Commit | Line | Data |
2c674647 |
1 | package Encode; |
51ef4e11 |
2 | use strict; |
2c674647 |
3 | |
b8a524e9 |
4 | our $VERSION = '0.02'; |
2c674647 |
5 | |
6 | require DynaLoader; |
7 | require Exporter; |
8 | |
51ef4e11 |
9 | our @ISA = qw(Exporter DynaLoader); |
2c674647 |
10 | |
4411f3b6 |
11 | # Public, encouraged API is exported by default |
51ef4e11 |
12 | our @EXPORT = qw ( |
4411f3b6 |
13 | encode |
14 | decode |
15 | encode_utf8 |
16 | decode_utf8 |
17 | find_encoding |
51ef4e11 |
18 | encodings |
4411f3b6 |
19 | ); |
20 | |
51ef4e11 |
21 | our @EXPORT_OK = |
2c674647 |
22 | qw( |
51ef4e11 |
23 | define_encoding |
24 | define_alias |
2c674647 |
25 | from_to |
26 | is_utf8 |
4411f3b6 |
27 | is_8bit |
28 | is_16bit |
a12c0f56 |
29 | utf8_upgrade |
30 | utf8_downgrade |
4411f3b6 |
31 | _utf8_on |
32 | _utf8_off |
2c674647 |
33 | ); |
34 | |
35 | bootstrap Encode (); |
36 | |
4411f3b6 |
37 | # Documentation moved after __END__ for speed - NI-S |
2c674647 |
38 | |
bf230f3d |
39 | use Carp; |
40 | |
51ef4e11 |
41 | # Make a %encoding package variable to allow a certain amount of cheating |
42 | our %encoding; |
43 | my @alias; # ordered matching list |
44 | my %alias; # cached known aliases |
f7ac3676 |
45 | |
6d6a7c8d |
46 | # 0 1 2 3 4 5 6 7 8 9 10 |
47 | our @latin2iso_num = ( 0, 1, 2, 3, 4, 9, 10, 13, 14, 15, 16 ); |
48 | |
f7ac3676 |
49 | our %winlatin2cp = ( |
50 | 'Latin1' => 1252, |
51 | 'Latin2' => 1250, |
52 | 'Cyrillic' => 1251, |
f7ac3676 |
53 | 'Greek' => 1253, |
54 | 'Turkish' => 1254, |
55 | 'Hebrew' => 1255, |
56 | 'Arabic' => 1256, |
57 | 'Baltic' => 1257, |
58 | 'Vietnamese' => 1258, |
59 | ); |
5345d506 |
60 | |
656753f8 |
61 | sub encodings |
62 | { |
63 | my ($class) = @_; |
51ef4e11 |
64 | return keys %encoding; |
65 | } |
66 | |
67 | sub findAlias |
68 | { |
69 | my $class = shift; |
70 | local $_ = shift; |
1e616cf5 |
71 | # print "# findAlias $_\n"; |
51ef4e11 |
72 | unless (exists $alias{$_}) |
656753f8 |
73 | { |
51ef4e11 |
74 | for (my $i=0; $i < @alias; $i += 2) |
656753f8 |
75 | { |
51ef4e11 |
76 | my $alias = $alias[$i]; |
77 | my $val = $alias[$i+1]; |
78 | my $new; |
79 | if (ref($alias) eq 'Regexp' && $_ =~ $alias) |
5345d506 |
80 | { |
51ef4e11 |
81 | $new = eval $val; |
82 | } |
83 | elsif (ref($alias) eq 'CODE') |
84 | { |
85 | $new = &{$alias}($val) |
86 | } |
5ad8ef52 |
87 | elsif (lc($_) eq lc($alias)) |
51ef4e11 |
88 | { |
89 | $new = $val; |
90 | } |
91 | if (defined($new)) |
92 | { |
93 | next if $new eq $_; # avoid (direct) recursion on bugs |
94 | my $enc = (ref($new)) ? $new : find_encoding($new); |
95 | if ($enc) |
5345d506 |
96 | { |
51ef4e11 |
97 | $alias{$_} = $enc; |
98 | last; |
5345d506 |
99 | } |
100 | } |
656753f8 |
101 | } |
5345d506 |
102 | } |
51ef4e11 |
103 | return $alias{$_}; |
5345d506 |
104 | } |
105 | |
51ef4e11 |
106 | sub define_alias |
5345d506 |
107 | { |
51ef4e11 |
108 | while (@_) |
5345d506 |
109 | { |
51ef4e11 |
110 | my ($alias,$name) = splice(@_,0,2); |
111 | push(@alias, $alias => $name); |
656753f8 |
112 | } |
51ef4e11 |
113 | } |
114 | |
016cb72c |
115 | # Allow variants of iso-8859-1 etc. |
d6089a2a |
116 | define_alias( qr/^iso[-_]?(\d+)[-_](\d+)$/i => '"iso-$1-$2"' ); |
016cb72c |
117 | |
7faf300d |
118 | # At least HP-UX has these. |
119 | define_alias( qr/^iso8859(\d+)$/i => '"iso-8859-$1"' ); |
120 | |
f7ac3676 |
121 | # More HP stuff. |
122 | define_alias( qr/^(?:hp-)?(arabic|greek|hebrew|kana|roman|thai|turkish)8$/i => '"${1}8"' ); |
123 | |
8a361256 |
124 | # The Official name of ASCII. |
125 | define_alias( qr/^ANSI[-_]?X3\.4[-_]?1968$/i => '"ascii"' ); |
126 | |
58d53262 |
127 | # This is a font issue, not an encoding issue. |
128 | # (The currency symbol of the Latin 1 upper half |
129 | # has been redefined as the euro symbol.) |
130 | define_alias( qr/^(.+)\@euro$/i => '"$1"' ); |
131 | |
016cb72c |
132 | # Allow latin-1 style names as well |
7faf300d |
133 | define_alias( qr/^(?:iso[-_]?)?latin[-_]?(\d+)$/i => '"iso-8859-$latin2iso_num[$1]"' ); |
016cb72c |
134 | |
f7ac3676 |
135 | # Allow winlatin1 style names as well |
cf91068f |
136 | define_alias( qr/^win(latin[12]|cyrillic|baltic|greek|turkish|hebrew|arabic|baltic|vietnamese)$/i => '"cp$winlatin2cp{\u$1}"' ); |
f7ac3676 |
137 | |
016cb72c |
138 | # Common names for non-latin prefered MIME names |
139 | define_alias( 'ascii' => 'US-ascii', |
140 | 'cyrillic' => 'iso-8859-5', |
141 | 'arabic' => 'iso-8859-6', |
142 | 'greek' => 'iso-8859-7', |
f7ac3676 |
143 | 'hebrew' => 'iso-8859-8', |
144 | 'thai' => 'iso-8859-11', |
145 | 'tis620' => 'iso-8859-11', |
146 | ); |
016cb72c |
147 | |
7faf300d |
148 | # At least AIX has IBM-NNN (surprisingly...) instead of cpNNN. |
149 | define_alias( qr/^ibm[-_]?(\d\d\d\d?)$/i => '"cp$1"'); |
150 | |
58d53262 |
151 | # Standardize on the dashed versions. |
152 | define_alias( qr/^utf8$/i => 'utf-8' ); |
7faf300d |
153 | define_alias( qr/^koi8r$/i => 'koi8-r' ); |
f7ac3676 |
154 | define_alias( qr/^koi8u$/i => 'koi8-u' ); |
155 | |
156 | # TODO: HP-UX '8' encodings arabic8 greek8 hebrew8 kana8 thai8 turkish8 |
157 | # TODO: HP-UX '15' encodings japanese15 korean15 roi15 |
158 | # TODO: Cyrillic encoding ISO-IR-111 (useful?) |
159 | # TODO: Chinese encodings GB18030 GBK Big5-HSKCS EUC-TW |
160 | # TODO: Armenian encoding ARMSCII-8 |
161 | # TODO: Hebrew encoding ISO-8859-8-1 |
162 | # TODO: Thai encoding TCVN |
163 | # TODO: Korean encoding Johab |
56a543c5 |
164 | # TODO: Vietnamese encodings VPS |
f7ac3676 |
165 | # TODO: Japanese encoding JIS (not the same as SJIS) |
