2 # $Id: Encode.pm,v 2.15 2006/04/06 15:44:11 dankogai Exp dankogai $
6 our $VERSION = sprintf "%d.%02d", q$Revision: 2.15 $ =~ /(\d+)/g;
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 str2bytes bytes2str
18 encodings find_encoding clone_encoding
21 our @FB_FLAGS = qw(DIE_ON_ERR WARN_ON_ERR RETURN_ON_ERR LEAVE_SRC
22 PERLQQ HTMLCREF XMLCREF STOP_AT_PARTIAL);
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('renew') 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};
119 sub find_encoding($;$)
121 my ($name, $skip_external) = @_;
122 return __PACKAGE__->getEncoding($name,$skip_external);
125 sub resolve_alias($){
126 my $obj = find_encoding(shift);
127 defined $obj and return $obj->name;
131 sub clone_encoding($){
132 my $obj = find_encoding(shift);
134 eval { require Storable };
136 return Storable::dclone($obj);
141 my ($name, $string, $check) = @_;
142 return undef unless defined $string;
143 $string .= '' if ref $string; # stringify;
145 my $enc = find_encoding($name);
146 unless(defined $enc){
148 Carp::croak("Unknown encoding '$name'");
150 my $octets = $enc->encode($string,$check);
151 $_[1] = $string if $check and !($check & LEAVE_SRC());
154 *str2bytes = \&encode;
158 my ($name,$octets,$check) = @_;
159 return undef unless defined $octets;
160 $octets .= '' if ref $octets;
162 my $enc = find_encoding($name);
163 unless(defined $enc){
165 Carp::croak("Unknown encoding '$name'");
167 my $string = $enc->decode($octets,$check);
168 $_[1] = $octets if $check and !($check & LEAVE_SRC());
171 *bytes2str = \&decode;
175 my ($string,$from,$to,$check) = @_;
176 return undef unless defined $string;
178 my $f = find_encoding($from);
181 Carp::croak("Unknown encoding '$from'");
183 my $t = find_encoding($to);
186 Carp::croak("Unknown encoding '$to'");
188 my $uni = $f->decode($string);
189 $_[0] = $string = $t->encode($uni,$check);
190 return undef if ($check && length($uni));
191 return defined($_[0]) ? length($string) : undef ;
203 my ($str, $check) = @_;
204 return $str if is_utf8($str);
206 return decode("utf8", $str, $check);
208 return decode("utf8", $str);
213 predefine_encodings(1);
216 # This is to restore %Encoding if really needed;
219 sub predefine_encodings{
220 use Encode::Encoding;
221 no warnings 'redefine';
224 # was in Encode::UTF_EBCDIC
225 package Encode::UTF_EBCDIC;
226 push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding';
228 my ($obj,$str,$chk) = @_;
230 for (my $i = 0; $i < length($str); $i++) {
232 chr(utf8::unicode_to_native(ord(substr($str,$i,1))));
238 my ($obj,$str,$chk) = @_;
240 for (my $i = 0; $i < length($str); $i++) {
242 chr(utf8::native_to_unicode(ord(substr($str,$i,1))));
247 $Encode::Encoding{Unicode} =
248 bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC";
250 package Encode::Internal;
251 push @Encode::Internal::ISA, 'Encode::Encoding';
253 my ($obj,$str,$chk) = @_;
259 $Encode::Encoding{Unicode} =
260 bless {Name => "Internal"} => "Encode::Internal";
264 # was in Encode::utf8
265 package Encode::utf8;
266 push @Encode::utf8::ISA, 'Encode::Encoding';
269 Encode::DEBUG and warn __PACKAGE__, " XS on";
270 *decode = \&decode_xs;
271 *encode = \&encode_xs;
273 Encode::DEBUG and warn __PACKAGE__, " XS off";
275 my ($obj,$octets,$chk) = @_;
276 my $str = Encode::decode_utf8($octets);
284 my ($obj,$string,$chk) = @_;
285 my $octets = Encode::encode_utf8($string);
290 *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk)
291 my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk
292 my ($rdst, $rsrc, $rpos) = \@_[1,2,3];
294 if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) {
295 $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm));
296 $$rpos = $npos + length($trm);
299 $$rdst .= substr($$rsrc, $pos);
300 $$rpos = length($$rsrc);
303 $Encode::Encoding{utf8} =
304 bless {Name => "utf8"} => "Encode::utf8";
305 $Encode::Encoding{"utf-8-strict"} =
306 bless {Name => "utf-8-strict", strict_utf8 => 1 } => "Encode::utf8";
316 Encode - character encodings
322 =head2 Table of Contents
324 Encode consists of a collection of modules whose details are too big
325 to fit in one document. This POD itself explains the top-level APIs
326 and general topics at a glance. For other topics and more details,
