Commit | Line | Data |
---|---|---|
920dae64 AT |
1 | # |
2 | # $Id: Encode.pm,v 2.12 2005/09/08 14:17:17 dankogai Exp dankogai $ | |
3 | # | |
4 | package Encode; | |
5 | use strict; | |
6 | our $VERSION = sprintf "%d.%02d", q$Revision: 2.12 $ =~ /(\d+)/g; | |
7 | sub DEBUG () { 0 } | |
8 | use XSLoader (); | |
9 | XSLoader::load(__PACKAGE__, $VERSION); | |
10 | ||
11 | require Exporter; | |
12 | use base qw/Exporter/; | |
13 | ||
14 | # Public, encouraged API is exported by default | |
15 | ||
16 | our @EXPORT = qw( | |
17 | decode decode_utf8 encode encode_utf8 | |
18 | encodings find_encoding clone_encoding | |
19 | ); | |
20 | ||
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); | |
25 | ||
26 | our @EXPORT_OK = | |
27 | ( | |
28 | qw( | |
29 | _utf8_off _utf8_on define_encoding from_to is_16bit is_8bit | |
30 | is_utf8 perlio_ok resolve_alias utf8_downgrade utf8_upgrade | |
31 | ), | |
32 | @FB_FLAGS, @FB_CONSTS, | |
33 | ); | |
34 | ||
35 | our %EXPORT_TAGS = | |
36 | ( | |
37 | all => [ @EXPORT, @EXPORT_OK ], | |
38 | fallbacks => [ @FB_CONSTS ], | |
39 | fallback_all => [ @FB_CONSTS, @FB_FLAGS ], | |
40 | ); | |
41 | ||
42 | # Documentation moved after __END__ for speed - NI-S | |
43 | ||
44 | our $ON_EBCDIC = (ord("A") == 193); | |
45 | ||
46 | use Encode::Alias; | |
47 | ||
48 | # Make a %Encoding package variable to allow a certain amount of cheating | |
49 | our %Encoding; | |
50 | our %ExtModule; | |
51 | require Encode::Config; | |
52 | eval { require Encode::ConfigLocal }; | |
53 | ||
54 | sub encodings | |
55 | { | |
56 | my $class = shift; | |
57 | my %enc; | |
58 | if (@_ and $_[0] eq ":all"){ | |
59 | %enc = ( %Encoding, %ExtModule ); | |
60 | }else{ | |
61 | %enc = %Encoding; | |
62 | for my $mod (map {m/::/o ? $_ : "Encode::$_" } @_){ | |
63 | DEBUG and warn $mod; | |
64 | for my $enc (keys %ExtModule){ | |
65 | $ExtModule{$enc} eq $mod and $enc{$enc} = $mod; | |
66 | } | |
67 | } | |
68 | } | |
69 | return | |
70 | sort { lc $a cmp lc $b } | |
71 | grep {!/^(?:Internal|Unicode|Guess)$/o} keys %enc; | |
72 | } | |
73 | ||
74 | sub perlio_ok{ | |
75 | my $obj = ref($_[0]) ? $_[0] : find_encoding($_[0]); | |
76 | $obj->can("perlio_ok") and return $obj->perlio_ok(); | |
77 | return 0; # safety net | |
78 | } | |
79 | ||
80 | sub define_encoding | |
81 | { | |
82 | my $obj = shift; | |
83 | my $name = shift; | |
84 | $Encoding{$name} = $obj; | |
85 | my $lc = lc($name); | |
86 | define_alias($lc => $obj) unless $lc eq $name; | |
87 | while (@_){ | |
88 | my $alias = shift; | |
89 | define_alias($alias, $obj); | |
90 | } | |
91 | return $obj; | |
92 | } | |
93 | ||
94 | sub getEncoding | |
95 | { | |
96 | my ($class, $name, $skip_external) = @_; | |
97 | ||
98 | ref($name) && $name->can('renew') and return $name; | |
99 | exists $Encoding{$name} and return $Encoding{$name}; | |
100 | my $lc = lc $name; | |
101 | exists $Encoding{$lc} and return $Encoding{$lc}; | |
102 | ||
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 $oc; | |
107 | ||
108 | unless ($skip_external) | |
109 | { | |
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}; | |
114 | } | |
115 | } | |
116 | return; | |
117 | } | |
118 | ||
119 | sub find_encoding($;$) | |
120 | { | |
121 | my ($name, $skip_external) = @_; | |
122 | return __PACKAGE__->getEncoding($name,$skip_external); | |
123 | } | |
124 | ||
125 | sub resolve_alias($){ | |
126 | my $obj = find_encoding(shift); | |
127 | defined $obj and return $obj->name; | |
128 | return; | |
129 | } | |
130 | ||
131 | sub clone_encoding($){ | |
132 | my $obj = find_encoding(shift); | |
133 | ref $obj or return; | |
134 | eval { require Storable }; | |
135 | $@ and return; | |
136 | return Storable::dclone($obj); | |
137 | } | |
138 | ||
139 | sub encode($$;$) | |
140 | { | |
141 | my ($name, $string, $check) = @_; | |
142 | return undef unless defined $string; | |
143 | $string .= '' if ref $string; # stringify; | |
144 | $check ||=0; | |
