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
45 # 0 1 2 3 4 5 6 7 8 9 10
46 our @latin2iso_num = ( 0, 1, 2, 3, 4, 9, 10, 13, 14, 15, 16 );
52 return keys %encoding;
59 unless (exists $alias{$_})
61 for (my $i=0; $i < @alias; $i += 2)
63 my $alias = $alias[$i];
64 my $val = $alias[$i+1];
66 if (ref($alias) eq 'Regexp' && $_ =~ $alias)
70 elsif (ref($alias) eq 'CODE')
72 $new = &{$alias}($val)
74 elsif (lc($_) eq lc($alias))
80 next if $new eq $_; # avoid (direct) recursion on bugs
81 my $enc = (ref($new)) ? $new : find_encoding($new);
97 my ($alias,$name) = splice(@_,0,2);
98 push(@alias, $alias => $name);
102 # Allow variants of iso-8859-1 etc.
103 define_alias( qr/^iso[-_]?(\d+)[-_](\d+)$/i => '"iso-$1-$2"' );
105 # At least HP-UX has these.
106 define_alias( qr/^iso8859(\d+)$/i => '"iso-8859-$1"' );
108 # This is a font issue, not an encoding issue.
109 # (The currency symbol of the Latin 1 upper half
110 # has been redefined as the euro symbol.)
111 define_alias( qr/^(.+)\@euro$/i => '"$1"' );
113 # Allow latin-1 style names as well
114 define_alias( qr/^(?:iso[-_]?)?latin[-_]?(\d+)$/i => '"iso-8859-$latin2iso_num[$1]"' );
116 # Common names for non-latin prefered MIME names
117 define_alias( 'ascii' => 'US-ascii',
118 'cyrillic' => 'iso-8859-5',
119 'arabic' => 'iso-8859-6',
120 'greek' => 'iso-8859-7',
121 'hebrew' => 'iso-8859-8');
123 # At least AIX has IBM-NNN (surprisingly...) instead of cpNNN.
124 define_alias( qr/^ibm[-_]?(\d\d\d\d?)$/i => '"cp$1"');
126 # Standardize on the dashed versions.
127 define_alias( qr/^utf8$/i => 'utf-8' );
128 define_alias( qr/^koi8r$/i => 'koi8-r' );
130 # TODO: the HP-UX '8' encodings: arabic8 greek8 hebrew8 roman8 turkish8
131 # TODO: the Thai Encoding tis620
132 # TODO: the Chinese Encoding gb18030
133 # TODO: what is the Japanese 'ujis' encoding seen in some Linuxes?
135 # Map white space and _ to '-'
136 define_alias( qr/^(\S+)[\s_]+(.*)$/i => '"$1-$2"' );
142 $encoding{$name} = $obj;
144 define_alias($lc => $obj) unless $lc eq $name;
148 define_alias($alias,$obj);
155 my ($class,$name) = @_;
157 if (ref($name) && $name->can('new_sequence'))
161 if (exists $encoding{$name})
163 return $encoding{$name};
167 return $class->findAlias($name);
174 return __PACKAGE__->getEncoding($name);
179 my ($name,$string,$check) = @_;
180 my $enc = find_encoding($name);
181 croak("Unknown encoding '$name'") unless defined $enc;
182 my $octets = $enc->encode($string,$check);
183 return undef if ($check && length($string));
189 my ($name,$octets,$check) = @_;
190 my $enc = find_encoding($name);
191 croak("Unknown encoding '$name'") unless defined $enc;
192 my $string = $enc->decode($octets,$check);
193 return undef if ($check && length($octets));
199 my ($string,$from,$to,$check) = @_;
200 my $f = find_encoding($from);
201 croak("Unknown encoding '$from'") unless defined $f;
202 my $t = find_encoding($to);
203 croak("Unknown encoding '$to'") unless defined $t;
204 my $uni = $f->decode($string,$check);
205 return undef if ($check && length($string));
206 $string = $t->encode($uni,$check);
207 return undef if ($check && length($uni));
208 return length($_[0] = $string);
221 return undef unless utf8::decode($str);
225 package Encode::Encoding;
226 # Base class for classes which implement encodings
231 my $canonical = shift;
232 $obj = bless { Name => $canonical },$obj unless ref $obj;
233 # warn "$canonical => $obj\n";
234 Encode::define_encoding($obj, $canonical, @_);
237 sub name { shift->{'Name'} }
239 # Temporary legacy methods
240 sub toUnicode { shift->decode(@_) }
241 sub fromUnicode { shift->encode(@_) }
243 sub new_sequence { return $_[0] }
246 use base 'Encode::Encoding';
248 package Encode::Internal;
249 use base 'Encode::Encoding';
251 # Dummy package that provides the encode interface but leaves data
252 # as UTF-X encoded. It is here so that from_to() works.