330 --------------------------------------------------------
331 Encode::Alias Alias definitions to encodings
332 Encode::Encoding Encode Implementation Base Class
333 Encode::Supported List of Supported Encodings
334 Encode::CN Simplified Chinese Encodings
335 Encode::JP Japanese Encodings
336 Encode::KR Korean Encodings
337 Encode::TW Traditional Chinese Encodings
338 --------------------------------------------------------
342 The C<Encode> module provides the interfaces between Perl's strings
343 and the rest of the system. Perl strings are sequences of
346 The repertoire of characters that Perl can represent is at least that
347 defined by the Unicode Consortium. On most platforms the ordinal
348 values of the characters (as returned by C<ord(ch)>) is the "Unicode
349 codepoint" for the character (the exceptions are those platforms where
350 the legacy encoding is some variant of EBCDIC rather than a super-set
351 of ASCII - see L<perlebcdic>).
353 Traditionally, computer data has been moved around in 8-bit chunks
354 often called "bytes". These chunks are also known as "octets" in
355 networking standards. Perl is widely used to manipulate data of many
356 types - not only strings of characters representing human or computer
357 languages but also "binary" data being the machine's representation of
358 numbers, pixels in an image - or just about anything.
360 When Perl is processing "binary data", the programmer wants Perl to
361 process "sequences of bytes". This is not a problem for Perl - as a
362 byte has 256 possible values, it easily fits in Perl's much larger
371 I<character>: a character in the range 0..(2**32-1) (or more).
372 (What Perl's strings are made of.)
376 I<byte>: a character in the range 0..255
377 (A special case of a Perl character.)
381 I<octet>: 8 bits of data, with ordinal values 0..255
382 (Term for bytes passed to or from a non-Perl context, e.g. a disk file.)
386 =head1 PERL ENCODING API
390 =item $octets = encode(ENCODING, $string [, CHECK])
392 Encodes a string from Perl's internal form into I<ENCODING> and returns
393 a sequence of octets. ENCODING can be either a canonical name or
394 an alias. For encoding names and aliases, see L</"Defining Aliases">.
395 For CHECK, see L</"Handling Malformed Data">.
397 For example, to convert a string from Perl's internal format to
398 iso-8859-1 (also known as Latin1),
400 $octets = encode("iso-8859-1", $string);
402 B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets
403 B<may not be equal to> $string. Though they both contain the same data, the utf8 flag
404 for $octets is B<always> off. When you encode anything, utf8 flag of
405 the result is always off, even when it contains completely valid utf8
406 string. See L</"The UTF-8 flag"> below.
408 If the $string is C<undef> then C<undef> is returned.
410 =item $string = decode(ENCODING, $octets [, CHECK])
412 Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's
413 internal form and returns the resulting string. As in encode(),
414 ENCODING can be either a canonical name or an alias. For encoding names
415 and aliases, see L</"Defining Aliases">. For CHECK, see
416 L</"Handling Malformed Data">.
418 For example, to convert ISO-8859-1 data to a string in Perl's internal format:
420 $string = decode("iso-8859-1", $octets);
422 B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string
423 B<may not be equal to> $octets. Though they both contain the same data,
424 the utf8 flag for $string is on unless $octets entirely consists of
425 ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag">
428 If the $string is C<undef> then C<undef> is returned.
430 =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
432 Converts B<in-place> data between two encodings. The data in $octets
433 must be encoded as octets and not as characters in Perl's internal
434 format. For example, to convert ISO-8859-1 data to Microsoft's CP1250
437 from_to($octets, "iso-8859-1", "cp1250");
439 and to convert it back:
441 from_to($octets, "cp1250", "iso-8859-1");
443 Note that because the conversion happens in place, the data to be
444 converted cannot be a string constant; it must be a scalar variable.