145 | my $enc = find_encoding($name); | |
146 | unless(defined $enc){ | |
147 | require Carp; | |
148 | Carp::croak("Unknown encoding '$name'"); | |
149 | } | |
150 | my $octets = $enc->encode($string,$check); | |
151 | $_[1] = $string if $check and !($check & LEAVE_SRC()); | |
152 | return $octets; | |
153 | } | |
154 | ||
155 | sub decode($$;$) | |
156 | { | |
157 | my ($name,$octets,$check) = @_; | |
158 | return undef unless defined $octets; | |
159 | $octets .= '' if ref $octets; | |
160 | $check ||=0; | |
161 | my $enc = find_encoding($name); | |
162 | unless(defined $enc){ | |
163 | require Carp; | |
164 | Carp::croak("Unknown encoding '$name'"); | |
165 | } | |
166 | my $string = $enc->decode($octets,$check); | |
167 | $_[1] = $octets if $check and !($check & LEAVE_SRC()); | |
168 | return $string; | |
169 | } | |
170 | ||
171 | sub from_to($$$;$) | |
172 | { | |
173 | my ($string,$from,$to,$check) = @_; | |
174 | return undef unless defined $string; | |
175 | $check ||=0; | |
176 | my $f = find_encoding($from); | |
177 | unless (defined $f){ | |
178 | require Carp; | |
179 | Carp::croak("Unknown encoding '$from'"); | |
180 | } | |
181 | my $t = find_encoding($to); | |
182 | unless (defined $t){ | |
183 | require Carp; | |
184 | Carp::croak("Unknown encoding '$to'"); | |
185 | } | |
186 | my $uni = $f->decode($string,$check); | |
187 | return undef if ($check && length($string)); | |
188 | $string = $t->encode($uni,$check); | |
189 | return undef if ($check && length($uni)); | |
190 | return defined($_[0] = $string) ? length($string) : undef ; | |
191 | } | |
192 | ||
193 | sub encode_utf8($) | |
194 | { | |
195 | my ($str) = @_; | |
196 | utf8::encode($str); | |
197 | return $str; | |
198 | } | |
199 | ||
200 | sub decode_utf8($;$) | |
201 | { | |
202 | my ($str, $check) = @_; | |
203 | if ($check){ | |
204 | return decode("utf8", $str, $check); | |
205 | }else{ | |
206 | return decode("utf8", $str); | |
207 | return $str; | |
208 | } | |
209 | } | |
210 | ||
211 | predefine_encodings(1); | |
212 | ||
213 | # | |
214 | # This is to restore %Encoding if really needed; | |
215 | # | |
216 | ||
217 | sub predefine_encodings{ | |
218 | use Encode::Encoding; | |
219 | no warnings 'redefine'; | |
220 | my $use_xs = shift; | |
221 | if ($ON_EBCDIC) { | |
222 | # was in Encode::UTF_EBCDIC | |
223 | package Encode::UTF_EBCDIC; | |
224 | push @Encode::UTF_EBCDIC::ISA, 'Encode::Encoding'; | |
225 | *decode = sub{ | |
226 | my ($obj,$str,$chk) = @_; | |
227 | my $res = ''; | |
228 | for (my $i = 0; $i < length($str); $i++) { | |
229 | $res .= | |
230 | chr(utf8::unicode_to_native(ord(substr($str,$i,1)))); | |
231 | } | |
232 | $_[1] = '' if $chk; | |
233 | return $res; | |
234 | }; | |
235 | *encode = sub{ | |
236 | my ($obj,$str,$chk) = @_; | |
237 | my $res = ''; | |
238 | for (my $i = 0; $i < length($str); $i++) { | |
239 | $res .= | |
240 | chr(utf8::native_to_unicode(ord(substr($str,$i,1)))); | |
241 | } | |
242 | $_[1] = '' if $chk; | |
243 | return $res; | |
244 | }; | |
245 | $Encode::Encoding{Unicode} = | |
246 | bless {Name => "UTF_EBCDIC"} => "Encode::UTF_EBCDIC"; | |
247 | } else { | |
248 | package Encode::Internal; | |
249 | push @Encode::Internal::ISA, 'Encode::Encoding'; | |
250 | *decode = sub{ | |
251 | my ($obj,$str,$chk) = @_; | |
252 | utf8::upgrade($str); | |
253 | $_[1] = '' if $chk; | |
254 | return $str; | |
255 | }; | |
256 | *encode = \&decode; | |
257 | $Encode::Encoding{Unicode} = | |
258 | bless {Name => "Internal"} => "Encode::Internal"; | |
259 | } | |
260 | ||
261 | { | |
262 | # was in Encode::utf8 | |
263 | package Encode::utf8; | |
264 | push @Encode::utf8::ISA, 'Encode::Encoding'; | |
265 | # | |
266 | if ($use_xs){ | |
267 | Encode::DEBUG and warn __PACKAGE__, " XS on"; | |
268 | *decode = \&decode_xs; | |
269 | *encode = \&encode_xs; | |
270 | }else{ | |
271 | Encode::DEBUG and warn __PACKAGE__, " XS off"; | |
272 | *decode = sub{ | |
273 | my ($obj,$octets,$chk) = @_; | |
