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 );
65 return keys %encoding;
72 unless (exists $alias{$_})
74 for (my $i=0; $i < @alias; $i += 2)
76 my $alias = $alias[$i];
77 my $val = $alias[$i+1];
79 if (ref($alias) eq 'Regexp' && $_ =~ $alias)
83 elsif (ref($alias) eq 'CODE')
85 $new = &{$alias}($val)
87 elsif (lc($_) eq lc($alias))
93 next if $new eq $_; # avoid (direct) recursion on bugs
94 my $enc = (ref($new)) ? $new : find_encoding($new);
110 my ($alias,$name) = splice(@_,0,2);
111 push(@alias, $alias => $name);
115 # Allow variants of iso-8859-1 etc.
116 define_alias( qr/^iso[-_]?(\d+)[-_](\d+)$/i => '"iso-$1-$2"' );
118 # At least HP-UX has these.
119 define_alias( qr/^iso8859(\d+)$/i => '"iso-8859-$1"' );
122 define_alias( qr/^(?:hp-)?(arabic|greek|hebrew|kana|roman|thai|turkish)8$/i => '"${1}8"' );
124 # This is a font issue, not an encoding issue.
125 # (The currency symbol of the Latin 1 upper half
126 # has been redefined as the euro symbol.)
127 define_alias( qr/^(.+)\@euro$/i => '"$1"' );
129 # Allow latin-1 style names as well
130 define_alias( qr/^(?:iso[-_]?)?latin[-_]?(\d+)$/i => '"iso-8859-$latin2iso_num[$1]"' );
132 # Allow winlatin1 style names as well
133 define_alias( qr/^win(latin[12]|cyrillic|baltic|greek|turkish|hebrew|arabic|baltic|vietnamese)$/i => '"cp$winlatin2cp{\u$1}"' );
135 # Common names for non-latin prefered MIME names
136 define_alias( 'ascii' => 'US-ascii',
137 'cyrillic' => 'iso-8859-5',
138 'arabic' => 'iso-8859-6',
139 'greek' => 'iso-8859-7',
140 'hebrew' => 'iso-8859-8',
141 'thai' => 'iso-8859-11',
142 'tis620' => 'iso-8859-11',
145 # At least AIX has IBM-NNN (surprisingly...) instead of cpNNN.
146 define_alias( qr/^ibm[-_]?(\d\d\d\d?)$/i => '"cp$1"');
148 # Standardize on the dashed versions.
149 define_alias( qr/^utf8$/i => 'utf-8' );
150 define_alias( qr/^koi8r$/i => 'koi8-r' );
151 define_alias( qr/^koi8u$/i => 'koi8-u' );
153 # TODO: HP-UX '8' encodings arabic8 greek8 hebrew8 kana8 thai8 turkish8
154 # TODO: HP-UX '15' encodings japanese15 korean15 roi15
155 # TODO: Cyrillic encoding ISO-IR-111 (useful?)
156 # TODO: Chinese encodings GB18030 GBK Big5-HSKCS EUC-TW
157 # TODO: Armenian encoding ARMSCII-8
158 # TODO: Hebrew encoding ISO-8859-8-1
159 # TODO: Thai encoding TCVN
160 # TODO: Korean encoding Johab
161 # TODO: Vietnamese encodings VISCII VPS
162 # TODO: Japanese encoding JIS (not the same as SJIS)
163 # TODO: Mac Asian+African encodings: Arabic Armenian Bengali Burmese
164 # ChineseSimp ChineseTrad Devanagari Ethiopic ExtArabic
165 # Farsi Georgian Gujarati Gurmukhi Hebrew Japanese
166 # Kannada Khmer Korean Laotian Malayalam Mongolian
167 # Oriya Sinhalese Symbol Tamil Telugu Tibetan Vietnamese
168 # TODO: what is the Japanese 'UJIS' encoding seen in some Linuxes?