446 from_to() returns the length of the converted string in octets on
447 success, I<undef> on error.
449 B<CAVEAT>: The following operations look the same but are not quite so;
451 from_to($data, "iso-8859-1", "utf8"); #1
452 $data = decode("iso-8859-1", $data); #2
454 Both #1 and #2 make $data consist of a completely valid UTF-8 string
455 but only #2 turns utf8 flag on. #1 is equivalent to
457 $data = encode("utf8", decode("iso-8859-1", $data));
459 See L</"The UTF-8 flag"> below.
461 =item $octets = encode_utf8($string);
463 Equivalent to C<$octets = encode("utf8", $string);> The characters
464 that comprise $string are encoded in Perl's internal format and the
465 result is returned as a sequence of octets. All possible
466 characters have a UTF-8 representation so this function cannot fail.
469 =item $string = decode_utf8($octets [, CHECK]);
471 equivalent to C<$string = decode("utf8", $octets [, CHECK])>.
472 The sequence of octets represented by
473 $octets is decoded from UTF-8 into a sequence of logical
474 characters. Not all sequences of octets form valid UTF-8 encodings, so
475 it is possible for this call to fail. For CHECK, see
476 L</"Handling Malformed Data">.
480 =head2 Listing available encodings
483 @list = Encode->encodings();
485 Returns a list of the canonical names of the available encodings that
486 are loaded. To get a list of all available encodings including the
487 ones that are not loaded yet, say
489 @all_encodings = Encode->encodings(":all");
491 Or you can give the name of a specific module.
493 @with_jp = Encode->encodings("Encode::JP");
495 When "::" is not in the name, "Encode::" is assumed.
497 @ebcdic = Encode->encodings("EBCDIC");
499 To find out in detail which encodings are supported by this package,
500 see L<Encode::Supported>.
502 =head2 Defining Aliases
504 To add a new alias to a given encoding, use:
508 define_alias(newName => ENCODING);
510 After that, newName can be used as an alias for ENCODING.
511 ENCODING may be either the name of an encoding or an
514 But before you do so, make sure the alias is nonexistent with
515 C<resolve_alias()>, which returns the canonical name thereof.
518 Encode::resolve_alias("latin1") eq "iso-8859-1" # true
519 Encode::resolve_alias("iso-8859-12") # false; nonexistent
520 Encode::resolve_alias($name) eq $name # true if $name is canonical
522 resolve_alias() does not need C<use Encode::Alias>; it can be
523 exported via C<use Encode qw(resolve_alias)>.
525 See L<Encode::Alias> for details.
527 =head1 Encoding via PerlIO
529 If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode
530 and encode directly via a filehandle. The following two examples
531 are totally identical in their functionality.
534 open my $in, "<:encoding(shiftjis)", $infile or die;
535 open my $out, ">:encoding(euc-jp)", $outfile or die;
536 while(<$in>){ print $out $_; }
539 open my $in, "<", $infile or die;
540 open my $out, ">", $outfile or die;
542 from_to($_, "shiftjis", "euc-jp", 1);
546 Unfortunately, it may be that encodings are PerlIO-savvy. You can check
547 if your encoding is supported by PerlIO by calling the C<perlio_ok>
550 Encode::perlio_ok("hz"); # False
551 find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available
553 use Encode qw(perlio_ok); # exported upon request
556 Fortunately, all encodings that come with Encode core are PerlIO-savvy
557 except for hz and ISO-2022-kr. For gory details, see
558 L<Encode::Encoding> and L<Encode::PerlIO>.
560 =head1 Handling Malformed Data
562 The optional I<CHECK> argument tells Encode what to do when it
563 encounters malformed data. Without CHECK, Encode::FB_DEFAULT ( == 0 )
566 As of version 2.12 Encode supports coderef values for CHECK. See below.
570 =item B<NOTE:> Not all encoding support this feature
572 Some encodings ignore I<CHECK> argument. For example,
573 L<Encode::Unicode> ignores I<CHECK> and it always croaks on error.
577 Now here is the list of I<CHECK> values available
581 =item I<CHECK> = Encode::FB_DEFAULT ( == 0)
583 If I<CHECK> is 0, (en|de)code will put a I<substitution character> in
584 place of a malformed character. When you encode, E<lt>subcharE<gt>
585 will be used. When you decode the code point C<0xFFFD> is used. If
586 the data is supposed to be UTF-8, an optional lexical warning
587 (category utf8) is given.
