1Encode(3pm)            Perl Programmers Reference Guide            Encode(3pm)
2
3
4

NAME

6       Encode - character encodings
7

SYNOPSIS

9           use Encode;
10
11   Table of Contents
12       Encode consists of a collection of modules whose details are too big to
13       fit in one document.  This POD itself explains the top-level APIs and
14       general topics at a glance.  For other topics and more details, see the
15       PODs below:
16
17         Name                          Description
18         --------------------------------------------------------
19         Encode::Alias         Alias definitions to encodings
20         Encode::Encoding      Encode Implementation Base Class
21         Encode::Supported     List of Supported Encodings
22         Encode::CN            Simplified Chinese Encodings
23         Encode::JP            Japanese Encodings
24         Encode::KR            Korean Encodings
25         Encode::TW            Traditional Chinese Encodings
26         --------------------------------------------------------
27

DESCRIPTION

29       The "Encode" module provides the interfaces between Perl's strings and
30       the rest of the system.  Perl strings are sequences of characters.
31
32       The repertoire of characters that Perl can represent is at least that
33       defined by the Unicode Consortium. On most platforms the ordinal values
34       of the characters (as returned by "ord(ch)") is the "Unicode codepoint"
35       for the character (the exceptions are those platforms where the legacy
36       encoding is some variant of EBCDIC rather than a super-set of ASCII -
37       see perlebcdic).
38
39       Traditionally, computer data has been moved around in 8-bit chunks
40       often called "bytes". These chunks are also known as "octets" in
41       networking standards. Perl is widely used to manipulate data of many
42       types - not only strings of characters representing human or computer
43       languages but also "binary" data being the machine's representation of
44       numbers, pixels in an image - or just about anything.
45
46       When Perl is processing "binary data", the programmer wants Perl to
47       process "sequences of bytes". This is not a problem for Perl - as a
48       byte has 256 possible values, it easily fits in Perl's much larger
49       "logical character".
50
51   TERMINOLOGY
52       · character: a character in the range 0..(2**32-1) (or more).  (What
53         Perl's strings are made of.)
54
55       · byte: a character in the range 0..255 (A special case of a Perl
56         character.)
57
58       · octet: 8 bits of data, with ordinal values 0..255 (Term for bytes
59         passed to or from a non-Perl context, e.g. a disk file.)
60