166 | # TODO: Mac Asian+African encodings: Arabic Armenian Bengali Burmese |
167 | # ChineseSimp ChineseTrad Devanagari Ethiopic ExtArabic |
168 | # Farsi Georgian Gujarati Gurmukhi Hebrew Japanese |
169 | # Kannada Khmer Korean Laotian Malayalam Mongolian |
170 | # Oriya Sinhalese Symbol Tamil Telugu Tibetan Vietnamese |
171 | # TODO: what is the Japanese 'UJIS' encoding seen in some Linuxes? |
58d53262 |
172 | |
016cb72c |
173 | # Map white space and _ to '-' |
174 | define_alias( qr/^(\S+)[\s_]+(.*)$/i => '"$1-$2"' ); |
175 | |
51ef4e11 |
176 | sub define_encoding |
177 | { |
178 | my $obj = shift; |
179 | my $name = shift; |
180 | $encoding{$name} = $obj; |
181 | my $lc = lc($name); |
182 | define_alias($lc => $obj) unless $lc eq $name; |
183 | while (@_) |
656753f8 |
184 | { |
51ef4e11 |
185 | my $alias = shift; |
186 | define_alias($alias,$obj); |
656753f8 |
187 | } |
51ef4e11 |
188 | return $obj; |
656753f8 |
189 | } |
190 | |
656753f8 |
191 | sub getEncoding |
192 | { |
193 | my ($class,$name) = @_; |
5345d506 |
194 | my $enc; |
0f43fc90 |
195 | if (ref($name) && $name->can('new_sequence')) |
196 | { |
197 | return $name; |
198 | } |
1e616cf5 |
199 | my $lc = lc $name; |
51ef4e11 |
200 | if (exists $encoding{$name}) |
656753f8 |
201 | { |
51ef4e11 |
202 | return $encoding{$name}; |
203 | } |
1e616cf5 |
204 | if (exists $encoding{$lc}) |
51ef4e11 |
205 | { |
1e616cf5 |
206 | return $encoding{$lc}; |
656753f8 |
207 | } |
1e616cf5 |
208 | |
209 | my $oc = $class->findAlias($name); |
210 | return $oc if defined $oc; |
211 | return $class->findAlias($lc) if $lc ne $name; |
212 | |
213 | return; |
656753f8 |
214 | } |
215 | |
4411f3b6 |
216 | sub find_encoding |
217 | { |
218 | my ($name) = @_; |
219 | return __PACKAGE__->getEncoding($name); |
220 | } |
221 | |
222 | sub encode |
223 | { |
224 | my ($name,$string,$check) = @_; |
225 | my $enc = find_encoding($name); |
226 | croak("Unknown encoding '$name'") unless defined $enc; |
50d26985 |
227 | my $octets = $enc->encode($string,$check); |
4411f3b6 |
228 | return undef if ($check && length($string)); |
229 | return $octets; |
230 | } |
231 | |
232 | sub decode |
233 | { |
234 | my ($name,$octets,$check) = @_; |
235 | my $enc = find_encoding($name); |
236 | croak("Unknown encoding '$name'") unless defined $enc; |
50d26985 |
237 | my $string = $enc->decode($octets,$check); |
96d6357c |
238 | $_[1] = $octets if $check; |
4411f3b6 |
239 | return $string; |
240 | } |
241 | |
242 | sub from_to |
243 | { |
244 | my ($string,$from,$to,$check) = @_; |
245 | my $f = find_encoding($from); |
246 | croak("Unknown encoding '$from'") unless defined $f; |
247 | my $t = find_encoding($to); |
248 | croak("Unknown encoding '$to'") unless defined $t; |
50d26985 |
249 | my $uni = $f->decode($string,$check); |
4411f3b6 |
250 | return undef if ($check && length($string)); |
50d26985 |
251 | $string = $t->encode($uni,$check); |
4411f3b6 |
252 | return undef if ($check && length($uni)); |
253 | return length($_[0] = $string); |
254 | } |
255 | |
256 | sub encode_utf8 |
257 | { |
258 | my ($str) = @_; |
1b026014 |
259 | utf8::encode($str); |
4411f3b6 |
260 | return $str; |
261 | } |
262 | |
263 | sub decode_utf8 |
264 | { |
265 | my ($str) = @_; |
1b026014 |
266 | return undef unless utf8::decode($str); |
4411f3b6 |
267 | return $str; |
268 | } |
269 | |
50d26985 |
270 | package Encode::Encoding; |
271 | # Base class for classes which implement encodings |
4edaa979 |
272 | |
51ef4e11 |
273 | sub Define |
274 | { |
275 | my $obj = shift; |
276 | my $canonical = shift; |
277 | $obj = bless { Name => $canonical },$obj unless ref $obj; |
278 | # warn "$canonical => $obj\n"; |
279 | Encode::define_encoding($obj, $canonical, @_); |
280 | } |
281 | |
282 | sub name { shift->{'Name'} } |
283 | |
50d26985 |
284 | # Temporary legacy methods |
4edaa979 |
285 | sub toUnicode { shift->decode(@_) } |
286 | sub fromUnicode { shift->encode(@_) } |
287 | |
288 | sub new_sequence { return $_[0] } |
50d26985 |
289 | |
290 | package Encode::XS; |
291 | use base 'Encode::Encoding'; |
292 | |
5ad8ef52 |
293 | package Encode::Internal; |
50d26985 |
294 | use base 'Encode::Encoding'; |
656753f8 |
295 | |
9b37254d |
296 | # Dummy package that provides the encode interface but leaves data |
1b026014 |
297 | # as UTF-X encoded. It is here so that from_to() works. |
656753f8 |
298 | |
5ad8ef52 |
299 | __PACKAGE__->Define('Internal'); |
300 | |
301 | Encode::define_alias( 'Unicode' => 'Internal' ) if ord('A') == 65; |
656753f8 |
302 | |
50d26985 |
303 | sub decode |
a12c0f56 |
304 | { |
305 | my ($obj,$str,$chk) = @_; |
1b026014 |
306 | utf8::upgrade($str); |
a12c0f56 |
307 | $_[1] = '' if $chk; |
308 | return $str; |
309 | } |
656753f8 |
310 | |
50d26985 |
311 | *encode = \&decode; |
656753f8 |
312 | |
5ad8ef52 |
313 | package Encoding::Unicode; |
314 | use base 'Encode::Encoding'; |
315 | |
316 | __PACKAGE__->Define('Unicode') unless ord('A') == 65; |
317 | |
318 | sub decode |
319 | { |
320 | my ($obj,$str,$chk) = @_; |
321 | my $res = ''; |
322 | for (my $i = 0; $i < length($str); $i++) |
323 | { |
324 | $res .= chr(utf8::unicode_to_native(ord(substr($str,$i,1)))); |
325 | } |
326 | $_[1] = '' if $chk; |
327 | return $res; |
328 | } |
329 | |
330 | sub encode |
331 | { |
332 | my ($obj,$str,$chk) = @_; |
333 | my $res = ''; |
334 | for (my $i = 0; $i < length($str); $i++) |
335 | { |
336 | $res .= chr(utf8::native_to_unicode(ord(substr($str,$i,1)))); |
337 | } |
338 | $_[1] = '' if $chk; |
339 | return $res; |
340 | } |
341 | |
342 | |
4411f3b6 |
343 | package Encode::utf8; |
50d26985 |
