-1");
742 Note that because the conversion happens in place, the data to be
743 converted cannot be a string constant, it must be a scalar variable.
747 =head2 Handling Malformed Data
749 If CHECK is not set, C<undef> is returned. If the data is supposed to
750 be UTF-8, an optional lexical warning (category utf8) is given. If
751 CHECK is true but not a code reference, dies.
753 It would desirable to have a way to indicate that transform should use
754 the encodings "replacement character" - no such mechanism is defined yet.
756 It is also planned to allow I<CHECK> to be a code reference.
758 This is not yet implemented as there are design issues with what its
759 arguments should be and how it returns its results.
765 Passed remaining fragment of string being processed.
766 Modifies it in place to remove bytes/characters it can understand
767 and returns a string used to represent them.
771 my $ch = substr($_[0],0,1,'');
772 return sprintf("\x{%02X}",ord($ch);
775 This scheme is close to how underlying C code for Encode works, but gives
776 the fixup routine very little context.
780 Passed original string, and an index into it of the problem area, and
781 output string so far. Appends what it will to output string and
782 returns new index into original string. For example:
785 # my ($s,$i,$d) = @_;
786 my $ch = substr($_[0],$_[1],1);
787 $_[2] .= sprintf("\x{%02X}",ord($ch);
791 This scheme gives maximal control to the fixup routine but is more
792 complicated to code, and may need internals of Encode to be tweaked to
793 keep original string intact.
799 Multiple return values rather than in-place modifications.
801 Index into the string could be pos($str) allowing s/\G...//.
807 The Unicode consortium defines the UTF-8 standard as a way of encoding
808 the entire Unicode repertiore as sequences of octets. This encoding is
809 expected to become very widespread. Perl can use this form internaly
810 to represent strings, so conversions to and from this form are
811 particularly efficient (as octets in memory do not have to change,
812 just the meta-data that tells Perl how to treat them).
818 $bytes = encode_utf8($string);
820 The characters that comprise string are encoded in Perl's superset of UTF-8
821 and the resulting octets returned as a sequence of bytes. All possible
822 characters have a UTF-8 representation so this function cannot fail.
826 $string = decode_utf8($bytes [,CHECK]);
828 The sequence of octets represented by $bytes is decoded from UTF-8
829 into a sequence of logical characters. Not all sequences of octets
830 form valid UTF-8 encodings, so it is possible for this call to fail.
831 For CHECK see L</"Handling Malformed Data">.
835 =head2 Other Encodings of Unicode
837 UTF-16 is similar to UCS-2, 16 bit or 2-byte chunks. UCS-2 can only
838 represent 0..0xFFFF, while UTF-16 has a I<surrogate pair> scheme which
839 allows it to cover the whole Unicode range.
841 Surrogates are code points set aside to encode the 0x01000..0x10FFFF
842 range of Unicode code points in pairs of 16-bit units. The I<high
843 surrogates> are the range 0xD800..0xDBFF, and the I<low surrogates>
844 are the range 0xDC00..0xDFFFF. The surrogate encoding is
846 $hi = ($uni - 0x10000) / 0x400 + 0xD800;
847 $lo = ($uni - 0x10000) % 0x400 + 0xDC00;
851 $uni = 0x10000 + ($hi - 0xD8000) * 0x400 + ($lo - 0xDC00);
853 Encode implements big-endian UCS-2 aliased to "iso-10646-1" as that
854 happens to be the name used by that representation when used with X11
857 UTF-32 or UCS-4 is 32-bit or 4-byte chunks. Perl's logical characters
858 can be considered as being in this form without encoding. An encoding
859 to transfer strings in this form (e.g. to write them to a file) would
862 pack('L*', unpack('U*', $string)); # native
864 pack('V*', unpack('U*', $string)); # little-endian
866 pack('N*', unpack('U*', $string)); # big-endian
868 depending on the endianness required.
870 No UTF-32 encodings are implemented yet.
872 Both UCS-2 and UCS-4 style encodings can have "byte order marks" by
873 representing the code point 0xFFFE as the very first thing in a file.
875 =head2 Listing available encodings
877 use Encode qw(encodings);
880 Returns a list of the canonical names of the available encodings.
882 =head2 Defining Aliases
884 use Encode qw(define_alias);
885 define_alias( newName => ENCODING);
887 Allows newName to be used as am alias for ENCODING. ENCODING may be
888 either the name of an encoding or and encoding object (as above).
890 Currently I<newName> can be specified in the following ways:
894 =item As a simple string.
896 =item As a qr// compiled regular expression, e.g.:
898 define_alias( qr/^iso8859-(\d+)$/i => '"iso-8859-$1"' );
900 In this case if I<ENCODING> is not a reference it is C<eval>-ed to
901 allow C<$1> etc. to be subsituted. The example is one way to names as
902 used in X11 font names to alias the MIME names for the iso-8859-*
905 =item As a code reference, e.g.:
907 define_alias( sub { return /^iso8859-(\d+)$/i ? "iso-8859-$1" : undef } , '');
909 In this case C<$_> will be set to the name that is being looked up and
910 I<ENCODING> is passed to the sub as its first argument. The example
911 is another way to names as used in X11 font names to alias the MIME
912 names for the iso-8859-* family.