170 # Map white space and _ to '-'
171 define_alias( qr/^(\S+)[\s_]+(.*)$/i => '"$1-$2"' );
177 $encoding{$name} = $obj;
179 define_alias($lc => $obj) unless $lc eq $name;
183 define_alias($alias,$obj);
190 my ($class,$name) = @_;
192 if (ref($name) && $name->can('new_sequence'))
196 if (exists $encoding{$name})
198 return $encoding{$name};
202 return $class->findAlias($name);
209 return __PACKAGE__->getEncoding($name);
214 my ($name,$string,$check) = @_;
215 my $enc = find_encoding($name);
216 croak("Unknown encoding '$name'") unless defined $enc;
217 my $octets = $enc->encode($string,$check);
218 return undef if ($check && length($string));
224 my ($name,$octets,$check) = @_;
225 my $enc = find_encoding($name);
226 croak("Unknown encoding '$name'") unless defined $enc;
227 my $string = $enc->decode($octets,$check);
228 return undef if ($check && length($octets));
234 my ($string,$from,$to,$check) = @_;
235 my $f = find_encoding($from);
236 croak("Unknown encoding '$from'") unless defined $f;
237 my $t = find_encoding($to);
238 croak("Unknown encoding '$to'") unless defined $t;
239 my $uni = $f->decode($string,$check);
240 return undef if ($check && length($string));
241 $string = $t->encode($uni,$check);
242 return undef if ($check && length($uni));
243 return length($_[0] = $string);
256 return undef unless utf8::decode($str);
260 package Encode::Encoding;
261 # Base class for classes which implement encodings
266 my $canonical = shift;
267 $obj = bless { Name => $canonical },$obj unless ref $obj;
268 # warn "$canonical => $obj\n";
269 Encode::define_encoding($obj, $canonical, @_);
272 sub name { shift->{'Name'} }
274 # Temporary legacy methods
275 sub toUnicode { shift->decode(@_) }
276 sub fromUnicode { shift->encode(@_) }
278 sub new_sequence { return $_[0] }
281 use base 'Encode::Encoding';
283 package Encode::Internal;
284 use base 'Encode::Encoding';
286 # Dummy package that provides the encode interface but leaves data
287 # as UTF-X encoded. It is here so that from_to() works.
289 __PACKAGE__->Define('Internal');
291 Encode::define_alias( 'Unicode' => 'Internal' ) if ord('A') == 65;
295 my ($obj,$str,$chk) = @_;
303 package Encoding::Unicode;
304 use base 'Encode::Encoding';
306 __PACKAGE__->Define('Unicode') unless ord('A') == 65;
310 my ($obj,$str,$chk) = @_;
312 for (my $i = 0; $i < length($str); $i++)
314 $res .= chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
322 my ($obj,$str,$chk) = @_;
324 for (my $i = 0; $i < length($str); $i++)
326 $res .= chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
333 package Encode::utf8;
334 use base 'Encode::Encoding';
335 # package to allow long-hand
336 # $octets = encode( utf8 => $string );
339 __PACKAGE__->Define(qw(UTF-8 utf8));
343 my ($obj,$octets,$chk) = @_;
344 my $str = Encode::decode_utf8($octets);
355 my ($obj,$string,$chk) = @_;
356 my $octets = Encode::encode_utf8($string);
361 package Encode::iso10646_1;
362 use base 'Encode::Encoding';
363 # Encoding is 16-bit network order Unicode (no surogates)
364 # Used for X font encodings
366 __PACKAGE__->Define(qw(UCS-2 iso-10646-1));
370 my ($obj,$str,$chk) = @_;
374 my $code = unpack('n',substr($str,0,2,'')) & 0xffff;
377 $_[1] = $str if $chk;
384 my ($obj,$uni,$chk) = @_;
388 my $ch = substr($uni,0,1,'');
395 $str .= pack('n',$x);
397 $_[1] = $uni if $chk;
401 package Encode::ucs_2le;
402 use base 'Encode::Encoding';
404 __PACKAGE__->Define(qw(UCS-2le UCS-2LE ucs-2le));
408 my ($obj,$str,$chk) = @_;
412 my $code = unpack('v',substr($str,0,2,'')) & 0xffff;
415 $_[1] = $str if $chk;
422 my ($obj,$uni,$chk) = @_;
426 my $ch = substr($uni,0,1,'');
433 $str .= pack('v',$x);
435 $_[1] = $uni if $chk;
439 # switch back to Encode package in case we ever add AutoLoader
448 Encode - character encodings
456 The C<Encode> module provides the interfaces between Perl's strings
457 and the rest of the system. Perl strings are sequences of B<characters>.