344 | use base 'Encode::Encoding'; |
4411f3b6 |
345 | # package to allow long-hand |
346 | # $octets = encode( utf8 => $string ); |
347 | # |
348 | |
51ef4e11 |
349 | __PACKAGE__->Define(qw(UTF-8 utf8)); |
4411f3b6 |
350 | |
50d26985 |
351 | sub decode |
4411f3b6 |
352 | { |
353 | my ($obj,$octets,$chk) = @_; |
2a936312 |
354 | my $str = Encode::decode_utf8($octets); |
4411f3b6 |
355 | if (defined $str) |
356 | { |
357 | $_[1] = '' if $chk; |
358 | return $str; |
359 | } |
360 | return undef; |
361 | } |
362 | |
50d26985 |
363 | sub encode |
4411f3b6 |
364 | { |
365 | my ($obj,$string,$chk) = @_; |
2a936312 |
366 | my $octets = Encode::encode_utf8($string); |
4411f3b6 |
367 | $_[1] = '' if $chk; |
368 | return $octets; |
4411f3b6 |
369 | } |
370 | |
9b37254d |
371 | package Encode::iso10646_1; |
50d26985 |
372 | use base 'Encode::Encoding'; |
51ef4e11 |
373 | # Encoding is 16-bit network order Unicode (no surogates) |
9b37254d |
374 | # Used for X font encodings |
87714904 |
375 | |
8040349a |
376 | __PACKAGE__->Define(qw(UCS-2 iso-10646-1)); |
87714904 |
377 | |
50d26985 |
378 | sub decode |
87714904 |
379 | { |
380 | my ($obj,$str,$chk) = @_; |
381 | my $uni = ''; |
382 | while (length($str)) |
383 | { |
5dcbab34 |
384 | my $code = unpack('n',substr($str,0,2,'')) & 0xffff; |
87714904 |
385 | $uni .= chr($code); |
386 | } |
387 | $_[1] = $str if $chk; |
8040349a |
388 | utf8::upgrade($uni); |
87714904 |
389 | return $uni; |
390 | } |
391 | |
50d26985 |
392 | sub encode |
87714904 |
393 | { |
394 | my ($obj,$uni,$chk) = @_; |
395 | my $str = ''; |
396 | while (length($uni)) |
397 | { |
398 | my $ch = substr($uni,0,1,''); |
399 | my $x = ord($ch); |
400 | unless ($x < 32768) |
401 | { |
402 | last if ($chk); |
403 | $x = 0; |
404 | } |
5dcbab34 |
405 | $str .= pack('n',$x); |
656753f8 |
406 | } |
bf230f3d |
407 | $_[1] = $uni if $chk; |
656753f8 |
408 | return $str; |
409 | } |
410 | |
79019f4f |
411 | package Encode::ucs_2le; |
412 | use base 'Encode::Encoding'; |
413 | |
414 | __PACKAGE__->Define(qw(UCS-2le UCS-2LE ucs-2le)); |
415 | |
416 | sub decode |
417 | { |
418 | my ($obj,$str,$chk) = @_; |
419 | my $uni = ''; |
420 | while (length($str)) |
421 | { |
422 | my $code = unpack('v',substr($str,0,2,'')) & 0xffff; |
423 | $uni .= chr($code); |
424 | } |
425 | $_[1] = $str if $chk; |
426 | utf8::upgrade($uni); |
427 | return $uni; |
428 | } |
429 | |
430 | sub encode |
431 | { |
432 | my ($obj,$uni,$chk) = @_; |
433 | my $str = ''; |
434 | while (length($uni)) |
435 | { |
436 | my $ch = substr($uni,0,1,''); |
437 | my $x = ord($ch); |
438 | unless ($x < 32768) |
439 | { |
440 | last if ($chk); |
441 | $x = 0; |
442 | } |
443 | $str .= pack('v',$x); |
444 | } |
445 | $_[1] = $uni if $chk; |
446 | return $str; |
447 | } |
448 | |
4411f3b6 |
449 | # switch back to Encode package in case we ever add AutoLoader |
450 | package Encode; |
451 | |
656753f8 |
452 | 1; |
453 | |
2a936312 |
454 | __END__ |
455 | |
4411f3b6 |
456 | =head1 NAME |
457 | |
458 | Encode - character encodings |
459 | |
460 | =head1 SYNOPSIS |
461 | |
462 | use Encode; |
463 | |
464 | =head1 DESCRIPTION |
465 | |
47bfe92f |
466 | The C<Encode> module provides the interfaces between Perl's strings |
467 | and the rest of the system. Perl strings are sequences of B<characters>. |
4411f3b6 |
468 | |
469 | The repertoire of characters that Perl can represent is at least that |
47bfe92f |
470 | defined by the Unicode Consortium. On most platforms the ordinal |
471 | values of the characters (as returned by C<ord(ch)>) is the "Unicode |
472 | codepoint" for the character (the exceptions are those platforms where |
473 | the legacy encoding is some variant of EBCDIC rather than a super-set |
474 | of ASCII - see L<perlebcdic>). |
4411f3b6 |
475 | |
476 | Traditionaly computer data has been moved around in 8-bit chunks |
477 | often called "bytes". These chunks are also known as "octets" in |
478 | networking standards. Perl is widely used to manipulate data of |
479 | many types - not only strings of characters representing human or |
480 | computer languages but also "binary" data being the machines representation |
481 | of numbers, pixels in an image - or just about anything. |
482 | |
47bfe92f |
483 | When Perl is processing "binary data" the programmer wants Perl to process |
484 | "sequences of bytes". This is not a problem for Perl - as a byte has 256 |
485 | possible values it easily fits in Perl's much larger "logical character". |
4411f3b6 |
486 | |
487 | =head2 TERMINOLOGY |
488 | |
4ac9195f |
489 | =over 4 |
4411f3b6 |
490 | |
491 | =item * |
492 | |
493 | I<character>: a character in the range 0..(2**32-1) (or more). |
47bfe92f |
494 | (What Perl's strings are made of.) |
4411f3b6 |
495 | |
496 | =item * |
497 | |
498 | I<byte>: a character in the range 0..255 |
47bfe92f |
499 | (A special case of a Perl character.) |
4411f3b6 |
500 | |
501 | =item * |
502 | |
503 | I<octet>: 8 bits of data, with ordinal values 0..255 |
47bfe92f |
504 | (Term for bytes passed to or from a non-Perl context, e.g. disk file.) |
4411f3b6 |
505 | |
506 | =back |
507 | |
508 | The marker [INTERNAL] marks Internal Implementation Details, in |
509 | general meant only for those who think they know what they are doing, |
510 | and such details may change in future releases. |
511 | |
512 | =head1 ENCODINGS |
513 | |
514 | =head2 Characteristics of an Encoding |
515 | |
516 | An encoding has a "repertoire" of characters that it can represent, |
517 | and for each representable character there is at least one sequence of |