PERL ENCODING API

62       $octets  = encode(ENCODING, $string [, CHECK])
63         Encodes a string from Perl's internal form into ENCODING and returns
64         a sequence of octets.  ENCODING can be either a canonical name or an
65         alias.  For encoding names and aliases, see "Defining Aliases".  For
66         CHECK, see "Handling Malformed Data".
67
68         For example, to convert a string from Perl's internal format to
69         iso-8859-1 (also known as Latin1),
70
71           $octets = encode("iso-8859-1", $string);
72
73         CAVEAT: When you run "$octets = encode("utf8", $string)", then
74         $octets may not be equal to $string.  Though they both contain the
75         same data, the UTF8 flag for $octets is always off.  When you encode
76         anything, UTF8 flag of the result is always off, even when it
77         contains completely valid utf8 string. See "The UTF8 flag" below.
78
79         If the $string is "undef" then "undef" is returned.
80
81       $string = decode(ENCODING, $octets [, CHECK])
82         Decodes a sequence of octets assumed to be in ENCODING into Perl's
83         internal form and returns the resulting string.  As in encode(),
84         ENCODING can be either a canonical name or an alias. For encoding
85         names and aliases, see "Defining Aliases".  For CHECK, see "Handling
86         Malformed Data".
87
88         For example, to convert ISO-8859-1 data to a string in Perl's
89         internal format:
90
91           $string = decode("iso-8859-1", $octets);
92
93         CAVEAT: When you run "$string = decode("utf8", $octets)", then
94         $string may not be equal to $octets.  Though they both contain the
95         same data, the UTF8 flag for $string is on unless $octets entirely
96         consists of ASCII data (or EBCDIC on EBCDIC machines).  See "The UTF8
97         flag" below.
98
99         If the $string is "undef" then "undef" is returned.
100
101       [$obj =] find_encoding(ENCODING)
102         Returns the encoding object corresponding to ENCODING.  Returns undef
103         if no matching ENCODING is find.
104
105         This object is what actually does the actual (en|de)coding.
106
107           $utf8 = decode($name, $bytes);
108
109         is in fact
110
111           $utf8 = do{
112             $obj = find_encoding($name);
113             croak qq(encoding "$name" not found) unless ref $obj;
114             $obj->decode($bytes)
115           };
116
117         with more error checking.
118
119         Therefore you can save time by reusing this object as follows;
120
121           my $enc = find_encoding("iso-8859-1");
122           while(<>){
123              my $utf8 = $enc->decode($_);
124              # and do someting with $utf8;
125           }
126
127         Besides "->decode" and "->encode", other methods are available as
128         well.  For instance, "-> name" returns the canonical name of the
129         encoding object.
130
131           find_encoding("latin1")->name; # iso-8859-1
132
133         See Encode::Encoding for details.
134
135       [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK])
136         Converts in-place data between two encodings. The data in $octets
137         must be encoded as octets and not as characters in Perl's internal
138         format. For example, to convert ISO-8859-1 data to Microsoft's CP1250
139         encoding:
140
141           from_to($octets, "iso-8859-1", "cp1250");
142
143         and to convert it back:
144
145           from_to($octets, "cp1250", "iso-8859-1");
146
147         Note that because the conversion happens in place, the data to be
148         converted cannot be a string constant; it must be a scalar variable.
149
150         from_to() returns the length of the converted string in octets on
151         success, undef on error.
152
153         CAVEAT: The following operations look the same but are not quite so;
154
155           from_to($data, "iso-8859-1", "utf8"); #1
156           $data = decode("iso-8859-1", $data);  #2
157
158         Both #1 and #2 make $data consist of a completely valid UTF-8 string
159         but only #2 turns UTF8 flag on.  #1 is equivalent to
160
161           $data = encode("utf8", decode("iso-8859-1", $data));
162
163         See "The UTF8 flag" below.
164
165         Also note that
166
167           from_to($octets, $from, $to, $check);
168
169         is equivalent to
170
171           $octets = encode($to, decode($from, $octets), $check);
172
173         Yes, it does not respect the $check during decoding.  It is
174         deliberately done that way.  If you need minute control, "decode"
175         then "encode" as follows;
176
177           $octets = encode($to, decode($from, $octets, $check_from), $check_to);
178
179       $octets = encode_utf8($string);
180         Equivalent to "$octets = encode("utf8", $string);" The characters
181         that comprise $string are encoded in Perl's internal format and the
182         result is returned as a sequence of octets. All possible characters
183         have a UTF-8 representation so this function cannot fail.
184
185       $string = decode_utf8($octets [, CHECK]);
186         equivalent to "$string = decode("utf8", $octets [, CHECK])".  The
187         sequence of octets represented by $octets is decoded from UTF-8 into
188         a sequence of logical characters. Not all sequences of octets form
189         valid UTF-8 encodings, so it is possible for this call to fail.  For
190         CHECK, see "Handling Malformed Data".
191
192   Listing available encodings
193         use Encode;
194         @list = Encode->encodings();
195
196       Returns a list of the canonical names of the available encodings that
197       are loaded.  To get a list of all available encodings including the
198       ones that are not loaded yet, say
199
200         @all_encodings = Encode->encodings(":all");
201
202       Or you can give the name of a specific module.
203
204         @with_jp = Encode->encodings("Encode::JP");
205
206       When "::" is not in the name, "Encode::" is assumed.
207
208         @ebcdic = Encode->encodings("EBCDIC");
209
210       To find out in detail which encodings are supported by this package,
211       see Encode::Supported.
212
213   Defining Aliases
214       To add a new alias to a given encoding, use:
215
216         use Encode;
217         use Encode::Alias;
218         define_alias(newName => ENCODING);
219
220       After that, newName can be used as an alias for ENCODING.  ENCODING may
221       be either the name of an encoding or an encoding object
222
223       But before you do so, make sure the alias is nonexistent with
224       "resolve_alias()", which returns the canonical name thereof.  i.e.
225
226         Encode::resolve_alias("latin1") eq "iso-8859-1" # true
227         Encode::resolve_alias("iso-8859-12")   # false; nonexistent
228         Encode::resolve_alias($name) eq $name  # true if $name is canonical
229
230       resolve_alias() does not need "use Encode::Alias"; it can be exported
231       via "use Encode qw(resolve_alias)".
232
233       See Encode::Alias for details.
234
235   Finding IANA Character Set Registry names
236       The canonical name of a given encoding does not necessarily agree with
237       IANA IANA Character Set Registry, commonly seen as "Content-Type:
238       text/plain; charset=whatever".  For most cases canonical names work but
239       sometimes it does not (notably 'utf-8-strict').
240
241       Therefore as of Encode version 2.21, a new method "mime_name()" is
242       added.
243
244         use Encode;
245         my $enc = find_encoding('UTF-8');
246         warn $enc->name;      # utf-8-strict
247         warn $enc->mime_name; # UTF-8
248
249       See also:  Encode::Encoding
250