589 =item I<CHECK> = Encode::FB_CROAK ( == 1)
591 If I<CHECK> is 1, methods will die on error immediately with an error
592 message. Therefore, when I<CHECK> is set to 1, you should trap the
593 error with eval{} unless you really want to let it die.
595 =item I<CHECK> = Encode::FB_QUIET
597 If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately
598 return the portion of the data that has been processed so far when an
599 error occurs. The data argument will be overwritten with everything
600 after that point (that is, the unprocessed part of data). This is
601 handy when you have to call decode repeatedly in the case where your
602 source data may contain partial multi-byte character sequences,
603 (i.e. you are reading with a fixed-width buffer). Here is a sample
604 code that does exactly this:
606 my $buffer = ''; my $string = '';
607 while(read $fh, $buffer, 256, length($buffer)){
608 $string .= decode($encoding, $buffer, Encode::FB_QUIET);
609 # $buffer now contains the unprocessed partial character
612 =item I<CHECK> = Encode::FB_WARN
614 This is the same as above, except that it warns on error. Handy when
615 you are debugging the mode above.
617 =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ)
619 =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF)
621 =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF)
623 For encodings that are implemented by Encode::XS, CHECK ==
624 Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode.
626 When you decode, C<\xI<HH>> will be inserted for a malformed character,
627 where I<HH> is the hex representation of the octet that could not be
628 decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted,
629 where I<HHHH> is the Unicode ID of the character that cannot be found
630 in the character repertoire of the encoding.
632 HTML/XML character reference modes are about the same, in place of
633 C<\x{I<HHHH>}>, HTML uses C<&#I<NNN>;> where I<NNN> is a decimal number and
634 XML uses C<&#xI<HHHH>;> where I<HHHH> is the hexadecimal number.
636 In Encode 2.10 or later, C<LEAVE_SRC> is also implied.
640 These modes are actually set via a bitmask. Here is how the FB_XX
641 constants are laid out. You can import the FB_XX constants via
642 C<use Encode qw(:fallbacks)>; you can import the generic bitmask
643 constants via C<use Encode qw(:fallback_all)>.
645 FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ
648 RETURN_ON_ERR 0x0004 X X
656 =head2 coderef for CHECK
658 As of Encode 2.12 CHECK can also be a code reference which takes the
659 ord value of unmapped caharacter as an argument and returns a string
660 that represents the fallback character. For instance,
662 $ascii = encode("ascii", $utf8, sub{ sprintf "<U+%04X>", shift });
664 Acts like FB_PERLQQ but E<lt>U+I<XXXX>E<gt> is used instead of
667 =head1 Defining Encodings
669 To define a new encoding, use:
671 use Encode qw(define_encoding);
672 define_encoding($object, 'canonicalName' [, alias...]);
674 I<canonicalName> will be associated with I<$object>. The object
675 should provide the interface described in L<Encode::Encoding>.
676 If more than two arguments are provided then additional
677 arguments are taken as aliases for I<$object>.
679 See L<Encode::Encoding> for more details.
681 =head1 The UTF-8 flag
683 Before the introduction of utf8 support in perl, The C<eq> operator
684 just compared the strings represented by two scalars. Beginning with
685 perl 5.8, C<eq> compares two strings with simultaneous consideration
686 of I<the utf8 flag>. To explain why we made it so, I will quote page
687 402 of C<Programming Perl, 3rd ed.>
693 Old byte-oriented programs should not spontaneously break on the old
694 byte-oriented data they used to work on.
698 Old byte-oriented programs should magically start working on the new
699 character-oriented data when appropriate.
703 Programs should run just as fast in the new character-oriented mode
704 as in the old byte-oriented mode.
708 Perl should remain one language, rather than forking into a
709 byte-oriented Perl and a character-oriented Perl.