518 | octets that represents it. |
519 | |
520 | =head2 Types of Encodings |
521 | |
522 | Encodings can be divided into the following types: |
523 | |
524 | =over 4 |
525 | |
526 | =item * Fixed length 8-bit (or less) encodings. |
527 | |
528 | Each character is a single octet so may have a repertoire of up to |
529 | 256 characters. ASCII and iso-8859-* are typical examples. |
530 | |
531 | =item * Fixed length 16-bit encodings |
532 | |
533 | Each character is two octets so may have a repertoire of up to |
47bfe92f |
534 | 65 536 characters. Unicode's UCS-2 is an example. Also used for |
4411f3b6 |
535 | encodings for East Asian languages. |
536 | |
537 | =item * Fixed length 32-bit encodings. |
538 | |
539 | Not really very "encoded" encodings. The Unicode code points |
540 | are just represented as 4-octet integers. None the less because |
541 | different architectures use different representations of integers |
542 | (so called "endian") there at least two disctinct encodings. |
543 | |
544 | =item * Multi-byte encodings |
545 | |
546 | The number of octets needed to represent a character varies. |
547 | UTF-8 is a particularly complex but regular case of a multi-byte |
548 | encoding. Several East Asian countries use a multi-byte encoding |
549 | where 1-octet is used to cover western roman characters and Asian |
550 | characters get 2-octets. |
551 | (UTF-16 is strictly a multi-byte encoding taking either 2 or 4 octets |
552 | to represent a Unicode code point.) |
553 | |
554 | =item * "Escape" encodings. |
555 | |
556 | These encodings embed "escape sequences" into the octet sequence |
557 | which describe how the following octets are to be interpreted. |
558 | The iso-2022-* family is typical. Following the escape sequence |
559 | octets are encoded by an "embedded" encoding (which will be one |
560 | of the above types) until another escape sequence switches to |
561 | a different "embedded" encoding. |
562 | |
563 | These schemes are very flexible and can handle mixed languages but are |
47bfe92f |
564 | very complex to process (and have state). No escape encodings are |
565 | implemented for Perl yet. |
4411f3b6 |
566 | |
567 | =back |
568 | |
569 | =head2 Specifying Encodings |
570 | |
571 | Encodings can be specified to the API described below in two ways: |
572 | |
573 | =over 4 |
574 | |
575 | =item 1. By name |
576 | |
47bfe92f |
577 | Encoding names are strings with characters taken from a restricted |
578 | repertoire. See L</"Encoding Names">. |
4411f3b6 |
579 | |
580 | =item 2. As an object |
581 | |
582 | Encoding objects are returned by C<find_encoding($name)>. |
583 | |
584 | =back |
585 | |
586 | =head2 Encoding Names |
587 | |
588 | Encoding names are case insensitive. White space in names is ignored. |
47bfe92f |
589 | In addition an encoding may have aliases. Each encoding has one |
590 | "canonical" name. The "canonical" name is chosen from the names of |
591 | the encoding by picking the first in the following sequence: |
4411f3b6 |
592 | |
593 | =over 4 |
594 | |
78255929 |
595 | =item * The MIME name as defined in IETF RFCs. |
4411f3b6 |
596 | |
597 | =item * The name in the IANA registry. |
598 | |
d1be9408 |
599 | =item * The name used by the organization that defined it. |
4411f3b6 |
600 | |
601 | =back |
602 | |
603 | Because of all the alias issues, and because in the general case |
604 | encodings have state C<Encode> uses the encoding object internally |
605 | once an operation is in progress. |
606 | |
21938dfa |
607 | As of Perl 5.8.0, at least the following encodings are recognized |
608 | (the => marks aliases): |
609 | |
610 | ASCII |
611 | |
612 | US-ASCII => ASCII |
613 | |
614 | The Unicode: |
615 | |
616 | UTF-8 |
617 | UTF-16 |
618 | UCS-2 |
619 | |
620 | ISO 10646-1 => UCS-2 |
621 | |
622 | The ISO 8859 and KOI: |
623 | |
624 | ISO 8859-1 ISO 8859-6 ISO 8859-11 KOI8-F |
625 | ISO 8859-2 ISO 8859-7 (12 doesn't exist) KOI8-R |
56a543c5 |
626 | ISO 8859-3 ISO 8859-8 ISO 8859-13 KOI8-U |
21938dfa |
627 | ISO 8859-4 ISO 8859-9 ISO 8859-14 |
628 | ISO 8859-5 ISO 8859-10 ISO 8859-15 |
629 | ISO 8859-16 |
630 | |
631 | Latin1 => 8859-1 Latin6 => 8859-10 |
632 | Latin2 => 8859-2 Latin7 => 8859-13 |
633 | Latin3 => 8859-3 Latin8 => 8859-14 |
634 | Latin4 => 8859-4 Latin9 => 8859-15 |
635 | Latin5 => 8859-9 Latin10 => 8859-16 |
636 | |
637 | Cyrillic => 8859-5 |
638 | Arabic => 8859-6 |
639 | Greek => 8859-7 |
640 | Hebrew => 8859-8 |
641 | Thai => 8859-11 |
642 | TIS620 => 8859-11 |
643 | |
644 | The CJKV: Chinese, Japanese, Korean, Vietnamese: |
645 | |
646 | ISO 2022 ISO 2022 JP-1 JIS 0201 GB 1988 Big5 EUC-CN |
56a543c5 |
647 | ISO 2022 CN ISO 2022 JP-2 JIS 0208 GB 2312 HZ EUC-JP |
21938dfa |
648 | ISO 2022 JP ISO 2022 KR JIS 0210 GB 12345 CNS 11643 EUC-JP-0212 |
56a543c5 |
649 | Shift-JIS EUC-KR |
21938dfa |
650 | VISCII |
651 | |
652 | The PC codepages: |
653 | |
654 | CP37 CP852 CP861 CP866 CP949 CP1251 CP1256 |
655 | CP424 CP855 CP862 CP869 CP950 CP1252 CP1257 |
656 | CP737 CP856 CP863 CP874 CP1006 CP1253 CP1258 |
657 | CP775 CP857 CP864 CP932 CP1047 CP1254 |
658 | CP850 CP860 CP865 CP936 CP1250 CP1255 |
659 | |
660 | WinLatin1 => CP1252 |
661 | WinLatin2 => CP1250 |
662 | WinCyrillic => CP1251 |
663 | WinGreek => CP1253 |
664 | WinTurkiskh => CP1254 |
665 | WinHebrew => CP1255 |
666 | WinArabic => CP1256 |
667 | WinBaltic => CP1257 |
668 | WinVietnamese => CP1258 |
669 | |
4a42e14c |
670 | (All the CPI<NNN...> are available also as IBMI<NNN...>.) |
21938dfa |
671 | |
672 | The Mac codepages: |
673 | |
674 | MacCentralEuropean MacJapanese |