274 | my $str = Encode::decode_utf8($octets); | |
275 | if (defined $str) { | |
276 | $_[1] = '' if $chk; | |
277 | return $str; | |
278 | } | |
279 | return undef; | |
280 | }; | |
281 | *encode = sub { | |
282 | my ($obj,$string,$chk) = @_; | |
283 | my $octets = Encode::encode_utf8($string); | |
284 | $_[1] = '' if $chk; | |
285 | return $octets; | |
286 | }; | |
287 | } | |
288 | *cat_decode = sub{ # ($obj, $dst, $src, $pos, $trm, $chk) | |
289 | my ($obj, undef, undef, $pos, $trm) = @_; # currently ignores $chk | |
290 | my ($rdst, $rsrc, $rpos) = \@_[1,2,3]; | |
291 | use bytes; | |
292 | if ((my $npos = index($$rsrc, $trm, $pos)) >= 0) { | |
293 | $$rdst .= substr($$rsrc, $pos, $npos - $pos + length($trm)); | |
294 | $$rpos = $npos + length($trm); | |
295 | return 1; | |
296 | } | |
297 | $$rdst .= substr($$rsrc, $pos); | |
298 | $$rpos = length($$rsrc); | |
299 | return ''; | |
300 | }; | |
301 | $Encode::Encoding{utf8} = | |
302 | bless {Name => "utf8"} => "Encode::utf8"; | |
303 | $Encode::Encoding{"utf-8-strict"} = | |
304 | bless {Name => "utf-8-strict", strict_utf8 => 1 } => "Encode::utf8"; | |
305 | } | |
306 | } | |
307 | ||
308 | 1; | |
309 | ||
310 | __END__ | |
311 | ||
312 | =head1 NAME | |
313 | ||
314 | Encode - character encodings | |
315 | ||
316 | =head1 SYNOPSIS | |
317 | ||
318 | use Encode; | |
319 | ||
320 | =head2 Table of Contents | |
321 | ||
322 | Encode consists of a collection of modules whose details are too big | |
323 | to fit in one document. This POD itself explains the top-level APIs | |
324 | and general topics at a glance. For other topics and more details, | |
325 | see the PODs below: | |
326 | ||
327 | Name Description | |
328 | -------------------------------------------------------- | |
329 | Encode::Alias Alias definitions to encodings | |
330 | Encode::Encoding Encode Implementation Base Class | |
331 | Encode::Supported List of Supported Encodings | |
332 | Encode::CN Simplified Chinese Encodings | |
333 | Encode::JP Japanese Encodings | |
334 | Encode::KR Korean Encodings | |
335 | Encode::TW Traditional Chinese Encodings | |
336 | -------------------------------------------------------- | |
337 | ||
338 | =head1 DESCRIPTION | |
339 | ||
340 | The C<Encode> module provides the interfaces between Perl's strings | |
341 | and the rest of the system. Perl strings are sequences of | |
342 | B<characters>. | |
343 | ||
344 | The repertoire of characters that Perl can represent is at least that | |
345 | defined by the Unicode Consortium. On most platforms the ordinal | |
346 | values of the characters (as returned by C<ord(ch)>) is the "Unicode | |
347 | codepoint" for the character (the exceptions are those platforms where | |
348 | the legacy encoding is some variant of EBCDIC rather than a super-set | |
349 | of ASCII - see L<perlebcdic>). | |
350 | ||
351 | Traditionally, computer data has been moved around in 8-bit chunks | |
352 | often called "bytes". These chunks are also known as "octets" in | |
353 | networking standards. Perl is widely used to manipulate data of many | |
354 | types - not only strings of characters representing human or computer | |
355 | languages but also "binary" data being the machine's representation of | |
356 | numbers, pixels in an image - or just about anything. | |
357 | ||
358 | When Perl is processing "binary data", the programmer wants Perl to | |
359 | process "sequences of bytes". This is not a problem for Perl - as a | |
360 | byte has 256 possible values, it easily fits in Perl's much larger | |
361 | "logical character". | |
362 | ||
363 | =head2 TERMINOLOGY | |
364 | ||
365 | =over 2 | |
366 | ||
367 | =item * | |
368 | ||
369 | I<character>: a character in the range 0..(2**32-1) (or more). | |
370 | (What Perl's strings are made of.) | |
371 | ||
372 | =item * | |
373 | ||
374 | I<byte>: a character in the range 0..255 | |
375 | (A special case of a Perl character.) | |
376 | ||