713 Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0
714 was born and many features documented in the book remained
715 unimplemented for a long time. Perl 5.8 corrected this and the introduction
716 of the UTF-8 flag is one of them. You can think of this perl notion as of a
717 byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8
720 Here is how Encode takes care of the utf8 flag.
726 When you encode, the resulting utf8 flag is always off.
730 When you decode, the resulting utf8 flag is on unless you can
731 unambiguously represent data. Here is the definition of
734 After C<$utf8 = decode('foo', $octet);>,
736 When $octet is... The utf8 flag in $utf8 is
737 ---------------------------------------------
738 In ASCII only (or EBCDIC only) OFF
740 In any other Encoding ON
741 ---------------------------------------------
743 As you see, there is one exception, In ASCII. That way you can assume
744 Goal #1. And with Encode Goal #2 is assumed but you still have to be
745 careful in such cases mentioned in B<CAVEAT> paragraphs.
747 This utf8 flag is not visible in perl scripts, exactly for the same
748 reason you cannot (or you I<don't have to>) see if a scalar contains a
749 string, integer, or floating point number. But you can still peek
750 and poke these if you will. See the section below.
754 =head2 Messing with Perl's Internals
756 The following API uses parts of Perl's internals in the current
757 implementation. As such, they are efficient but may change.
761 =item is_utf8(STRING [, CHECK])
763 [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING.
764 If CHECK is true, also checks the data in STRING for being well-formed
765 UTF-8. Returns true if successful, false otherwise.
767 As of perl 5.8.1, L<utf8> also has utf8::is_utf8().
769 =item _utf8_on(STRING)
771 [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is
772 B<not> checked for being well-formed UTF-8. Do not use unless you
773 B<know> that the STRING is well-formed UTF-8. Returns the previous
774 state of the UTF-8 flag (so please don't treat the return value as
775 indicating success or failure), or C<undef> if STRING is not a string.
777 =item _utf8_off(STRING)
779 [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously.
780 Returns the previous state of the UTF-8 flag (so please don't treat the
781 return value as indicating success or failure), or C<undef> if STRING is
786 =head1 UTF-8 vs. utf8
788 ....We now view strings not as sequences of bytes, but as sequences
789 of numbers in the range 0 .. 2**32-1 (or in the case of 64-bit
790 computers, 0 .. 2**64-1) -- Programming Perl, 3rd ed.
792 That has been the perl's notion of UTF-8 but official UTF-8 is more
793 strict; Its ranges is much narrower (0 .. 10FFFF), some sequences are
794 not allowed (i.e. Those used in the surrogate pair, 0xFFFE, et al).
796 Now that is overruled by Larry Wall himself.
798 From: Larry Wall <larry@wall.org>
799 Date: December 04, 2004 11:51:58 JST
800 To: perl-unicode@perl.org
801 Subject: Re: Make Encode.pm support the real UTF-8
802 Message-Id: <20041204025158.GA28754@wall.org>
804 On Fri, Dec 03, 2004 at 10:12:12PM +0000, Tim Bunce wrote:
805 : I've no problem with 'utf8' being perl's unrestricted uft8 encoding,
806 : but "UTF-8" is the name of the standard and should give the
807 : corresponding behaviour.
809 For what it's worth, that's how I've always kept them straight in my
812 Also for what it's worth, Perl 6 will mostly default to strict but
813 make it easy to switch back to lax.
817 Do you copy? As of Perl 5.8.7, B<UTF-8> means strict, official UTF-8
818 while B<utf8> means liberal, lax, version thereof. And Encode version
819 2.10 or later thus groks the difference between C<UTF-8> and C"utf8".
821 encode("utf8", "\x{FFFF_FFFF}", 1); # okay
822 encode("UTF-8", "\x{FFFF_FFFF}", 1); # croaks
824 C<UTF-8> in Encode is actually a canonical name for C<utf-8-strict>.
825 Yes, the hyphen between "UTF" and "8" is important. Without it Encode
828 find_encoding("UTF-8")->name # is 'utf-8-strict'
829 find_encoding("utf-8")->name # ditto. names are case insensitive
830 find_encoding("utf_8")->name # ditto. "_" are treated as "-"
831 find_encoding("UTF8")->name # is 'utf8'.
837 L<Encode::Supported>,
844 the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt>
848 This project was originated by Nick Ing-Simmons and later maintained
849 by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full
850 list of people involved. For any questions, use
851 E<lt>perl-unicode@perl.orgE<gt> so we can all share.