459 The repertoire of characters that Perl can represent is at least that
460 defined by the Unicode Consortium. On most platforms the ordinal
461 values of the characters (as returned by C<ord(ch)>) is the "Unicode
462 codepoint" for the character (the exceptions are those platforms where
463 the legacy encoding is some variant of EBCDIC rather than a super-set
464 of ASCII - see L<perlebcdic>).
466 Traditionaly computer data has been moved around in 8-bit chunks
467 often called "bytes". These chunks are also known as "octets" in
468 networking standards. Perl is widely used to manipulate data of
469 many types - not only strings of characters representing human or
470 computer languages but also "binary" data being the machines representation
471 of numbers, pixels in an image - or just about anything.
473 When Perl is processing "binary data" the programmer wants Perl to process
474 "sequences of bytes". This is not a problem for Perl - as a byte has 256
475 possible values it easily fits in Perl's much larger "logical character".
483 I<character>: a character in the range 0..(2**32-1) (or more).
484 (What Perl's strings are made of.)
488 I<byte>: a character in the range 0..255
489 (A special case of a Perl character.)
493 I<octet>: 8 bits of data, with ordinal values 0..255
494 (Term for bytes passed to or from a non-Perl context, e.g. disk file.)
498 The marker [INTERNAL] marks Internal Implementation Details, in
499 general meant only for those who think they know what they are doing,
500 and such details may change in future releases.
504 =head2 Characteristics of an Encoding
506 An encoding has a "repertoire" of characters that it can represent,
507 and for each representable character there is at least one sequence of
508 octets that represents it.
510 =head2 Types of Encodings
512 Encodings can be divided into the following types:
516 =item * Fixed length 8-bit (or less) encodings.
518 Each character is a single octet so may have a repertoire of up to
519 256 characters. ASCII and iso-8859-* are typical examples.
521 =item * Fixed length 16-bit encodings
523 Each character is two octets so may have a repertoire of up to
524 65 536 characters. Unicode's UCS-2 is an example. Also used for
525 encodings for East Asian languages.
527 =item * Fixed length 32-bit encodings.
529 Not really very "encoded" encodings. The Unicode code points
530 are just represented as 4-octet integers. None the less because
531 different architectures use different representations of integers
532 (so called "endian") there at least two disctinct encodings.
534 =item * Multi-byte encodings
536 The number of octets needed to represent a character varies.
537 UTF-8 is a particularly complex but regular case of a multi-byte
538 encoding. Several East Asian countries use a multi-byte encoding
539 where 1-octet is used to cover western roman characters and Asian
540 characters get 2-octets.
541 (UTF-16 is strictly a multi-byte encoding taking either 2 or 4 octets
542 to represent a Unicode code point.)
544 =item * "Escape" encodings.
546 These encodings embed "escape sequences" into the octet sequence
547 which describe how the following octets are to be interpreted.
548 The iso-2022-* family is typical. Following the escape sequence
549 octets are encoded by an "embedded" encoding (which will be one
550 of the above types) until another escape sequence switches to
551 a different "embedded" encoding.
553 These schemes are very flexible and can handle mixed languages but are
554 very complex to process (and have state). No escape encodings are
555 implemented for Perl yet.
559 =head2 Specifying Encodings
561 Encodings can be specified to the API described below in two ways:
567 Encoding names are strings with characters taken from a restricted
568 repertoire. See L</"Encoding Names">.