377 | =item * | |
378 | ||
379 | I<octet>: 8 bits of data, with ordinal values 0..255 | |
380 | (Term for bytes passed to or from a non-Perl context, e.g. a disk file.) | |
381 | ||
382 | =back | |
383 | ||
384 | =head1 PERL ENCODING API | |
385 | ||
386 | =over 2 | |
387 | ||
388 | =item $octets = encode(ENCODING, $string [, CHECK]) | |
389 | ||
390 | Encodes a string from Perl's internal form into I<ENCODING> and returns | |
391 | a sequence of octets. ENCODING can be either a canonical name or | |
392 | an alias. For encoding names and aliases, see L</"Defining Aliases">. | |
393 | For CHECK, see L</"Handling Malformed Data">. | |
394 | ||
395 | For example, to convert a string from Perl's internal format to | |
396 | iso-8859-1 (also known as Latin1), | |
397 | ||
398 | $octets = encode("iso-8859-1", $string); | |
399 | ||
400 | B<CAVEAT>: When you run C<$octets = encode("utf8", $string)>, then $octets | |
401 | B<may not be equal to> $string. Though they both contain the same data, the utf8 flag | |
402 | for $octets is B<always> off. When you encode anything, utf8 flag of | |
403 | the result is always off, even when it contains completely valid utf8 | |
404 | string. See L</"The UTF-8 flag"> below. | |
405 | ||
406 | If the $string is C<undef> then C<undef> is returned. | |
407 | ||
408 | =item $string = decode(ENCODING, $octets [, CHECK]) | |
409 | ||
410 | Decodes a sequence of octets assumed to be in I<ENCODING> into Perl's | |
411 | internal form and returns the resulting string. As in encode(), | |
412 | ENCODING can be either a canonical name or an alias. For encoding names | |
413 | and aliases, see L</"Defining Aliases">. For CHECK, see | |
414 | L</"Handling Malformed Data">. | |
415 | ||
416 | For example, to convert ISO-8859-1 data to a string in Perl's internal format: | |
417 | ||
418 | $string = decode("iso-8859-1", $octets); | |
419 | ||
420 | B<CAVEAT>: When you run C<$string = decode("utf8", $octets)>, then $string | |
421 | B<may not be equal to> $octets. Though they both contain the same data, | |
422 | the utf8 flag for $string is on unless $octets entirely consists of | |
423 | ASCII data (or EBCDIC on EBCDIC machines). See L</"The UTF-8 flag"> | |
424 | below. | |
425 | ||
426 | If the $string is C<undef> then C<undef> is returned. | |
427 | ||
428 | =item [$length =] from_to($octets, FROM_ENC, TO_ENC [, CHECK]) | |
429 | ||
430 | Converts B<in-place> data between two encodings. The data in $octets | |
431 | must be encoded as octets and not as characters in Perl's internal | |
432 | format. For example, to convert ISO-8859-1 data to Microsoft's CP1250 | |
433 | encoding: | |
434 | ||
435 | from_to($octets, "iso-8859-1", "cp1250"); | |
436 | ||
437 | and to convert it back: | |
438 | ||
439 | from_to($octets, "cp1250", "iso-8859-1"); | |
440 | ||
441 | Note that because the conversion happens in place, the data to be | |
442 | converted cannot be a string constant; it must be a scalar variable. | |
443 | ||
444 | from_to() returns the length of the converted string in octets on | |
445 | success, I<undef> on error. | |
446 | ||
447 | B<CAVEAT>: The following operations look the same but are not quite so; | |
448 | ||
449 | from_to($data, "iso-8859-1", "utf8"); #1 | |
450 | $data = decode("iso-8859-1", $data); #2 | |
451 | ||
452 | Both #1 and #2 make $data consist of a completely valid UTF-8 string | |
453 | but only #2 turns utf8 flag on. #1 is equivalent to | |
454 | ||
455 | $data = encode("utf8", decode("iso-8859-1", $data)); | |
456 | ||
457 | See L</"The UTF-8 flag"> below. | |
458 | ||
459 | =item $octets = encode_utf8($string); | |
460 | ||
461 | Equivalent to C<$octets = encode("utf8", $string);> The characters | |
462 | that comprise $string are encoded in Perl's internal format and the | |
463 | result is returned as a sequence of octets. All possible | |
464 | characters have a UTF-8 representation so this function cannot fail. | |
465 | ||
466 | ||