56a543c5 |
675 | MacCroatian MacRoman |
676 | MacCyrillic MacRumanian |
677 | MacDingbats MacSami |
678 | MacGreek MacThai |
679 | MacIcelandic MacTurkish |
680 | MacUkraine |
21938dfa |
681 | |
682 | Miscellaneous: |
683 | |
684 | 7bit-greek IR-197 |
685 | 7bit-kana NeXTstep |
686 | 7bit-latin1 POSIX-BC |
687 | DingBats Roman8 |
688 | GSM 0338 Symbol |
689 | |
4411f3b6 |
690 | =head1 PERL ENCODING API |
691 | |
692 | =head2 Generic Encoding Interface |
693 | |
694 | =over 4 |
695 | |
696 | =item * |
697 | |
698 | $bytes = encode(ENCODING, $string[, CHECK]) |
699 | |
47bfe92f |
700 | Encodes string from Perl's internal form into I<ENCODING> and returns |
701 | a sequence of octets. For CHECK see L</"Handling Malformed Data">. |
4411f3b6 |
702 | |
703 | =item * |
704 | |
705 | $string = decode(ENCODING, $bytes[, CHECK]) |
706 | |
47bfe92f |
707 | Decode sequence of octets assumed to be in I<ENCODING> into Perl's |
708 | internal form and returns the resulting string. For CHECK see |
709 | L</"Handling Malformed Data">. |
710 | |
711 | =item * |
712 | |
713 | from_to($string, FROM_ENCODING, TO_ENCODING[, CHECK]) |
714 | |
2b106fbe |
715 | Convert B<in-place> the data between two encodings. How did the data |
716 | in $string originally get to be in FROM_ENCODING? Either using |
e9692b5b |
717 | encode() or through PerlIO: See L</"Encoding and IO">. For CHECK |
2b106fbe |
718 | see L</"Handling Malformed Data">. |
719 | |
720 | For example to convert ISO 8859-1 data to UTF-8: |
721 | |
722 | from_to($data, "iso-8859-1", "utf-8"); |
723 | |
724 | and to convert it back: |
725 | |
726 | from_to($data, "utf-8", "iso-8859-1"); |
4411f3b6 |
727 | |
ab97ca19 |
728 | Note that because the conversion happens in place, the data to be |
729 | converted cannot be a string constant, it must be a scalar variable. |
730 | |
4411f3b6 |
731 | =back |
732 | |
733 | =head2 Handling Malformed Data |
734 | |
735 | If CHECK is not set, C<undef> is returned. If the data is supposed to |
47bfe92f |
736 | be UTF-8, an optional lexical warning (category utf8) is given. If |
737 | CHECK is true but not a code reference, dies. |
4411f3b6 |
738 | |
47bfe92f |
739 | It would desirable to have a way to indicate that transform should use |
740 | the encodings "replacement character" - no such mechanism is defined yet. |
4411f3b6 |
741 | |
742 | It is also planned to allow I<CHECK> to be a code reference. |
743 | |
47bfe92f |
744 | This is not yet implemented as there are design issues with what its |
745 | arguments should be and how it returns its results. |
4411f3b6 |
746 | |
747 | =over 4 |
748 | |
749 | =item Scheme 1 |
750 | |
751 | Passed remaining fragment of string being processed. |
752 | Modifies it in place to remove bytes/characters it can understand |
753 | and returns a string used to represent them. |
754 | e.g. |
755 | |
756 | sub fixup { |
757 | my $ch = substr($_[0],0,1,''); |
758 | return sprintf("\x{%02X}",ord($ch); |
759 | } |
760 | |
761 | This scheme is close to how underlying C code for Encode works, but gives |
762 | the fixup routine very little context. |
763 | |
764 | =item Scheme 2 |
765 | |
47bfe92f |
766 | Passed original string, and an index into it of the problem area, and |
767 | output string so far. Appends what it will to output string and |
768 | returns new index into original string. For example: |
4411f3b6 |
769 | |
770 | sub fixup { |
771 | # my ($s,$i,$d) = @_; |
772 | my $ch = substr($_[0],$_[1],1); |
773 | $_[2] .= sprintf("\x{%02X}",ord($ch); |
774 | return $_[1]+1; |
775 | } |
776 | |
47bfe92f |
777 | This scheme gives maximal control to the fixup routine but is more |
778 | complicated to code, and may need internals of Encode to be tweaked to |
779 | keep original string intact. |
4411f3b6 |
780 | |
781 | =item Other Schemes |
782 | |
783 | Hybrids of above. |
784 | |
785 | Multiple return values rather than in-place modifications. |
786 | |
787 | Index into the string could be pos($str) allowing s/\G...//. |
788 | |
789 | =back |
790 | |
791 | =head2 UTF-8 / utf8 |
792 | |
793 | The Unicode consortium defines the UTF-8 standard as a way of encoding |
47bfe92f |
794 | the entire Unicode repertiore as sequences of octets. This encoding is |
795 | expected to become very widespread. Perl can use this form internaly |
796 | to represent strings, so conversions to and from this form are |
797 | particularly efficient (as octets in memory do not have to change, |
798 | just the meta-data that tells Perl how to treat them). |
4411f3b6 |
799 | |
800 | =over 4 |
801 | |
802 | =item * |
803 | |
804 | $bytes = encode_utf8($string); |
805 | |
47bfe92f |
806 | The characters that comprise string are encoded in Perl's superset of UTF-8 |
4411f3b6 |
807 | and the resulting octets returned as a sequence of bytes. All possible |
808 | characters have a UTF-8 representation so this function cannot fail. |
809 | |
810 | =item * |
811 | |
812 | $string = decode_utf8($bytes [,CHECK]); |
813 | |
47bfe92f |
814 | The sequence of octets represented by $bytes is decoded from UTF-8 |
815 | into a sequence of logical characters. Not all sequences of octets |
816 | form valid UTF-8 encodings, so it is possible for this call to fail. |
817 | For CHECK see L</"Handling Malformed Data">. |
4411f3b6 |
818 | |
819 | =back |
820 | |
821 | =head2 Other Encodings of Unicode |
822 | |
47bfe92f |
823 | UTF-16 is similar to UCS-2, 16 bit or 2-byte chunks. UCS-2 can only |
7a4efbb2 |
824 | represent 0..0xFFFF, while UTF-16 has a I<surrogate pair> scheme which |
47bfe92f |
825 | allows it to cover the whole Unicode range. |