570 =item 2. As an object
572 Encoding objects are returned by C<find_encoding($name)>.
576 =head2 Encoding Names
578 Encoding names are case insensitive. White space in names is ignored.
579 In addition an encoding may have aliases. Each encoding has one
580 "canonical" name. The "canonical" name is chosen from the names of
581 the encoding by picking the first in the following sequence:
585 =item * The MIME name as defined in IETF RFC-XXXX.
587 =item * The name in the IANA registry.
589 =item * The name used by the the organization that defined it.
593 Because of all the alias issues, and because in the general case
594 encodings have state C<Encode> uses the encoding object internally
595 once an operation is in progress.
597 =head1 PERL ENCODING API
599 =head2 Generic Encoding Interface
605 $bytes = encode(ENCODING, $string[, CHECK])
607 Encodes string from Perl's internal form into I<ENCODING> and returns
608 a sequence of octets. For CHECK see L</"Handling Malformed Data">.
612 $string = decode(ENCODING, $bytes[, CHECK])
614 Decode sequence of octets assumed to be in I<ENCODING> into Perl's
615 internal form and returns the resulting string. For CHECK see
616 L</"Handling Malformed Data">.
620 from_to($string, FROM_ENCODING, TO_ENCODING[, CHECK])
622 Convert B<in-place> the data between two encodings. How did the data
623 in $string originally get to be in FROM_ENCODING? Either using
624 encode() or through PerlIO: See L</"Encoding and IO">. For CHECK
625 see L</"Handling Malformed Data">.
627 For example to convert ISO 8859-1 data to UTF-8:
629 from_to($data, "iso-8859-1", "utf-8");
631 and to convert it back:
633 from_to($data, "utf-8", "iso-8859-1");
635 Note that because the conversion happens in place, the data to be
636 converted cannot be a string constant, it must be a scalar variable.
640 =head2 Handling Malformed Data
642 If CHECK is not set, C<undef> is returned. If the data is supposed to
643 be UTF-8, an optional lexical warning (category utf8) is given. If
644 CHECK is true but not a code reference, dies.
646 It would desirable to have a way to indicate that transform should use
647 the encodings "replacement character" - no such mechanism is defined yet.
649 It is also planned to allow I<CHECK> to be a code reference.
651 This is not yet implemented as there are design issues with what its
652 arguments should be and how it returns its results.
658 Passed remaining fragment of string being processed.
659 Modifies it in place to remove bytes/characters it can understand
660 and returns a string used to represent them.
664 my $ch = substr($_[0],0,1,'');
665 return sprintf("\x{%02X}",ord($ch);
668 This scheme is close to how underlying C code for Encode works, but gives
669 the fixup routine very little context.
673 Passed original string, and an index into it of the problem area, and
674 output string so far. Appends what it will to output string and
675 returns new index into original string. For example:
678 # my ($s,$i,$d) = @_;
679 my $ch = substr($_[0],$_[1],1);
680 $_[2] .= sprintf("\x{%02X}",ord($ch);
684 This scheme gives maximal control to the fixup routine but is more
685 complicated to code, and may need internals of Encode to be tweaked to
686 keep original string intact.
692 Multiple return values rather than in-place modifications.
694 Index into the string could be pos($str) allowing s/\G...//.
700 The Unicode consortium defines the UTF-8 standard as a way of encoding
701 the entire Unicode repertiore as sequences of octets. This encoding is
702 expected to become very widespread. Perl can use this form internaly
703 to represent strings, so conversions to and from this form are
704 particularly efficient (as octets in memory do not have to change,
705 just the meta-data that tells Perl how to treat them).
711 $bytes = encode_utf8($string);
713 The characters that comprise string are encoded in Perl's superset of UTF-8
714 and the resulting octets returned as a sequence of bytes. All possible
715 characters have a UTF-8 representation so this function cannot fail.
719 $string = decode_utf8($bytes [,CHECK]);
721 The sequence of octets represented by $bytes is decoded from UTF-8
722 into a sequence of logical characters. Not all sequences of octets
723 form valid UTF-8 encodings, so it is possible for this call to fail.