254 __PACKAGE__->Define('Internal');
256 Encode::define_alias( 'Unicode' => 'Internal' ) if ord('A') == 65;
260 my ($obj,$str,$chk) = @_;
268 package Encoding::Unicode;
269 use base 'Encode::Encoding';
271 __PACKAGE__->Define('Unicode') unless ord('A') == 65;
275 my ($obj,$str,$chk) = @_;
277 for (my $i = 0; $i < length($str); $i++)
279 $res .= chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
287 my ($obj,$str,$chk) = @_;
289 for (my $i = 0; $i < length($str); $i++)
291 $res .= chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
298 package Encode::utf8;
299 use base 'Encode::Encoding';
300 # package to allow long-hand
301 # $octets = encode( utf8 => $string );
304 __PACKAGE__->Define(qw(UTF-8 utf8));
308 my ($obj,$octets,$chk) = @_;
309 my $str = Encode::decode_utf8($octets);
320 my ($obj,$string,$chk) = @_;
321 my $octets = Encode::encode_utf8($string);
326 package Encode::iso10646_1;
327 use base 'Encode::Encoding';
328 # Encoding is 16-bit network order Unicode (no surogates)
329 # Used for X font encodings
331 __PACKAGE__->Define(qw(UCS-2 iso-10646-1));
335 my ($obj,$str,$chk) = @_;
339 my $code = unpack('n',substr($str,0,2,'')) & 0xffff;
342 $_[1] = $str if $chk;
349 my ($obj,$uni,$chk) = @_;
353 my $ch = substr($uni,0,1,'');
360 $str .= pack('n',$x);
362 $_[1] = $uni if $chk;
366 # switch back to Encode package in case we ever add AutoLoader
375 Encode - character encodings
383 The C<Encode> module provides the interfaces between Perl's strings
384 and the rest of the system. Perl strings are sequences of B<characters>.
386 The repertoire of characters that Perl can represent is at least that
387 defined by the Unicode Consortium. On most platforms the ordinal
388 values of the characters (as returned by C<ord(ch)>) is the "Unicode
389 codepoint" for the character (the exceptions are those platforms where
390 the legacy encoding is some variant of EBCDIC rather than a super-set
391 of ASCII - see L<perlebcdic>).
393 Traditionaly computer data has been moved around in 8-bit chunks
394 often called "bytes". These chunks are also known as "octets" in
395 networking standards. Perl is widely used to manipulate data of
396 many types - not only strings of characters representing human or
397 computer languages but also "binary" data being the machines representation
398 of numbers, pixels in an image - or just about anything.
400 When Perl is processing "binary data" the programmer wants Perl to process
401 "sequences of bytes". This is not a problem for Perl - as a byte has 256
402 possible values it easily fits in Perl's much larger "logical character".
410 I<character>: a character in the range 0..(2**32-1) (or more).
411 (What Perl's strings are made of.)
415 I<byte>: a character in the range 0..255
416 (A special case of a Perl character.)
420 I<octet>: 8 bits of data, with ordinal values 0..255
421 (Term for bytes passed to or from a non-Perl context, e.g. disk file.)
425 The marker [INTERNAL] marks Internal Implementation Details, in
426 general meant only for those who think they know what they are doing,
427 and such details may change in future releases.
431 =head2 Characteristics of an Encoding
433 An encoding has a "repertoire" of characters that it can represent,
434 and for each representable character there is at least one sequence of
435 octets that represents it.
437 =head2 Types of Encodings
439 Encodings can be divided into the following types:
443 =item * Fixed length 8-bit (or less) encodings.
445 Each character is a single octet so may have a repertoire of up to
446 256 characters. ASCII and iso-8859-* are typical examples.
448 =item * Fixed length 16-bit encodings
450 Each character is two octets so may have a repertoire of up to
451 65 536 characters. Unicode's UCS-2 is an example. Also used for
452 encodings for East Asian languages.