4411f3b6 |
826 | |
7a4efbb2 |
827 | Surrogates are code points set aside to encode the 0x01000..0x10FFFF |
828 | range of Unicode code points in pairs of 16-bit units. The I<high |
829 | surrogates> are the range 0xD800..0xDBFF, and the I<low surrogates> |
830 | are the range 0xDC00..0xDFFFF. The surrogate encoding is |
831 | |
832 | $hi = ($uni - 0x10000) / 0x400 + 0xD800; |
833 | $lo = ($uni - 0x10000) % 0x400 + 0xDC00; |
834 | |
835 | and the decoding is |
836 | |
837 | $uni = 0x10000 + ($hi - 0xD8000) * 0x400 + ($lo - 0xDC00); |
838 | |
8040349a |
839 | Encode implements big-endian UCS-2 aliased to "iso-10646-1" as that |
47bfe92f |
840 | happens to be the name used by that representation when used with X11 |
841 | fonts. |
4411f3b6 |
842 | |
843 | UTF-32 or UCS-4 is 32-bit or 4-byte chunks. Perl's logical characters |
844 | can be considered as being in this form without encoding. An encoding |
47bfe92f |
845 | to transfer strings in this form (e.g. to write them to a file) would |
846 | need to |
4411f3b6 |
847 | |
c079d275 |
848 | pack('L*', unpack('U*', $string)); # native |
4411f3b6 |
849 | or |
c079d275 |
850 | pack('V*', unpack('U*', $string)); # little-endian |
4411f3b6 |
851 | or |
c079d275 |
852 | pack('N*', unpack('U*', $string)); # big-endian |
4411f3b6 |
853 | |
c079d275 |
854 | depending on the endianness required. |
4411f3b6 |
855 | |
51ef4e11 |
856 | No UTF-32 encodings are implemented yet. |
4411f3b6 |
857 | |
47bfe92f |
858 | Both UCS-2 and UCS-4 style encodings can have "byte order marks" by |
859 | representing the code point 0xFFFE as the very first thing in a file. |
4411f3b6 |
860 | |
51ef4e11 |
861 | =head2 Listing available encodings |
862 | |
863 | use Encode qw(encodings); |
864 | @list = encodings(); |
865 | |
866 | Returns a list of the canonical names of the available encodings. |
867 | |
868 | =head2 Defining Aliases |
869 | |
870 | use Encode qw(define_alias); |
871 | define_alias( newName => ENCODING); |
872 | |
47bfe92f |
873 | Allows newName to be used as am alias for ENCODING. ENCODING may be |
874 | either the name of an encoding or and encoding object (as above). |
51ef4e11 |
875 | |
876 | Currently I<newName> can be specified in the following ways: |
877 | |
878 | =over 4 |
879 | |
880 | =item As a simple string. |
881 | |
882 | =item As a qr// compiled regular expression, e.g.: |
883 | |
884 | define_alias( qr/^iso8859-(\d+)$/i => '"iso-8859-$1"' ); |
885 | |
47bfe92f |
886 | In this case if I<ENCODING> is not a reference it is C<eval>-ed to |
887 | allow C<$1> etc. to be subsituted. The example is one way to names as |
888 | used in X11 font names to alias the MIME names for the iso-8859-* |
889 | family. |
51ef4e11 |
890 | |
891 | =item As a code reference, e.g.: |
892 | |
893 | define_alias( sub { return /^iso8859-(\d+)$/i ? "iso-8859-$1" : undef } , ''); |
894 | |
895 | In this case C<$_> will be set to the name that is being looked up and |
47bfe92f |
896 | I<ENCODING> is passed to the sub as its first argument. The example |
897 | is another way to names as used in X11 font names to alias the MIME |
898 | names for the iso-8859-* family. |
51ef4e11 |
899 | |
900 | =back |
901 | |
902 | =head2 Defining Encodings |
903 | |
e9692b5b |
904 | use Encode qw(define_alias); |
905 | define_encoding( $object, 'canonicalName' [,alias...]); |
51ef4e11 |
906 | |
47bfe92f |
907 | Causes I<canonicalName> to be associated with I<$object>. The object |
908 | should provide the interface described in L</"IMPLEMENTATION CLASSES"> |
909 | below. If more than two arguments are provided then additional |
910 | arguments are taken as aliases for I<$object> as for C<define_alias>. |
51ef4e11 |
911 | |
4411f3b6 |
912 | =head1 Encoding and IO |
913 | |
914 | It is very common to want to do encoding transformations when |
915 | reading or writing files, network connections, pipes etc. |
47bfe92f |
916 | If Perl is configured to use the new 'perlio' IO system then |
4411f3b6 |
917 | C<Encode> provides a "layer" (See L<perliol>) which can transform |
918 | data as it is read or written. |
919 | |
8e86646e |
920 | Here is how the blind poet would modernise the encoding: |
921 | |
42234700 |
922 | use Encode; |
8e86646e |
923 | open(my $iliad,'<:encoding(iso-8859-7)','iliad.greek'); |
924 | open(my $utf8,'>:utf8','iliad.utf8'); |
925 | my @epic = <$iliad>; |
926 | print $utf8 @epic; |
927 | close($utf8); |
928 | close($illiad); |
4411f3b6 |
929 | |
930 | In addition the new IO system can also be configured to read/write |
931 | UTF-8 encoded characters (as noted above this is efficient): |
932 | |
e9692b5b |
933 | open(my $fh,'>:utf8','anything'); |
934 | print $fh "Any \x{0021} string \N{SMILEY FACE}\n"; |
4411f3b6 |
935 | |
936 | Either of the above forms of "layer" specifications can be made the default |
937 | for a lexical scope with the C<use open ...> pragma. See L<open>. |
938 | |
939 | Once a handle is open is layers can be altered using C<binmode>. |
940 | |
47bfe92f |
941 | Without any such configuration, or if Perl itself is built using |
4411f3b6 |
942 | system's own IO, then write operations assume that file handle accepts |
943 | only I<bytes> and will C<die> if a character larger than 255 is |
944 | written to the handle. When reading, each octet from the handle |
945 | becomes a byte-in-a-character. Note that this default is the same |
47bfe92f |
946 | behaviour as bytes-only languages (including Perl before v5.6) would |
947 | have, and is sufficient to handle native 8-bit encodings |