467 | =item $string = decode_utf8($octets [, CHECK]); | |
468 | ||
469 | equivalent to C<$string = decode("utf8", $octets [, CHECK])>. | |
470 | The sequence of octets represented by | |
471 | $octets is decoded from UTF-8 into a sequence of logical | |
472 | characters. Not all sequences of octets form valid UTF-8 encodings, so | |
473 | it is possible for this call to fail. For CHECK, see | |
474 | L</"Handling Malformed Data">. | |
475 | ||
476 | =back | |
477 | ||
478 | =head2 Listing available encodings | |
479 | ||
480 | use Encode; | |
481 | @list = Encode->encodings(); | |
482 | ||
483 | Returns a list of the canonical names of the available encodings that | |
484 | are loaded. To get a list of all available encodings including the | |
485 | ones that are not loaded yet, say | |
486 | ||
487 | @all_encodings = Encode->encodings(":all"); | |
488 | ||
489 | Or you can give the name of a specific module. | |
490 | ||
491 | @with_jp = Encode->encodings("Encode::JP"); | |
492 | ||
493 | When "::" is not in the name, "Encode::" is assumed. | |
494 | ||
495 | @ebcdic = Encode->encodings("EBCDIC"); | |
496 | ||
497 | To find out in detail which encodings are supported by this package, | |
498 | see L<Encode::Supported>. | |
499 | ||
500 | =head2 Defining Aliases | |
501 | ||
502 | To add a new alias to a given encoding, use: | |
503 | ||
504 | use Encode; | |
505 | use Encode::Alias; | |
506 | define_alias(newName => ENCODING); | |
507 | ||
508 | After that, newName can be used as an alias for ENCODING. | |
509 | ENCODING may be either the name of an encoding or an | |
510 | I<encoding object> | |
511 | ||
512 | But before you do so, make sure the alias is nonexistent with | |
513 | C<resolve_alias()>, which returns the canonical name thereof. | |
514 | i.e. | |
515 | ||
516 | Encode::resolve_alias("latin1") eq "iso-8859-1" # true | |
517 | Encode::resolve_alias("iso-8859-12") # false; nonexistent | |
518 | Encode::resolve_alias($name) eq $name # true if $name is canonical | |
519 | ||
520 | resolve_alias() does not need C<use Encode::Alias>; it can be | |
521 | exported via C<use Encode qw(resolve_alias)>. | |
522 | ||
523 | See L<Encode::Alias> for details. | |
524 | ||
525 | =head1 Encoding via PerlIO | |
526 | ||
527 | If your perl supports I<PerlIO> (which is the default), you can use a PerlIO layer to decode | |
528 | and encode directly via a filehandle. The following two examples | |
529 | are totally identical in their functionality. | |
530 | ||
531 | # via PerlIO | |
532 | open my $in, "<:encoding(shiftjis)", $infile or die; | |
533 | open my $out, ">:encoding(euc-jp)", $outfile or die; | |
534 | while(<$in>){ print $out $_; } | |
535 | ||
536 | # via from_to | |
537 | open my $in, "<", $infile or die; | |
538 | open my $out, ">", $outfile or die; | |
539 | while(<$in>){ | |
540 | from_to($_, "shiftjis", "euc-jp", 1); | |
541 | print $out $_; | |
542 | } | |
543 | ||
544 | Unfortunately, it may be that encodings are PerlIO-savvy. You can check | |
545 | if your encoding is supported by PerlIO by calling the C<perlio_ok> | |
546 | method. | |
547 | ||
548 | Encode::perlio_ok("hz"); # False | |
549 | find_encoding("euc-cn")->perlio_ok; # True where PerlIO is available | |
550 | ||
551 | use Encode qw(perlio_ok); # exported upon request | |
552 | perlio_ok("euc-jp") | |
553 | ||
554 | Fortunately, all encodings that come with Encode core are PerlIO-savvy | |
555 | except for hz and ISO-2022-kr. For gory details, see | |
556 | L<Encode::Encoding> and L<Encode::PerlIO>. | |
557 | ||
558 | =head1 Handling Malformed Data | |
559 | ||
560 | The optional I<CHECK> argument tells Encode what to do when it | |
561 | encounters malformed data. Without CHECK, Encode::FB_DEFAULT ( == 0 ) | |
562 | is assumed. | |
563 | ||
564 | As of version 2.12 Encode supports coderef values for CHECK. See below. | |
565 | ||
566 | =over 2 | |
567 | ||
568 | =item B<NOTE:> Not all encoding support this feature | |
569 | ||