916 =head2 Defining Encodings
918 use Encode qw(define_alias);
919 define_encoding( $object, 'canonicalName' [,alias...]);
921 Causes I<canonicalName> to be associated with I<$object>. The object
922 should provide the interface described in L</"IMPLEMENTATION CLASSES">
923 below. If more than two arguments are provided then additional
924 arguments are taken as aliases for I<$object> as for C<define_alias>.
926 =head1 Encoding and IO
928 It is very common to want to do encoding transformations when
929 reading or writing files, network connections, pipes etc.
930 If Perl is configured to use the new 'perlio' IO system then
931 C<Encode> provides a "layer" (See L<perliol>) which can transform
932 data as it is read or written.
934 Here is how the blind poet would modernise the encoding:
937 open(my $iliad,'<:encoding(iso-8859-7)','iliad.greek');
938 open(my $utf8,'>:utf8','iliad.utf8');
944 In addition the new IO system can also be configured to read/write
945 UTF-8 encoded characters (as noted above this is efficient):
947 open(my $fh,'>:utf8','anything');
948 print $fh "Any \x{0021} string \N{SMILEY FACE}\n";
950 Either of the above forms of "layer" specifications can be made the default
951 for a lexical scope with the C<use open ...> pragma. See L<open>.
953 Once a handle is open is layers can be altered using C<binmode>.
955 Without any such configuration, or if Perl itself is built using
956 system's own IO, then write operations assume that file handle accepts
957 only I<bytes> and will C<die> if a character larger than 255 is
958 written to the handle. When reading, each octet from the handle
959 becomes a byte-in-a-character. Note that this default is the same
960 behaviour as bytes-only languages (including Perl before v5.6) would
961 have, and is sufficient to handle native 8-bit encodings
962 e.g. iso-8859-1, EBCDIC etc. and any legacy mechanisms for handling
963 other encodings and binary data.
965 In other cases it is the programs responsibility to transform
966 characters into bytes using the API above before doing writes, and to
967 transform the bytes read from a handle into characters before doing
968 "character operations" (e.g. C<lc>, C</\W+/>, ...).
970 You can also use PerlIO to convert larger amounts of data you don't
971 want to bring into memory. For example to convert between ISO 8859-1
972 (Latin 1) and UTF-8 (or UTF-EBCDIC in EBCDIC machines):
974 open(F, "<:encoding(iso-8859-1)", "data.txt") or die $!;
975 open(G, ">:utf8", "data.utf") or die $!;
976 while (<F>) { print G }
978 # Could also do "print G <F>" but that would pull
979 # the whole file into memory just to write it out again.
983 open(my $f, "<:encoding(cp1252)")
984 open(my $g, ">:encoding(iso-8859-2)")
985 open(my $h, ">:encoding(latin9)") # iso-8859-15
987 See L<PerlIO> for more information.
989 See also L<encoding> for how to change the default encoding of the
992 =head1 Encoding How to ...
998 =item * IO with mixed content (faking iso-2020-*)
1000 =item * MIME's Content-Length:
1002 =item * UTF-8 strings in binary data.
1004 =item * Perl/Encode wrappers on non-Unicode XS modules.
1008 =head1 Messing with Perl's Internals
1010 The following API uses parts of Perl's internals in the current
1011 implementation. As such they are efficient, but may change.
1015 =item * is_utf8(STRING [, CHECK])
1017 [INTERNAL] Test whether the UTF-8 flag is turned on in the STRING.
1018 If CHECK is true, also checks the data in STRING for being well-formed
1019 UTF-8. Returns true if successful, false otherwise.
1021 =item * valid_utf8(STRING)
1023 [INTERNAL] Test whether STRING is in a consistent state. Will return
1024 true if string is held as bytes, or is well-formed UTF-8 and has the
1025 UTF-8 flag on. Main reason for this routine is to allow Perl's
1026 testsuite to check that operations have left strings in a consistent
1033 [INTERNAL] Turn on the UTF-8 flag in STRING. The data in STRING is
1034 B<not> checked for being well-formed UTF-8. Do not use unless you
1035 B<know> that the STRING is well-formed UTF-8. Returns the previous
1036 state of the UTF-8 flag (so please don't test the return value as
1037 I<not> success or failure), or C<undef> if STRING is not a string.
1043 [INTERNAL] Turn off the UTF-8 flag in STRING. Do not use frivolously.
1044 Returns the previous state of the UTF-8 flag (so please don't test the
1045 return value as I<not> success or failure), or C<undef> if STRING is
1050 =head1 IMPLEMENTATION CLASSES
1052 As mentioned above encodings are (in the current implementation at least)
1053 defined by objects. The mapping of encoding name to object is via the
1056 The values of the hash can currently be either strings or objects.
1057 The string form may go away in the future. The string form occurs
1058 when C<encodings()> has scanned C<@INC> for loadable encodings but has
1059 not actually loaded the encoding in question. This is because the
1060 current "loading" process is all Perl and a bit slow.