454 =item * Fixed length 32-bit encodings.
456 Not really very "encoded" encodings. The Unicode code points
457 are just represented as 4-octet integers. None the less because
458 different architectures use different representations of integers
459 (so called "endian") there at least two disctinct encodings.
461 =item * Multi-byte encodings
463 The number of octets needed to represent a character varies.
464 UTF-8 is a particularly complex but regular case of a multi-byte
465 encoding. Several East Asian countries use a multi-byte encoding
466 where 1-octet is used to cover western roman characters and Asian
467 characters get 2-octets.
468 (UTF-16 is strictly a multi-byte encoding taking either 2 or 4 octets
469 to represent a Unicode code point.)
471 =item * "Escape" encodings.
473 These encodings embed "escape sequences" into the octet sequence
474 which describe how the following octets are to be interpreted.
475 The iso-2022-* family is typical. Following the escape sequence
476 octets are encoded by an "embedded" encoding (which will be one
477 of the above types) until another escape sequence switches to
478 a different "embedded" encoding.
480 These schemes are very flexible and can handle mixed languages but are
481 very complex to process (and have state). No escape encodings are
482 implemented for Perl yet.
486 =head2 Specifying Encodings
488 Encodings can be specified to the API described below in two ways:
494 Encoding names are strings with characters taken from a restricted
495 repertoire. See L</"Encoding Names">.
497 =item 2. As an object
499 Encoding objects are returned by C<find_encoding($name)>.
503 =head2 Encoding Names
505 Encoding names are case insensitive. White space in names is ignored.
506 In addition an encoding may have aliases. Each encoding has one
507 "canonical" name. The "canonical" name is chosen from the names of
508 the encoding by picking the first in the following sequence:
512 =item * The MIME name as defined in IETF RFC-XXXX.
514 =item * The name in the IANA registry.
516 =item * The name used by the the organization that defined it.
520 Because of all the alias issues, and because in the general case
521 encodings have state C<Encode> uses the encoding object internally
522 once an operation is in progress.
524 =head1 PERL ENCODING API
526 =head2 Generic Encoding Interface
532 $bytes = encode(ENCODING, $string[, CHECK])
534 Encodes string from Perl's internal form into I<ENCODING> and returns
535 a sequence of octets. For CHECK see L</"Handling Malformed Data">.
539 $string = decode(ENCODING, $bytes[, CHECK])
541 Decode sequence of octets assumed to be in I<ENCODING> into Perl's
542 internal form and returns the resulting string. For CHECK see
543 L</"Handling Malformed Data">.
547 from_to($string, FROM_ENCODING, TO_ENCODING[, CHECK])
549 Convert B<in-place> the data between two encodings. How did the data
550 in $string originally get to be in FROM_ENCODING? Either using
551 encode() or through PerlIO: See L</"Encoding and IO">. For CHECK
552 see L</"Handling Malformed Data">.
554 For example to convert ISO 8859-1 data to UTF-8:
556 from_to($data, "iso-8859-1", "utf-8");
558 and to convert it back:
560 from_to($data, "utf-8", "iso-8859-1");
562 Note that because the conversion happens in place, the data to be
563 converted cannot be a string constant, it must be a scalar variable.
567 =head2 Handling Malformed Data
569 If CHECK is not set, C<undef> is returned. If the data is supposed to
570 be UTF-8, an optional lexical warning (category utf8) is given. If
571 CHECK is true but not a code reference, dies.
573 It would desirable to have a way to indicate that transform should use
574 the encodings "replacement character" - no such mechanism is defined yet.
576 It is also planned to allow I<CHECK> to be a code reference.
578 This is not yet implemented as there are design issues with what its
579 arguments should be and how it returns its results.
585 Passed remaining fragment of string being processed.
586 Modifies it in place to remove bytes/characters it can understand
587 and returns a string used to represent them.
591 my $ch = substr($_[0],0,1,'');
592 return sprintf("\x{%02X}",ord($ch);
595 This scheme is close to how underlying C code for Encode works, but gives
596 the fixup routine very little context.
600 Passed original string, and an index into it of the problem area, and
601 output string so far. Appends what it will to output string and
602 returns new index into original string. For example:
605 # my ($s,$i,$d) = @_;
606 my $ch = substr($_[0],$_[1],1);
607 $_[2] .= sprintf("\x{%02X}",ord($ch);
611 This scheme gives maximal control to the fixup routine but is more
612 complicated to code, and may need internals of Encode to be tweaked to
613 keep original string intact.