Encoding via PerlIO

252       If your perl supports PerlIO (which is the default), you can use a
253       PerlIO layer to decode and encode directly via a filehandle.  The
254       following two examples are totally identical in their functionality.
255
256         # via PerlIO
257         open my $in,  "<:encoding(shiftjis)", $infile  or die;
258         open my $out, ">:encoding(euc-jp)",   $outfile or die;
259         while(<$in>){ print $out $_; }
260
261         # via from_to
262         open my $in,  "<", $infile  or die;
263         open my $out, ">", $outfile or die;
264         while(<$in>){
265           from_to($_, "shiftjis", "euc-jp", 1);
266           print $out $_;
267         }
268
269       Unfortunately, it may be that encodings are PerlIO-savvy.  You can
270       check if your encoding is supported by PerlIO by calling the
271       "perlio_ok" method.
272
273         Encode::perlio_ok("hz");             # False
274         find_encoding("euc-cn")->perlio_ok;  # True where PerlIO is available
275
276         use Encode qw(perlio_ok);            # exported upon request
277         perlio_ok("euc-jp")
278
279       Fortunately, all encodings that come with Encode core are PerlIO-savvy
280       except for hz and ISO-2022-kr.  For gory details, see Encode::Encoding
281       and Encode::PerlIO.
282

Handling Malformed Data

284       The optional CHECK argument tells Encode what to do when it encounters
285       malformed data.  Without CHECK, Encode::FB_DEFAULT ( == 0 ) is assumed.
286
287       As of version 2.12 Encode supports coderef values for CHECK.  See
288       below.
289
290       NOTE: Not all encoding support this feature
291         Some encodings ignore CHECK argument.  For example, Encode::Unicode
292         ignores CHECK and it always croaks on error.
293
294       Now here is the list of CHECK values available
295
296       CHECK = Encode::FB_DEFAULT ( == 0)
297         If CHECK is 0, (en|de)code will put a substitution character in place
298         of a malformed character.  When you encode, <subchar> will be used.
299         When you decode the code point 0xFFFD is used.  If the data is
300         supposed to be UTF-8, an optional lexical warning (category utf8) is
301         given.
302
303       CHECK = Encode::FB_CROAK ( == 1)
304         If CHECK is 1, methods will die on error immediately with an error
305         message.  Therefore, when CHECK is set to 1,  you should trap the
306         error with eval{} unless you really want to let it die.
307
308       CHECK = Encode::FB_QUIET
309         If CHECK is set to Encode::FB_QUIET, (en|de)code will immediately
310         return the portion of the data that has been processed so far when an
311         error occurs. The data argument will be overwritten with everything
312         after that point (that is, the unprocessed part of data).  This is
313         handy when you have to call decode repeatedly in the case where your
314         source data may contain partial multi-byte character sequences, (i.e.
315         you are reading with a fixed-width buffer). Here is a sample code
316         that does exactly this:
317
318           my $buffer = ''; my $string = '';
319           while(read $fh, $buffer, 256, length($buffer)){
320             $string .= decode($encoding, $buffer, Encode::FB_QUIET);
321             # $buffer now contains the unprocessed partial character
322           }
323
324       CHECK = Encode::FB_WARN
325         This is the same as above, except that it warns on error.  Handy when
326         you are debugging the mode above.
327
328       perlqq mode (CHECK = Encode::FB_PERLQQ)
329       HTML charref mode (CHECK = Encode::FB_HTMLCREF)
330       XML charref mode (CHECK = Encode::FB_XMLCREF)
331         For encodings that are implemented by Encode::XS, CHECK ==
332         Encode::FB_PERLQQ turns (en|de)code into "perlqq" fallback mode.
333
334         When you decode, "\xHH" will be inserted for a malformed character,
335         where HH is the hex representation of the octet  that could not be
336         decoded to utf8.  And when you encode, "\x{HHHH}" will be inserted,
337         where HHHH is the Unicode ID of the character that cannot be found in
338         the character repertoire of the encoding.
339
340         HTML/XML character reference modes are about the same, in place of
341         "\x{HHHH}", HTML uses "&#NNN;" where NNN is a decimal number and XML
342         uses "&#xHHHH;" where HHHH is the hexadecimal number.
343
344         In Encode 2.10 or later, "LEAVE_SRC" is also implied.
345
346       The bitmask
347         These modes are actually set via a bitmask.  Here is how the FB_XX
348         constants are laid out.  You can import the FB_XX constants via "use
349         Encode qw(:fallbacks)"; you can import the generic bitmask constants
350         via "use Encode qw(:fallback_all)".
351
352                              FB_DEFAULT FB_CROAK FB_QUIET FB_WARN  FB_PERLQQ
353          DIE_ON_ERR    0x0001             X
354          WARN_ON_ERR   0x0002                               X
355          RETURN_ON_ERR 0x0004                      X        X
356          LEAVE_SRC     0x0008                                        X
357          PERLQQ        0x0100                                        X
358          HTMLCREF      0x0200
359          XMLCREF       0x0400
360
361       Encode::LEAVE_SRC
362         If the "Encode::LEAVE_SRC" bit is not set, but CHECK is, then the
363         second argument to "encode()" or "decode()" may be assigned to by the
364         functions. If you're not interested in this, then bitwise-or the
365         bitmask with it.
366
367   coderef for CHECK
368       As of Encode 2.12 CHECK can also be a code reference which takes the
369       ord value of unmapped caharacter as an argument and returns a string
370       that represents the fallback character.  For instance,
371
372         $ascii = encode("ascii", $utf8, sub{ sprintf "<U+%04X>", shift });
373
374       Acts like FB_PERLQQ but <U+XXXX> is used instead of \x{XXXX}.
375