948 | e.g. iso-8859-1, EBCDIC etc. and any legacy mechanisms for handling |
949 | other encodings and binary data. |
950 | |
951 | In other cases it is the programs responsibility to transform |
952 | characters into bytes using the API above before doing writes, and to |
953 | transform the bytes read from a handle into characters before doing |
954 | "character operations" (e.g. C<lc>, C</\W+/>, ...). |
955 | |
47bfe92f |
956 | You can also use PerlIO to convert larger amounts of data you don't |
957 | want to bring into memory. For example to convert between ISO 8859-1 |
958 | (Latin 1) and UTF-8 (or UTF-EBCDIC in EBCDIC machines): |
959 | |
e9692b5b |
960 | open(F, "<:encoding(iso-8859-1)", "data.txt") or die $!; |
961 | open(G, ">:utf8", "data.utf") or die $!; |
962 | while (<F>) { print G } |
963 | |
964 | # Could also do "print G <F>" but that would pull |
965 | # the whole file into memory just to write it out again. |
966 | |
967 | More examples: |
47bfe92f |
968 | |
e9692b5b |
969 | open(my $f, "<:encoding(cp1252)") |
970 | open(my $g, ">:encoding(iso-8859-2)") |
971 | open(my $h, ">:encoding(latin9)") # iso-8859-15 |
47bfe92f |
972 | |
973 | See L<PerlIO> for more information. |
4411f3b6 |
974 | |
1768d7eb |
975 | See also L<encoding> for how to change the default encoding of the |
d521382b |
976 | data in your script. |
1768d7eb |
977 | |
4411f3b6 |
978 | =head1 Encoding How to ... |
979 | |
980 | To do: |
981 | |
982 | =over 4 |
983 | |
984 | =item * IO with mixed content (faking iso-2020-*) |
985 | |
986 | =item * MIME's Content-Length: |
987 | |
988 | =item * UTF-8 strings in binary data. |
989 | |
47bfe92f |
990 | =item * Perl/Encode wrappers on non-Unicode XS modules. |
4411f3b6 |
991 | |
992 | =back |
993 | |
994 | =head1 Messing with Perl's Internals |
995 | |
47bfe92f |
996 | The following API uses parts of Perl's internals in the current |
997 | implementation. As such they are efficient, but may change. |
4411f3b6 |
998 | |
999 | =over 4 |
1000 | |
4411f3b6 |
1001 | =item * is_utf8(STRING [, CHECK]) |
1002 | |
1003 | [INTERNAL] Test whether the UTF-8 flag is turned on in the STRING. |
47bfe92f |
1004 | If CHECK is true, also checks the data in STRING for being well-formed |
1005 | UTF-8. Returns true if successful, false otherwise. |
4411f3b6 |
1006 | |
1007 | =item * valid_utf8(STRING) |
1008 | |
47bfe92f |
1009 | [INTERNAL] Test whether STRING is in a consistent state. Will return |
1010 | true if string is held as bytes, or is well-formed UTF-8 and has the |
1011 | UTF-8 flag on. Main reason for this routine is to allow Perl's |
1012 | testsuite to check that operations have left strings in a consistent |
1013 | state. |
4411f3b6 |
1014 | |
1015 | =item * |
1016 | |
1017 | _utf8_on(STRING) |
1018 | |
1019 | [INTERNAL] Turn on the UTF-8 flag in STRING. The data in STRING is |
1020 | B<not> checked for being well-formed UTF-8. Do not use unless you |
1021 | B<know> that the STRING is well-formed UTF-8. Returns the previous |
1022 | state of the UTF-8 flag (so please don't test the return value as |
1023 | I<not> success or failure), or C<undef> if STRING is not a string. |
1024 | |
1025 | =item * |
1026 | |
1027 | _utf8_off(STRING) |
1028 | |
1029 | [INTERNAL] Turn off the UTF-8 flag in STRING. Do not use frivolously. |
1030 | Returns the previous state of the UTF-8 flag (so please don't test the |
1031 | return value as I<not> success or failure), or C<undef> if STRING is |
1032 | not a string. |
1033 | |
1034 | =back |
1035 | |
4edaa979 |
1036 | =head1 IMPLEMENTATION CLASSES |
1037 | |
1038 | As mentioned above encodings are (in the current implementation at least) |
1039 | defined by objects. The mapping of encoding name to object is via the |
51ef4e11 |
1040 | C<%encodings> hash. |
4edaa979 |
1041 | |
1042 | The values of the hash can currently be either strings or objects. |
1043 | The string form may go away in the future. The string form occurs |
1044 | when C<encodings()> has scanned C<@INC> for loadable encodings but has |
1045 | not actually loaded the encoding in question. This is because the |
47bfe92f |
1046 | current "loading" process is all Perl and a bit slow. |
4edaa979 |
1047 | |
47bfe92f |
1048 | Once an encoding is loaded then value of the hash is object which |
1049 | implements the encoding. The object should provide the following |
1050 | interface: |
4edaa979 |
1051 | |
1052 | =over 4 |
1053 | |
1054 | =item -E<gt>name |
1055 | |
1056 | Should return the string representing the canonical name of the encoding. |
1057 | |
1058 | =item -E<gt>new_sequence |
1059 | |
47bfe92f |
1060 | This is a placeholder for encodings with state. It should return an |
1061 | object which implements this interface, all current implementations |
1062 | return the original object. |
4edaa979 |
1063 | |
1064 | =item -E<gt>encode($string,$check) |
1065 | |
47bfe92f |
1066 | Should return the octet sequence representing I<$string>. If I<$check> |
1067 | is true it should modify I<$string> in place to remove the converted |
1068 | part (i.e. the whole string unless there is an error). If an error |
1069 | occurs it should return the octet sequence for the fragment of string |
1070 | that has been converted, and modify $string in-place to remove the |
1071 | converted part leaving it starting with the problem fragment. |
4edaa979 |
1072 | |
47bfe92f |
1073 | If check is is false then C<encode> should make a "best effort" to |
1074 | convert the string - for example by using a replacement character. |
4edaa979 |
1075 | |
1076 | =item -E<gt>decode($octets,$check) |
1077 | |
47bfe92f |