619 Multiple return values rather than in-place modifications.
621 Index into the string could be pos($str) allowing s/\G...//.
627 The Unicode consortium defines the UTF-8 standard as a way of encoding
628 the entire Unicode repertiore as sequences of octets. This encoding is
629 expected to become very widespread. Perl can use this form internaly
630 to represent strings, so conversions to and from this form are
631 particularly efficient (as octets in memory do not have to change,
632 just the meta-data that tells Perl how to treat them).
638 $bytes = encode_utf8($string);
640 The characters that comprise string are encoded in Perl's superset of UTF-8
641 and the resulting octets returned as a sequence of bytes. All possible
642 characters have a UTF-8 representation so this function cannot fail.
646 $string = decode_utf8($bytes [,CHECK]);
648 The sequence of octets represented by $bytes is decoded from UTF-8
649 into a sequence of logical characters. Not all sequences of octets
650 form valid UTF-8 encodings, so it is possible for this call to fail.
651 For CHECK see L</"Handling Malformed Data">.
655 =head2 Other Encodings of Unicode
657 UTF-16 is similar to UCS-2, 16 bit or 2-byte chunks. UCS-2 can only
658 represent 0..0xFFFF, while UTF-16 has a "surrogate pair" scheme which
659 allows it to cover the whole Unicode range.
661 Encode implements big-endian UCS-2 aliased to "iso-10646-1" as that
662 happens to be the name used by that representation when used with X11
665 UTF-32 or UCS-4 is 32-bit or 4-byte chunks. Perl's logical characters
666 can be considered as being in this form without encoding. An encoding
667 to transfer strings in this form (e.g. to write them to a file) would
670 pack('L',map(chr($_),split(//,$string))); # native
672 pack('V',map(chr($_),split(//,$string))); # little-endian
674 pack('N',map(chr($_),split(//,$string))); # big-endian
676 depending on the endian required.
678 No UTF-32 encodings are implemented yet.
680 Both UCS-2 and UCS-4 style encodings can have "byte order marks" by
681 representing the code point 0xFFFE as the very first thing in a file.
683 =head2 Listing available encodings
685 use Encode qw(encodings);
688 Returns a list of the canonical names of the available encodings.
690 =head2 Defining Aliases
692 use Encode qw(define_alias);
693 define_alias( newName => ENCODING);
695 Allows newName to be used as am alias for ENCODING. ENCODING may be
696 either the name of an encoding or and encoding object (as above).
698 Currently I<newName> can be specified in the following ways:
702 =item As a simple string.
704 =item As a qr// compiled regular expression, e.g.:
706 define_alias( qr/^iso8859-(\d+)$/i => '"iso-8859-$1"' );
708 In this case if I<ENCODING> is not a reference it is C<eval>-ed to
709 allow C<$1> etc. to be subsituted. The example is one way to names as
710 used in X11 font names to alias the MIME names for the iso-8859-*
713 =item As a code reference, e.g.:
715 define_alias( sub { return /^iso8859-(\d+)$/i ? "iso-8859-$1" : undef } , '');
717 In this case C<$_> will be set to the name that is being looked up and
718 I<ENCODING> is passed to the sub as its first argument. The example
719 is another way to names as used in X11 font names to alias the MIME
720 names for the iso-8859-* family.
724 =head2 Defining Encodings
726 use Encode qw(define_alias);
727 define_encoding( $object, 'canonicalName' [,alias...]);
729 Causes I<canonicalName> to be associated with I<$object>. The object
730 should provide the interface described in L</"IMPLEMENTATION CLASSES">
731 below. If more than two arguments are provided then additional
732 arguments are taken as aliases for I<$object> as for C<define_alias>.
734 =head1 Encoding and IO
736 It is very common to want to do encoding transformations when
737 reading or writing files, network connections, pipes etc.
738 If Perl is configured to use the new 'perlio' IO system then
739 C<Encode> provides a "layer" (See L<perliol>) which can transform
740 data as it is read or written.