Defining Encodings

377       To define a new encoding, use:
378
379           use Encode qw(define_encoding);
380           define_encoding($object, 'canonicalName' [, alias...]);
381
382       canonicalName will be associated with $object.  The object should
383       provide the interface described in Encode::Encoding.  If more than two
384       arguments are provided then additional arguments are taken as aliases
385       for $object.
386
387       See Encode::Encoding for more details.
388

The UTF8 flag

390       Before the introduction of Unicode support in perl, The "eq" operator
391       just compared the strings represented by two scalars. Beginning with
392       perl 5.8, "eq" compares two strings with simultaneous consideration of
393       the UTF8 flag. To explain why we made it so, I will quote page 402 of
394       "Programming Perl, 3rd ed."
395
396       Goal #1:
397         Old byte-oriented programs should not spontaneously break on the old
398         byte-oriented data they used to work on.
399
400       Goal #2:
401         Old byte-oriented programs should magically start working on the new
402         character-oriented data when appropriate.
403
404       Goal #3:
405         Programs should run just as fast in the new character-oriented mode
406         as in the old byte-oriented mode.
407
408       Goal #4:
409         Perl should remain one language, rather than forking into a byte-
410         oriented Perl and a character-oriented Perl.
411
412       Back when "Programming Perl, 3rd ed." was written, not even Perl 5.6.0
413       was born and many features documented in the book remained
414       unimplemented for a long time.  Perl 5.8 corrected this and the
415       introduction of the UTF8 flag is one of them.  You can think of this
416       perl notion as of a byte-oriented mode (UTF8 flag off) and a character-
417       oriented mode (UTF8 flag on).
418
419       Here is how Encode takes care of the UTF8 flag.
420
421       · When you encode, the resulting UTF8 flag is always off.
422
423       · When you decode, the resulting UTF8 flag is on unless you can
424         unambiguously represent data.  Here is the definition of dis-
425         ambiguity.
426
427         After "$utf8 = decode('foo', $octet);",
428
429           When $octet is...   The UTF8 flag in $utf8 is
430           ---------------------------------------------
431           In ASCII only (or EBCDIC only)            OFF
432           In ISO-8859-1                              ON
433           In any other Encoding                      ON
434           ---------------------------------------------
435
436         As you see, there is one exception, In ASCII.  That way you can
437         assume Goal #1.  And with Encode Goal #2 is assumed but you still
438         have to be careful in such cases mentioned in CAVEAT paragraphs.
439
440         This UTF8 flag is not visible in perl scripts, exactly for the same
441         reason you cannot (or you don't have to) see if a scalar contains a
442         string, integer, or floating point number.   But you can still peek
443         and poke these if you will.  See the section below.
444
445   Messing with Perl's Internals
446       The following API uses parts of Perl's internals in the current
447       implementation.  As such, they are efficient but may change.
448
449       is_utf8(STRING [, CHECK])
450         [INTERNAL] Tests whether the UTF8 flag is turned on in the STRING.
451         If CHECK is true, also checks the data in STRING for being well-
452         formed UTF-8.  Returns true if successful, false otherwise.
453
454         As of perl 5.8.1, utf8 also has utf8::is_utf8().
455
456       _utf8_on(STRING)
457         [INTERNAL] Turns on the UTF8 flag in STRING.  The data in STRING is
458         not checked for being well-formed UTF-8.  Do not use unless you know
459         that the STRING is well-formed UTF-8.  Returns the previous state of
460         the UTF8 flag (so please don't treat the return value as indicating
461         success or failure), or "undef" if STRING is not a string.
462
463         This function does not work on tainted values.
464
465       _utf8_off(STRING)
466         [INTERNAL] Turns off the UTF8 flag in STRING.  Do not use
467         frivolously.  Returns the previous state of the UTF8 flag (so please
468         don't treat the return value as indicating success or failure), or
469         "undef" if STRING is not a string.
470
471         This function does not work on tainted values.
472