724 For CHECK see L</"Handling Malformed Data">.
728 =head2 Other Encodings of Unicode
730 UTF-16 is similar to UCS-2, 16 bit or 2-byte chunks. UCS-2 can only
731 represent 0..0xFFFF, while UTF-16 has a "surrogate pair" scheme which
732 allows it to cover the whole Unicode range.
734 Encode implements big-endian UCS-2 aliased to "iso-10646-1" as that
735 happens to be the name used by that representation when used with X11
738 UTF-32 or UCS-4 is 32-bit or 4-byte chunks. Perl's logical characters
739 can be considered as being in this form without encoding. An encoding
740 to transfer strings in this form (e.g. to write them to a file) would
743 pack('L*', unpack('U*', $string)); # native
745 pack('V*', unpack('U*', $string)); # little-endian
747 pack('N*', unpack('U*', $string)); # big-endian
749 depending on the endianness required.
751 No UTF-32 encodings are implemented yet.
753 Both UCS-2 and UCS-4 style encodings can have "byte order marks" by
754 representing the code point 0xFFFE as the very first thing in a file.
756 =head2 Listing available encodings
758 use Encode qw(encodings);
761 Returns a list of the canonical names of the available encodings.
763 =head2 Defining Aliases
765 use Encode qw(define_alias);
766 define_alias( newName => ENCODING);
768 Allows newName to be used as am alias for ENCODING. ENCODING may be
769 either the name of an encoding or and encoding object (as above).
771 Currently I<newName> can be specified in the following ways:
775 =item As a simple string.
777 =item As a qr// compiled regular expression, e.g.:
779 define_alias( qr/^iso8859-(\d+)$/i => '"iso-8859-$1"' );
781 In this case if I<ENCODING> is not a reference it is C<eval>-ed to
782 allow C<$1> etc. to be subsituted. The example is one way to names as
783 used in X11 font names to alias the MIME names for the iso-8859-*
786 =item As a code reference, e.g.:
788 define_alias( sub { return /^iso8859-(\d+)$/i ? "iso-8859-$1" : undef } , '');
790 In this case C<$_> will be set to the name that is being looked up and
791 I<ENCODING> is passed to the sub as its first argument. The example
792 is another way to names as used in X11 font names to alias the MIME
793 names for the iso-8859-* family.
797 =head2 Defining Encodings
799 use Encode qw(define_alias);
800 define_encoding( $object, 'canonicalName' [,alias...]);
802 Causes I<canonicalName> to be associated with I<$object>. The object
803 should provide the interface described in L</"IMPLEMENTATION CLASSES">
804 below. If more than two arguments are provided then additional
805 arguments are taken as aliases for I<$object> as for C<define_alias>.
807 =head1 Encoding and IO
809 It is very common to want to do encoding transformations when
810 reading or writing files, network connections, pipes etc.
811 If Perl is configured to use the new 'perlio' IO system then
812 C<Encode> provides a "layer" (See L<perliol>) which can transform
813 data as it is read or written.
815 Here is how the blind poet would modernise the encoding:
818 open(my $iliad,'<:encoding(iso-8859-7)','iliad.greek');
819 open(my $utf8,'>:utf8','iliad.utf8');
825 In addition the new IO system can also be configured to read/write
826 UTF-8 encoded characters (as noted above this is efficient):
828 open(my $fh,'>:utf8','anything');
829 print $fh "Any \x{0021} string \N{SMILEY FACE}\n";
831 Either of the above forms of "layer" specifications can be made the default
832 for a lexical scope with the C<use open ...> pragma. See L<open>.
834 Once a handle is open is layers can be altered using C<binmode>.
836 Without any such configuration, or if Perl itself is built using
837 system's own IO, then write operations assume that file handle accepts
838 only I<bytes> and will C<die> if a character larger than 255 is
839 written to the handle. When reading, each octet from the handle
840 becomes a byte-in-a-character. Note that this default is the same
841 behaviour as bytes-only languages (including Perl before v5.6) would
842 have, and is sufficient to handle native 8-bit encodings
843 e.g. iso-8859-1, EBCDIC etc. and any legacy mechanisms for handling
844 other encodings and binary data.