570 | Some encodings ignore I<CHECK> argument. For example, | |
571 | L<Encode::Unicode> ignores I<CHECK> and it always croaks on error. | |
572 | ||
573 | =back | |
574 | ||
575 | Now here is the list of I<CHECK> values available | |
576 | ||
577 | =over 2 | |
578 | ||
579 | =item I<CHECK> = Encode::FB_DEFAULT ( == 0) | |
580 | ||
581 | If I<CHECK> is 0, (en|de)code will put a I<substitution character> in | |
582 | place of a malformed character. When you encode, E<lt>subcharE<gt> | |
583 | will be used. When you decode the code point C<0xFFFD> is used. If | |
584 | the data is supposed to be UTF-8, an optional lexical warning | |
585 | (category utf8) is given. | |
586 | ||
587 | =item I<CHECK> = Encode::FB_CROAK ( == 1) | |
588 | ||
589 | If I<CHECK> is 1, methods will die on error immediately with an error | |
590 | message. Therefore, when I<CHECK> is set to 1, you should trap the | |
591 | error with eval{} unless you really want to let it die. | |
592 | ||
593 | =item I<CHECK> = Encode::FB_QUIET | |
594 | ||
595 | If I<CHECK> is set to Encode::FB_QUIET, (en|de)code will immediately | |
596 | return the portion of the data that has been processed so far when an | |
597 | error occurs. The data argument will be overwritten with everything | |
598 | after that point (that is, the unprocessed part of data). This is | |
599 | handy when you have to call decode repeatedly in the case where your | |
600 | source data may contain partial multi-byte character sequences, | |
601 | (i.e. you are reading with a fixed-width buffer). Here is a sample | |
602 | code that does exactly this: | |
603 | ||
604 | my $buffer = ''; my $string = ''; | |
605 | while(read $fh, $buffer, 256, length($buffer)){ | |
606 | $string .= decode($encoding, $buffer, Encode::FB_QUIET); | |
607 | # $buffer now contains the unprocessed partial character | |
608 | } | |
609 | ||
610 | =item I<CHECK> = Encode::FB_WARN | |
611 | ||
612 | This is the same as above, except that it warns on error. Handy when | |
613 | you are debugging the mode above. | |
614 | ||
615 | =item perlqq mode (I<CHECK> = Encode::FB_PERLQQ) | |
616 | ||
617 | =item HTML charref mode (I<CHECK> = Encode::FB_HTMLCREF) | |
618 | ||
619 | =item XML charref mode (I<CHECK> = Encode::FB_XMLCREF) | |
620 | ||
621 | For encodings that are implemented by Encode::XS, CHECK == | |
622 | Encode::FB_PERLQQ turns (en|de)code into C<perlqq> fallback mode. | |
623 | ||
624 | When you decode, C<\xI<HH>> will be inserted for a malformed character, | |
625 | where I<HH> is the hex representation of the octet that could not be | |
626 | decoded to utf8. And when you encode, C<\x{I<HHHH>}> will be inserted, | |
627 | where I<HHHH> is the Unicode ID of the character that cannot be found | |
628 | in the character repertoire of the encoding. | |
629 | ||
630 | HTML/XML character reference modes are about the same, in place of | |
631 | C<\x{I<HHHH>}>, HTML uses C<&#I<NNN>;> where I<NNN> is a decimal number and | |
632 | XML uses C<&#xI<HHHH>;> where I<HHHH> is the hexadecimal number. | |
633 | ||
634 | In Encode 2.10 or later, C<LEAVE_SRC> is also implied. | |
635 | ||
636 | =item The bitmask | |
637 | ||
638 | These modes are actually set via a bitmask. Here is how the FB_XX | |
639 | constants are laid out. You can import the FB_XX constants via | |
640 | C<use Encode qw(:fallbacks)>; you can import the generic bitmask | |
641 | constants via C<use Encode qw(:fallback_all)>. | |
642 | ||
643 | FB_DEFAULT FB_CROAK FB_QUIET FB_WARN FB_PERLQQ | |
644 | DIE_ON_ERR 0x0001 X | |
645 | WARN_ON_ERR 0x0002 X | |
646 | RETURN_ON_ERR 0x0004 X X | |
647 | LEAVE_SRC 0x0008 X | |
648 | PERLQQ 0x0100 X | |
649 | HTMLCREF 0x0200 | |
650 | XMLCREF 0x0400 | |
651 | ||
652 | =back | |
653 | ||
654 | =head2 coderef for CHECK | |
655 | ||
656 | As of Encode 2.12 CHECK can also be a code reference which takes the | |
657 | ord value of unmapped caharacter as an argument and returns a string | |