UTF-8 vs. utf8 vs. UTF8

474         ....We now view strings not as sequences of bytes, but as sequences
475         of numbers in the range 0 .. 2**32-1 (or in the case of 64-bit
476         computers, 0 .. 2**64-1) -- Programming Perl, 3rd ed.
477
478       That has been the perl's notion of UTF-8 but official UTF-8 is more
479       strict; Its ranges is much narrower (0 .. 10FFFF), some sequences are
480       not allowed (i.e. Those used in the surrogate pair, 0xFFFE, et al).
481
482       Now that is overruled by Larry Wall himself.
483
484         From: Larry Wall <larry@wall.org>
485         Date: December 04, 2004 11:51:58 JST
486         To: perl-unicode@perl.org
487         Subject: Re: Make Encode.pm support the real UTF-8
488         Message-Id: <20041204025158.GA28754@wall.org>
489
490         On Fri, Dec 03, 2004 at 10:12:12PM +0000, Tim Bunce wrote:
491         : I've no problem with 'utf8' being perl's unrestricted uft8 encoding,
492         : but "UTF-8" is the name of the standard and should give the
493         : corresponding behaviour.
494
495         For what it's worth, that's how I've always kept them straight in my
496         head.
497
498         Also for what it's worth, Perl 6 will mostly default to strict but
499         make it easy to switch back to lax.
500
501         Larry
502
503       Do you copy?  As of Perl 5.8.7, UTF-8 means strict, official UTF-8
504       while utf8 means liberal, lax, version thereof.  And Encode version
505       2.10 or later thus groks the difference between "UTF-8" and C"utf8".
506
507         encode("utf8",  "\x{FFFF_FFFF}", 1); # okay
508         encode("UTF-8", "\x{FFFF_FFFF}", 1); # croaks
509
510       "UTF-8" in Encode is actually a canonical name for "utf-8-strict".
511       Yes, the hyphen between "UTF" and "8" is important.  Without it Encode
512       goes "liberal"
513
514         find_encoding("UTF-8")->name # is 'utf-8-strict'
515         find_encoding("utf-8")->name # ditto. names are case insensitive
516         find_encoding("utf_8")->name  # ditto. "_" are treated as "-"
517         find_encoding("UTF8")->name  # is 'utf8'.
518
519       The UTF8 flag is internally called UTF8, without a hyphen. It indicates
520       whether a string is internally encoded as utf8, also without a hypen.
521

SEE ALSO

523       Encode::Encoding, Encode::Supported, Encode::PerlIO, encoding,
524       perlebcdic, "open" in perlfunc, perlunicode, perluniintro, perlunifaq,
525       perlunitut utf8, the Perl Unicode Mailing List <perl-unicode@perl.org>
526

MAINTAINER

528       This project was originated by Nick Ing-Simmons and later maintained by
529       Dan Kogai <dankogai@dan.co.jp>.  See AUTHORS for a full list of people
530       involved.  For any questions, use <perl-unicode@perl.org> so we can all
531       share.
532
533       While Dan Kogai retains the copyright as a maintainer, the credit
534       should go to all those involoved.  See AUTHORS for those submitted
535       codes.
536
538       Copyright 2002-2006 Dan Kogai <dankogai@dan.co.jp>
539
540       This library is free software; you can redistribute it and/or modify it
541       under the same terms as Perl itself.
542
543
544
545perl v5.12.4                      2011-06-07                       Encode(3pm)
Impressum