1062 Once an encoding is loaded then value of the hash is object which
1063 implements the encoding. The object should provide the following
1070 Should return the string representing the canonical name of the encoding.
1072 =item -E<gt>new_sequence
1074 This is a placeholder for encodings with state. It should return an
1075 object which implements this interface, all current implementations
1076 return the original object.
1078 =item -E<gt>encode($string,$check)
1080 Should return the octet sequence representing I<$string>. If I<$check>
1081 is true it should modify I<$string> in place to remove the converted
1082 part (i.e. the whole string unless there is an error). If an error
1083 occurs it should return the octet sequence for the fragment of string
1084 that has been converted, and modify $string in-place to remove the
1085 converted part leaving it starting with the problem fragment.
1087 If check is is false then C<encode> should make a "best effort" to
1088 convert the string - for example by using a replacement character.
1090 =item -E<gt>decode($octets,$check)
1092 Should return the string that I<$octets> represents. If I<$check> is
1093 true it should modify I<$octets> in place to remove the converted part
1094 (i.e. the whole sequence unless there is an error). If an error
1095 occurs it should return the fragment of string that has been
1096 converted, and modify $octets in-place to remove the converted part
1097 leaving it starting with the problem fragment.
1099 If check is is false then C<decode> should make a "best effort" to
1100 convert the string - for example by using Unicode's "\x{FFFD}" as a
1101 replacement character.
1105 It should be noted that the check behaviour is different from the
1106 outer public API. The logic is that the "unchecked" case is useful
1107 when encoding is part of a stream which may be reporting errors
1108 (e.g. STDERR). In such cases it is desirable to get everything
1109 through somehow without causing additional errors which obscure the
1110 original one. Also the encoding is best placed to know what the
1111 correct replacement character is, so if that is the desired behaviour
1112 then letting low level code do it is the most efficient.
1114 In contrast if check is true, the scheme above allows the encoding to
1115 do as much as it can and tell layer above how much that was. What is
1116 lacking at present is a mechanism to report what went wrong. The most
1117 likely interface will be an additional method call to the object, or
1118 perhaps (to avoid forcing per-stream objects on otherwise stateless
1119 encodings) and additional parameter.
1121 It is also highly desirable that encoding classes inherit from
1122 C<Encode::Encoding> as a base class. This allows that class to define
1123 additional behaviour for all encoding objects. For example built in
1124 Unicode, UCS-2 and UTF-8 classes use :
1126 package Encode::MyEncoding;
1127 use base qw(Encode::Encoding);
1129 __PACKAGE__->Define(qw(myCanonical myAlias));
1131 To create an object with bless {Name => ...},$class, and call
1132 define_encoding. They inherit their C<name> method from
1133 C<Encode::Encoding>.
1135 =head2 Compiled Encodings
1137 F<Encode.xs> provides a class C<Encode::XS> which provides the
1138 interface described above. It calls a generic octet-sequence to
1139 octet-sequence "engine" that is driven by tables (defined in
1140 F<encengine.c>). The same engine is used for both encode and
1141 decode. C<Encode:XS>'s C<encode> forces Perl's characters to their
1142 UTF-8 form and then treats them as just another multibyte
1143 encoding. C<Encode:XS>'s C<decode> transforms the sequence and then
1144 turns the UTF-8-ness flag as that is the form that the tables are
1145 defined to produce. For details of the engine see the comments in
1148 The tables are produced by the Perl script F<compile> (the name needs
1149 to change so we can eventually install it somewhere). F<compile> can
1150 currently read two formats:
1156 This is a coined format used by Tcl. It is documented in
1157 Encode/EncodeFormat.pod.
1161 This is the semi-standard format used by IBM's ICU package.
1165 F<compile> can write the following forms:
1171 See above - the F<Encode/*.ucm> files provided with the distribution have
1172 been created from the original Tcl .enc files using this approach.
1176 Produces tables as C data structures - this is used to build in encodings
1177 into F<Encode.so>/F<Encode.dll>.
1181 In theory this allows encodings to be stand-alone loadable Perl
1182 extensions. The process has not yet been tested. The plan is to use
1183 this approach for large East Asian encodings.
1187 The set of encodings built-in to F<Encode.so>/F<Encode.dll> is
1188 determined by F<Makefile.PL>. The current set is as follows:
1192 =item ascii and iso-8859-*
1194 That is all the common 8-bit "western" encodings.
1196 =item IBM-1047 and two other variants of EBCDIC.
1198 These are the same variants that are supported by EBCDIC Perl as
1199 "native" encodings. They are included to prove "reversibility" of
1200 some constructs in EBCDIC Perl.
1202 =item symbol and dingbats as used by Tk on X11.
1204 (The reason Encode got started was to support Perl/Tk.)
1208 That set is rather ad hoc and has been driven by the needs of the
1209 tests rather than the needs of typical applications. It is likely
1214 L<perlunicode>, L<perlebcdic>, L<perlfunc/open>, L<PerlIO>, L<encoding>