742 Here is how the blind poet would modernise the encoding:
745 open(my $iliad,'<:encoding(iso-8859-7)','iliad.greek');
746 open(my $utf8,'>:utf8','iliad.utf8');
752 In addition the new IO system can also be configured to read/write
753 UTF-8 encoded characters (as noted above this is efficient):
755 open(my $fh,'>:utf8','anything');
756 print $fh "Any \x{0021} string \N{SMILEY FACE}\n";
758 Either of the above forms of "layer" specifications can be made the default
759 for a lexical scope with the C<use open ...> pragma. See L<open>.
761 Once a handle is open is layers can be altered using C<binmode>.
763 Without any such configuration, or if Perl itself is built using
764 system's own IO, then write operations assume that file handle accepts
765 only I<bytes> and will C<die> if a character larger than 255 is
766 written to the handle. When reading, each octet from the handle
767 becomes a byte-in-a-character. Note that this default is the same
768 behaviour as bytes-only languages (including Perl before v5.6) would
769 have, and is sufficient to handle native 8-bit encodings
770 e.g. iso-8859-1, EBCDIC etc. and any legacy mechanisms for handling
771 other encodings and binary data.
773 In other cases it is the programs responsibility to transform
774 characters into bytes using the API above before doing writes, and to
775 transform the bytes read from a handle into characters before doing
776 "character operations" (e.g. C<lc>, C</\W+/>, ...).
778 You can also use PerlIO to convert larger amounts of data you don't
779 want to bring into memory. For example to convert between ISO 8859-1
780 (Latin 1) and UTF-8 (or UTF-EBCDIC in EBCDIC machines):
782 open(F, "<:encoding(iso-8859-1)", "data.txt") or die $!;
783 open(G, ">:utf8", "data.utf") or die $!;
784 while (<F>) { print G }
786 # Could also do "print G <F>" but that would pull
787 # the whole file into memory just to write it out again.
791 open(my $f, "<:encoding(cp1252)")
792 open(my $g, ">:encoding(iso-8859-2)")
793 open(my $h, ">:encoding(latin9)") # iso-8859-15
795 See L<PerlIO> for more information.
797 =head1 Encoding How to ...
803 =item * IO with mixed content (faking iso-2020-*)
805 =item * MIME's Content-Length:
807 =item * UTF-8 strings in binary data.
809 =item * Perl/Encode wrappers on non-Unicode XS modules.
813 =head1 Messing with Perl's Internals
815 The following API uses parts of Perl's internals in the current
816 implementation. As such they are efficient, but may change.
820 =item * is_utf8(STRING [, CHECK])
822 [INTERNAL] Test whether the UTF-8 flag is turned on in the STRING.
823 If CHECK is true, also checks the data in STRING for being well-formed
824 UTF-8. Returns true if successful, false otherwise.
826 =item * valid_utf8(STRING)
828 [INTERNAL] Test whether STRING is in a consistent state. Will return
829 true if string is held as bytes, or is well-formed UTF-8 and has the
830 UTF-8 flag on. Main reason for this routine is to allow Perl's
831 testsuite to check that operations have left strings in a consistent
838 [INTERNAL] Turn on the UTF-8 flag in STRING. The data in STRING is
839 B<not> checked for being well-formed UTF-8. Do not use unless you
840 B<know> that the STRING is well-formed UTF-8. Returns the previous
841 state of the UTF-8 flag (so please don't test the return value as
842 I<not> success or failure), or C<undef> if STRING is not a string.
848 [INTERNAL] Turn off the UTF-8 flag in STRING. Do not use frivolously.
849 Returns the previous state of the UTF-8 flag (so please don't test the
850 return value as I<not> success or failure), or C<undef> if STRING is
855 =head1 IMPLEMENTATION CLASSES
857 As mentioned above encodings are (in the current implementation at least)
858 defined by objects. The mapping of encoding name to object is via the
861 The values of the hash can currently be either strings or objects.
862 The string form may go away in the future. The string form occurs
863 when C<encodings()> has scanned C<@INC> for loadable encodings but has
864 not actually loaded the encoding in question. This is because the
865 current "loading" process is all Perl and a bit slow.
867 Once an encoding is loaded then value of the hash is object which
868 implements the encoding. The object should provide the following
875 Should return the string representing the canonical name of the encoding.
877 =item -E<gt>new_sequence
879 This is a placeholder for encodings with state. It should return an
880 object which implements this interface, all current implementations
881 return the original object.