846 In other cases it is the programs responsibility to transform
847 characters into bytes using the API above before doing writes, and to
848 transform the bytes read from a handle into characters before doing
849 "character operations" (e.g. C<lc>, C</\W+/>, ...).
851 You can also use PerlIO to convert larger amounts of data you don't
852 want to bring into memory. For example to convert between ISO 8859-1
853 (Latin 1) and UTF-8 (or UTF-EBCDIC in EBCDIC machines):
855 open(F, "<:encoding(iso-8859-1)", "data.txt") or die $!;
856 open(G, ">:utf8", "data.utf") or die $!;
857 while (<F>) { print G }
859 # Could also do "print G <F>" but that would pull
860 # the whole file into memory just to write it out again.
864 open(my $f, "<:encoding(cp1252)")
865 open(my $g, ">:encoding(iso-8859-2)")
866 open(my $h, ">:encoding(latin9)") # iso-8859-15
868 See L<PerlIO> for more information.
870 =head1 Encoding How to ...
876 =item * IO with mixed content (faking iso-2020-*)
878 =item * MIME's Content-Length:
880 =item * UTF-8 strings in binary data.
882 =item * Perl/Encode wrappers on non-Unicode XS modules.
886 =head1 Messing with Perl's Internals
888 The following API uses parts of Perl's internals in the current
889 implementation. As such they are efficient, but may change.
893 =item * is_utf8(STRING [, CHECK])
895 [INTERNAL] Test whether the UTF-8 flag is turned on in the STRING.
896 If CHECK is true, also checks the data in STRING for being well-formed
897 UTF-8. Returns true if successful, false otherwise.
899 =item * valid_utf8(STRING)
901 [INTERNAL] Test whether STRING is in a consistent state. Will return
902 true if string is held as bytes, or is well-formed UTF-8 and has the
903 UTF-8 flag on. Main reason for this routine is to allow Perl's
904 testsuite to check that operations have left strings in a consistent
911 [INTERNAL] Turn on the UTF-8 flag in STRING. The data in STRING is
912 B<not> checked for being well-formed UTF-8. Do not use unless you
913 B<know> that the STRING is well-formed UTF-8. Returns the previous
914 state of the UTF-8 flag (so please don't test the return value as
915 I<not> success or failure), or C<undef> if STRING is not a string.
921 [INTERNAL] Turn off the UTF-8 flag in STRING. Do not use frivolously.
922 Returns the previous state of the UTF-8 flag (so please don't test the
923 return value as I<not> success or failure), or C<undef> if STRING is
928 =head1 IMPLEMENTATION CLASSES
930 As mentioned above encodings are (in the current implementation at least)
931 defined by objects. The mapping of encoding name to object is via the
934 The values of the hash can currently be either strings or objects.
935 The string form may go away in the future. The string form occurs
936 when C<encodings()> has scanned C<@INC> for loadable encodings but has
937 not actually loaded the encoding in question. This is because the
938 current "loading" process is all Perl and a bit slow.
940 Once an encoding is loaded then value of the hash is object which
941 implements the encoding. The object should provide the following
948 Should return the string representing the canonical name of the encoding.
950 =item -E<gt>new_sequence
952 This is a placeholder for encodings with state. It should return an
953 object which implements this interface, all current implementations
954 return the original object.
956 =item -E<gt>encode($string,$check)
958 Should return the octet sequence representing I<$string>. If I<$check>
959 is true it should modify I<$string> in place to remove the converted
960 part (i.e. the whole string unless there is an error). If an error
961 occurs it should return the octet sequence for the fragment of string
962 that has been converted, and modify $string in-place to remove the
963 converted part leaving it starting with the problem fragment.
965 If check is is false then C<encode> should make a "best effort" to
966 convert the string - for example by using a replacement character.