658 | that represents the fallback character. For instance, | |
659 | ||
660 | $ascii = encode("ascii", $utf8, sub{ sprintf "<U+%04X>", shift }); | |
661 | ||
662 | Acts like FB_PERLQQ but E<lt>U+I<XXXX>E<gt> is used instead of | |
663 | \x{I<XXXX>}. | |
664 | ||
665 | =head1 Defining Encodings | |
666 | ||
667 | To define a new encoding, use: | |
668 | ||
669 | use Encode qw(define_encoding); | |
670 | define_encoding($object, 'canonicalName' [, alias...]); | |
671 | ||
672 | I<canonicalName> will be associated with I<$object>. The object | |
673 | should provide the interface described in L<Encode::Encoding>. | |
674 | If more than two arguments are provided then additional | |
675 | arguments are taken as aliases for I<$object>. | |
676 | ||
677 | See L<Encode::Encoding> for more details. | |
678 | ||
679 | =head1 The UTF-8 flag | |
680 | ||
681 | Before the introduction of utf8 support in perl, The C<eq> operator | |
682 | just compared the strings represented by two scalars. Beginning with | |
683 | perl 5.8, C<eq> compares two strings with simultaneous consideration | |
684 | of I<the utf8 flag>. To explain why we made it so, I will quote page | |
685 | 402 of C<Programming Perl, 3rd ed.> | |
686 | ||
687 | =over 2 | |
688 | ||
689 | =item Goal #1: | |
690 | ||
691 | Old byte-oriented programs should not spontaneously break on the old | |
692 | byte-oriented data they used to work on. | |
693 | ||
694 | =item Goal #2: | |
695 | ||
696 | Old byte-oriented programs should magically start working on the new | |
697 | character-oriented data when appropriate. | |
698 | ||
699 | =item Goal #3: | |
700 | ||
701 | Programs should run just as fast in the new character-oriented mode | |
702 | as in the old byte-oriented mode. | |
703 | ||
704 | =item Goal #4: | |
705 | ||
706 | Perl should remain one language, rather than forking into a | |
707 | byte-oriented Perl and a character-oriented Perl. | |
708 | ||
709 | =back | |
710 | ||
711 | Back when C<Programming Perl, 3rd ed.> was written, not even Perl 5.6.0 | |
712 | was born and many features documented in the book remained | |
713 | unimplemented for a long time. Perl 5.8 corrected this and the introduction | |
714 | of the UTF-8 flag is one of them. You can think of this perl notion as of a | |
715 | byte-oriented mode (utf8 flag off) and a character-oriented mode (utf8 | |
716 | flag on). | |
717 | ||
718 | Here is how Encode takes care of the utf8 flag. | |
719 | ||
720 | =over 2 | |
721 | ||
722 | =item * | |
723 | ||
724 | When you encode, the resulting utf8 flag is always off. | |
725 | ||
726 | =item * | |
727 | ||
728 | When you decode, the resulting utf8 flag is on unless you can | |
729 | unambiguously represent data. Here is the definition of | |
730 | dis-ambiguity. | |
731 | ||
732 | After C<$utf8 = decode('foo', $octet);>, | |
733 | ||
734 | When $octet is... The utf8 flag in $utf8 is | |
735 | --------------------------------------------- | |
736 | In ASCII only (or EBCDIC only) OFF | |
737 | In ISO-8859-1 ON | |
738 | In any other Encoding ON | |
739 | --------------------------------------------- | |
740 | ||
741 | As you see, there is one exception, In ASCII. That way you can assume | |
742 | Goal #1. And with Encode Goal #2 is assumed but you still have to be | |
743 | careful in such cases mentioned in B<CAVEAT> paragraphs. | |
744 | ||
745 | This utf8 flag is not visible in perl scripts, exactly for the same | |
746 | reason you cannot (or you I<don't have to>) see if a scalar contains a | |
747 | string, integer, or floating point number. But you can still peek | |
748 | and poke these if you will. See the section below. | |
749 | ||
750 | =back | |
751 | ||
752 | =head2 Messing with Perl's Internals | |
753 | ||
754 | The following API uses parts of Perl's internals in the current | |
755 | implementation. As such, they are efficient but may change. | |
756 | ||
757 | =over 2 | |
758 | ||
759 | =item is_utf8(STRING [, CHECK]) | |