883 =item -E<gt>encode($string,$check)
885 Should return the octet sequence representing I<$string>. If I<$check>
886 is true it should modify I<$string> in place to remove the converted
887 part (i.e. the whole string unless there is an error). If an error
888 occurs it should return the octet sequence for the fragment of string
889 that has been converted, and modify $string in-place to remove the
890 converted part leaving it starting with the problem fragment.
892 If check is is false then C<encode> should make a "best effort" to
893 convert the string - for example by using a replacement character.
895 =item -E<gt>decode($octets,$check)
897 Should return the string that I<$octets> represents. If I<$check> is
898 true it should modify I<$octets> in place to remove the converted part
899 (i.e. the whole sequence unless there is an error). If an error
900 occurs it should return the fragment of string that has been
901 converted, and modify $octets in-place to remove the converted part
902 leaving it starting with the problem fragment.
904 If check is is false then C<decode> should make a "best effort" to
905 convert the string - for example by using Unicode's "\x{FFFD}" as a
906 replacement character.
910 It should be noted that the check behaviour is different from the
911 outer public API. The logic is that the "unchecked" case is useful
912 when encoding is part of a stream which may be reporting errors
913 (e.g. STDERR). In such cases it is desirable to get everything
914 through somehow without causing additional errors which obscure the
915 original one. Also the encoding is best placed to know what the
916 correct replacement character is, so if that is the desired behaviour
917 then letting low level code do it is the most efficient.
919 In contrast if check is true, the scheme above allows the encoding to
920 do as much as it can and tell layer above how much that was. What is
921 lacking at present is a mechanism to report what went wrong. The most
922 likely interface will be an additional method call to the object, or
923 perhaps (to avoid forcing per-stream objects on otherwise stateless
924 encodings) and additional parameter.
926 It is also highly desirable that encoding classes inherit from
927 C<Encode::Encoding> as a base class. This allows that class to define
928 additional behaviour for all encoding objects. For example built in
929 Unicode, UCS-2 and UTF-8 classes use :
931 package Encode::MyEncoding;
932 use base qw(Encode::Encoding);
934 __PACKAGE__->Define(qw(myCanonical myAlias));
936 To create an object with bless {Name => ...},$class, and call
937 define_encoding. They inherit their C<name> method from
940 =head2 Compiled Encodings
942 F<Encode.xs> provides a class C<Encode::XS> which provides the
943 interface described above. It calls a generic octet-sequence to
944 octet-sequence "engine" that is driven by tables (defined in
945 F<encengine.c>). The same engine is used for both encode and
946 decode. C<Encode:XS>'s C<encode> forces Perl's characters to their
947 UTF-8 form and then treats them as just another multibyte
948 encoding. C<Encode:XS>'s C<decode> transforms the sequence and then
949 turns the UTF-8-ness flag as that is the form that the tables are
950 defined to produce. For details of the engine see the comments in
953 The tables are produced by the Perl script F<compile> (the name needs
954 to change so we can eventually install it somewhere). F<compile> can
955 currently read two formats:
961 This is a coined format used by Tcl. It is documented in
962 Encode/EncodeFormat.pod.
966 This is the semi-standard format used by IBM's ICU package.
970 F<compile> can write the following forms:
976 See above - the F<Encode/*.ucm> files provided with the distribution have
977 been created from the original Tcl .enc files using this approach.
981 Produces tables as C data structures - this is used to build in encodings
982 into F<Encode.so>/F<Encode.dll>.
986 In theory this allows encodings to be stand-alone loadable Perl
987 extensions. The process has not yet been tested. The plan is to use
988 this approach for large East Asian encodings.
992 The set of encodings built-in to F<Encode.so>/F<Encode.dll> is
993 determined by F<Makefile.PL>. The current set is as follows:
997 =item ascii and iso-8859-*
999 That is all the common 8-bit "western" encodings.
1001 =item IBM-1047 and two other variants of EBCDIC.
1003 These are the same variants that are supported by EBCDIC Perl as
1004 "native" encodings. They are included to prove "reversibility" of
1005 some constructs in EBCDIC Perl.
1007 =item symbol and dingbats as used by Tk on X11.
1009 (The reason Encode got started was to support Perl/Tk.)
1013 That set is rather ad hoc and has been driven by the needs of the
1014 tests rather than the needs of typical applications. It is likely
1019 L<perlunicode>, L<perlebcdic>, L<perlfunc/open>, L<PerlIO>