968 =item -E<gt>decode($octets,$check)
970 Should return the string that I<$octets> represents. If I<$check> is
971 true it should modify I<$octets> in place to remove the converted part
972 (i.e. the whole sequence unless there is an error). If an error
973 occurs it should return the fragment of string that has been
974 converted, and modify $octets in-place to remove the converted part
975 leaving it starting with the problem fragment.
977 If check is is false then C<decode> should make a "best effort" to
978 convert the string - for example by using Unicode's "\x{FFFD}" as a
979 replacement character.
983 It should be noted that the check behaviour is different from the
984 outer public API. The logic is that the "unchecked" case is useful
985 when encoding is part of a stream which may be reporting errors
986 (e.g. STDERR). In such cases it is desirable to get everything
987 through somehow without causing additional errors which obscure the
988 original one. Also the encoding is best placed to know what the
989 correct replacement character is, so if that is the desired behaviour
990 then letting low level code do it is the most efficient.
992 In contrast if check is true, the scheme above allows the encoding to
993 do as much as it can and tell layer above how much that was. What is
994 lacking at present is a mechanism to report what went wrong. The most
995 likely interface will be an additional method call to the object, or
996 perhaps (to avoid forcing per-stream objects on otherwise stateless
997 encodings) and additional parameter.
999 It is also highly desirable that encoding classes inherit from
1000 C<Encode::Encoding> as a base class. This allows that class to define
1001 additional behaviour for all encoding objects. For example built in
1002 Unicode, UCS-2 and UTF-8 classes use :
1004 package Encode::MyEncoding;
1005 use base qw(Encode::Encoding);
1007 __PACKAGE__->Define(qw(myCanonical myAlias));
1009 To create an object with bless {Name => ...},$class, and call
1010 define_encoding. They inherit their C<name> method from
1011 C<Encode::Encoding>.
1013 =head2 Compiled Encodings
1015 F<Encode.xs> provides a class C<Encode::XS> which provides the
1016 interface described above. It calls a generic octet-sequence to
1017 octet-sequence "engine" that is driven by tables (defined in
1018 F<encengine.c>). The same engine is used for both encode and
1019 decode. C<Encode:XS>'s C<encode> forces Perl's characters to their
1020 UTF-8 form and then treats them as just another multibyte
1021 encoding. C<Encode:XS>'s C<decode> transforms the sequence and then
1022 turns the UTF-8-ness flag as that is the form that the tables are
1023 defined to produce. For details of the engine see the comments in
1026 The tables are produced by the Perl script F<compile> (the name needs
1027 to change so we can eventually install it somewhere). F<compile> can
1028 currently read two formats:
1034 This is a coined format used by Tcl. It is documented in
1035 Encode/EncodeFormat.pod.
1039 This is the semi-standard format used by IBM's ICU package.
1043 F<compile> can write the following forms:
1049 See above - the F<Encode/*.ucm> files provided with the distribution have
1050 been created from the original Tcl .enc files using this approach.
1054 Produces tables as C data structures - this is used to build in encodings
1055 into F<Encode.so>/F<Encode.dll>.
1059 In theory this allows encodings to be stand-alone loadable Perl
1060 extensions. The process has not yet been tested. The plan is to use
1061 this approach for large East Asian encodings.
1065 The set of encodings built-in to F<Encode.so>/F<Encode.dll> is
1066 determined by F<Makefile.PL>. The current set is as follows:
1070 =item ascii and iso-8859-*
1072 That is all the common 8-bit "western" encodings.
1074 =item IBM-1047 and two other variants of EBCDIC.
1076 These are the same variants that are supported by EBCDIC Perl as
1077 "native" encodings. They are included to prove "reversibility" of
1078 some constructs in EBCDIC Perl.
1080 =item symbol and dingbats as used by Tk on X11.
1082 (The reason Encode got started was to support Perl/Tk.)
1086 That set is rather ad hoc and has been driven by the needs of the
1087 tests rather than the needs of typical applications. It is likely
1092 L<perlunicode>, L<perlebcdic>, L<perlfunc/open>, L<PerlIO>