760 | ||
761 | [INTERNAL] Tests whether the UTF-8 flag is turned on in the STRING. | |
762 | If CHECK is true, also checks the data in STRING for being well-formed | |
763 | UTF-8. Returns true if successful, false otherwise. | |
764 | ||
765 | As of perl 5.8.1, L<utf8> also has utf8::is_utf8(). | |
766 | ||
767 | =item _utf8_on(STRING) | |
768 | ||
769 | [INTERNAL] Turns on the UTF-8 flag in STRING. The data in STRING is | |
770 | B<not> checked for being well-formed UTF-8. Do not use unless you | |
771 | B<know> that the STRING is well-formed UTF-8. Returns the previous | |
772 | state of the UTF-8 flag (so please don't treat the return value as | |
773 | indicating success or failure), or C<undef> if STRING is not a string. | |
774 | ||
775 | =item _utf8_off(STRING) | |
776 | ||
777 | [INTERNAL] Turns off the UTF-8 flag in STRING. Do not use frivolously. | |
778 | Returns the previous state of the UTF-8 flag (so please don't treat the | |
779 | return value as indicating success or failure), or C<undef> if STRING is | |
780 | not a string. | |
781 | ||
782 | =back | |
783 | ||
784 | =head1 UTF-8 vs. utf8 | |
785 | ||
786 | ....We now view strings not as sequences of bytes, but as sequences | |
787 | of numbers in the range 0 .. 2**32-1 (or in the case of 64-bit | |
788 | computers, 0 .. 2**64-1) -- Programming Perl, 3rd ed. | |
789 | ||
790 | That has been the perl's notion of UTF-8 but official UTF-8 is more | |
791 | strict; Its ranges is much narrower (0 .. 10FFFF), some sequences are | |
792 | not allowed (i.e. Those used in the surrogate pair, 0xFFFE, et al). | |
793 | ||
794 | Now that is overruled by Larry Wall himself. | |
795 | ||
796 | From: Larry Wall <larry@wall.org> | |
797 | Date: December 04, 2004 11:51:58 JST | |
798 | To: perl-unicode@perl.org | |
799 | Subject: Re: Make Encode.pm support the real UTF-8 | |
800 | Message-Id: <20041204025158.GA28754@wall.org> | |
801 | ||
802 | On Fri, Dec 03, 2004 at 10:12:12PM +0000, Tim Bunce wrote: | |
803 | : I've no problem with 'utf8' being perl's unrestricted uft8 encoding, | |
804 | : but "UTF-8" is the name of the standard and should give the | |
805 | : corresponding behaviour. | |
806 | ||
807 | For what it's worth, that's how I've always kept them straight in my | |
808 | head. | |
809 | ||
810 | Also for what it's worth, Perl 6 will mostly default to strict but | |
811 | make it easy to switch back to lax. | |
812 | ||
813 | Larry | |
814 | ||
815 | Do you copy? As of Perl 5.8.7, B<UTF-8> means strict, official UTF-8 | |
816 | while B<utf8> means liberal, lax, version thereof. And Encode version | |
817 | 2.10 or later thus groks the difference between C<UTF-8> and C"utf8". | |
818 | ||
819 | encode("utf8", "\x{FFFF_FFFF}", 1); # okay | |
820 | encode("UTF-8", "\x{FFFF_FFFF}", 1); # croaks | |
821 | ||
822 | C<UTF-8> in Encode is actually a canonical name for C<utf-8-strict>. | |
823 | Yes, the hyphen between "UTF" and "8" is important. Without it Encode | |
824 | goes "liberal" | |
825 | ||
826 | find_encoding("UTF-8")->name # is 'utf-8-strict' | |
827 | find_encoding("utf-8")->name # ditto. names are case insensitive | |
828 | find_encoding("utf8")->name # ditto. "_" are treated as "-" | |
829 | find_encoding("UTF8")->name # is 'utf8'. | |
830 | ||
831 | ||
832 | =head1 SEE ALSO | |
833 | ||
834 | L<Encode::Encoding>, | |
835 | L<Encode::Supported>, | |
836 | L<Encode::PerlIO>, | |
837 | L<encoding>, | |
838 | L<perlebcdic>, | |
839 | L<perlfunc/open>, | |
840 | L<perlunicode>, | |
841 | L<utf8>, | |
842 | the Perl Unicode Mailing List E<lt>perl-unicode@perl.orgE<gt> | |
843 | ||
844 | =head1 MAINTAINER | |
845 | ||
846 | This project was originated by Nick Ing-Simmons and later maintained | |
847 | by Dan Kogai E<lt>dankogai@dan.co.jpE<gt>. See AUTHORS for a full | |
848 | list of people involved. For any questions, use | |
849 | E<lt>perl-unicode@perl.orgE<gt> so we can all share. | |
850 | ||
851 | =cut |