1078 | Should return the string that I<$octets> represents. If I<$check> is |
1079 | true it should modify I<$octets> in place to remove the converted part |
1080 | (i.e. the whole sequence unless there is an error). If an error |
1081 | occurs it should return the fragment of string that has been |
1082 | converted, and modify $octets in-place to remove the converted part |
4edaa979 |
1083 | leaving it starting with the problem fragment. |
1084 | |
47bfe92f |
1085 | If check is is false then C<decode> should make a "best effort" to |
1086 | convert the string - for example by using Unicode's "\x{FFFD}" as a |
1087 | replacement character. |
4edaa979 |
1088 | |
1089 | =back |
1090 | |
47bfe92f |
1091 | It should be noted that the check behaviour is different from the |
1092 | outer public API. The logic is that the "unchecked" case is useful |
1093 | when encoding is part of a stream which may be reporting errors |
1094 | (e.g. STDERR). In such cases it is desirable to get everything |
1095 | through somehow without causing additional errors which obscure the |
1096 | original one. Also the encoding is best placed to know what the |
1097 | correct replacement character is, so if that is the desired behaviour |
1098 | then letting low level code do it is the most efficient. |
1099 | |
1100 | In contrast if check is true, the scheme above allows the encoding to |
1101 | do as much as it can and tell layer above how much that was. What is |
1102 | lacking at present is a mechanism to report what went wrong. The most |
1103 | likely interface will be an additional method call to the object, or |
1104 | perhaps (to avoid forcing per-stream objects on otherwise stateless |
1105 | encodings) and additional parameter. |
1106 | |
1107 | It is also highly desirable that encoding classes inherit from |
1108 | C<Encode::Encoding> as a base class. This allows that class to define |
1109 | additional behaviour for all encoding objects. For example built in |
1110 | Unicode, UCS-2 and UTF-8 classes use : |
51ef4e11 |
1111 | |
1112 | package Encode::MyEncoding; |
1113 | use base qw(Encode::Encoding); |
1114 | |
1115 | __PACKAGE__->Define(qw(myCanonical myAlias)); |
1116 | |
47bfe92f |
1117 | To create an object with bless {Name => ...},$class, and call |
1118 | define_encoding. They inherit their C<name> method from |
1119 | C<Encode::Encoding>. |
4edaa979 |
1120 | |
1121 | =head2 Compiled Encodings |
1122 | |
47bfe92f |
1123 | F<Encode.xs> provides a class C<Encode::XS> which provides the |
1124 | interface described above. It calls a generic octet-sequence to |
1125 | octet-sequence "engine" that is driven by tables (defined in |
1126 | F<encengine.c>). The same engine is used for both encode and |
1127 | decode. C<Encode:XS>'s C<encode> forces Perl's characters to their |
1128 | UTF-8 form and then treats them as just another multibyte |
1129 | encoding. C<Encode:XS>'s C<decode> transforms the sequence and then |
1130 | turns the UTF-8-ness flag as that is the form that the tables are |
1131 | defined to produce. For details of the engine see the comments in |
1132 | F<encengine.c>. |
1133 | |
1134 | The tables are produced by the Perl script F<compile> (the name needs |
1135 | to change so we can eventually install it somewhere). F<compile> can |
1136 | currently read two formats: |
4edaa979 |
1137 | |
1138 | =over 4 |
1139 | |
1140 | =item *.enc |
1141 | |
47bfe92f |
1142 | This is a coined format used by Tcl. It is documented in |
1143 | Encode/EncodeFormat.pod. |
4edaa979 |
1144 | |
1145 | =item *.ucm |
1146 | |
1147 | This is the semi-standard format used by IBM's ICU package. |
1148 | |
1149 | =back |
1150 | |
1151 | F<compile> can write the following forms: |
1152 | |
1153 | =over 4 |
1154 | |
1155 | =item *.ucm |
1156 | |
1157 | See above - the F<Encode/*.ucm> files provided with the distribution have |
1158 | been created from the original Tcl .enc files using this approach. |
1159 | |
1160 | =item *.c |
1161 | |
1162 | Produces tables as C data structures - this is used to build in encodings |
1163 | into F<Encode.so>/F<Encode.dll>. |
1164 | |
1165 | =item *.xs |
1166 | |
47bfe92f |
1167 | In theory this allows encodings to be stand-alone loadable Perl |
1168 | extensions. The process has not yet been tested. The plan is to use |
1169 | this approach for large East Asian encodings. |
4edaa979 |
1170 | |
1171 | =back |
1172 | |
47bfe92f |
1173 | The set of encodings built-in to F<Encode.so>/F<Encode.dll> is |
1174 | determined by F<Makefile.PL>. The current set is as follows: |
4edaa979 |
1175 | |
1176 | =over 4 |
1177 | |
1178 | =item ascii and iso-8859-* |
1179 | |
1180 | That is all the common 8-bit "western" encodings. |
1181 | |
1182 | =item IBM-1047 and two other variants of EBCDIC. |
1183 | |
47bfe92f |
1184 | These are the same variants that are supported by EBCDIC Perl as |
1185 | "native" encodings. They are included to prove "reversibility" of |
1186 | some constructs in EBCDIC Perl. |
4edaa979 |
1187 | |
1188 | =item symbol and dingbats as used by Tk on X11. |
1189 | |
47bfe92f |
1190 | (The reason Encode got started was to support Perl/Tk.) |
4edaa979 |
1191 | |
1192 | =back |
1193 | |
47bfe92f |
1194 | That set is rather ad hoc and has been driven by the needs of the |
1195 | tests rather than the needs of typical applications. It is likely |
1196 | to be rationalized. |
4edaa979 |
1197 | |
4411f3b6 |
1198 | =head1 SEE ALSO |
1199 | |
1768d7eb |
1200 | L<perlunicode>, L<perlebcdic>, L<perlfunc/open>, L<PerlIO>, L<encoding> |
4411f3b6 |
1201 | |
1202 | =cut |
1203 | |