Debian Codegraph @

upstream/debian-unstable/man3/encoding.3perl

raw
1 .\" -*- mode: troff; coding: utf-8 -*-
2 .\" Automatically generated by Pod::Man 5.0102 (Pod::Simple 3.45)
3 .\"
4 .\" Standard preamble:
5 .\" ========================================================================
6 .de Sp \" Vertical space (when we can't use .PP)
7 .if t .sp .5v
8 .if n .sp
9 ..
10 .de Vb \" Begin verbatim text
11 .ft CW
12 .nf
13 .ne \\$1
14 ..
15 .de Ve \" End verbatim text
16 .ft R
17 .fi
18 ..
19 .\" \*(C` and \*(C' are quotes in nroff, nothing in troff, for use with C<>.
20 .ie n \{\
21 . ds C` ""
22 . ds C' ""
23 'br\}
24 .el\{\
25 . ds C`
26 . ds C'
27 'br\}
28 .\"
29 .\" Escape single quotes in literal strings from groff's Unicode transform.
30 .ie \n(.g .ds Aq \(aq
31 .el .ds Aq '
32 .\"
33 .\" If the F register is >0, we'll generate index entries on stderr for
34 .\" titles (.TH), headers (.SH), subsections (.SS), items (.Ip), and index
35 .\" entries marked with X<> in POD. Of course, you'll have to process the
36 .\" output yourself in some meaningful fashion.
37 .\"
38 .\" Avoid warning from groff about undefined register 'F'.
39 .de IX
40 ..
41 .nr rF 0
42 .if \n(.g .if rF .nr rF 1
43 .if (\n(rF:(\n(.g==0)) \{\
44 . if \nF \{\
45 . de IX
46 . tm Index:\\$1\t\\n%\t"\\$2"
47 ..
48 . if !\nF==2 \{\
49 . nr % 0
50 . nr F 2
51 . \}
52 . \}
53 .\}
54 .rr rF
55 .\" ========================================================================
56 .\"
57 .IX Title "encoding 3perl"
58 .TH encoding 3perl 2025-11-16 "perl v5.40.1" "Perl Programmers Reference Guide"
59 .\" For nroff, turn off justification. Always turn off hyphenation; it makes
60 .\" way too many mistakes in technical documents.
61 .if n .ad l
62 .nh
63 .SH NAME
64 encoding \- allows you to write your script in non\-ASCII and non\-UTF\-8
65 .SH WARNING
66 .IX Header "WARNING"
67 This module has been deprecated since perl v5.18. See "DESCRIPTION" and
68 "BUGS".
69 .SH SYNOPSIS
70 .IX Header "SYNOPSIS"
71 .Vb 2
72 \& use encoding "greek"; # Perl like Greek to you?
73 \& use encoding "euc\-jp"; # Jperl!
74 \&
75 \& # or you can even do this if your shell supports your native encoding
76 \&
77 \& perl \-Mencoding=latin2 \-e\*(Aq...\*(Aq # Feeling centrally European?
78 \& perl \-Mencoding=euc\-kr \-e\*(Aq...\*(Aq # Or Korean?
79 \&
80 \& # more control
81 \&
82 \& # A simple euc\-cn => utf\-8 converter
83 \& use encoding "euc\-cn", STDOUT => "utf8"; while(<>){print};
84 \&
85 \& # "no encoding;" supported
86 \& no encoding;
87 \&
88 \& # an alternate way, Filter
89 \& use encoding "euc\-jp", Filter=>1;
90 \& # now you can use kanji identifiers \-\- in euc\-jp!
91 \&
92 \& # encode based on the current locale \- specialized purposes only;
93 \& # fraught with danger!!
94 \& use encoding \*(Aq:locale\*(Aq;
95 .Ve
96 .SH DESCRIPTION
97 .IX Header "DESCRIPTION"
98 This pragma is used to enable a Perl script to be written in encodings that
99 aren't strictly ASCII nor UTF\-8. It translates all or portions of the Perl
100 program script from a given encoding into UTF\-8, and changes the PerlIO layers
101 of \f(CW\*(C`STDIN\*(C'\fR and \f(CW\*(C`STDOUT\*(C'\fR to the encoding specified.
102 .PP
103 This pragma dates from the days when UTF\-8\-enabled editors were uncommon. But
104 that was long ago, and the need for it is greatly diminished. That, coupled
105 with the fact that it doesn't work with threads, along with other problems,
106 (see "BUGS") have led to its being deprecated. It is planned to remove this
107 pragma in a future Perl version. New code should be written in UTF\-8, and the
108 \&\f(CW\*(C`use utf8\*(C'\fR pragma used instead (see perluniintro and utf8 for details).
109 Old code should be converted to UTF\-8, via something like the recipe in the
110 "SYNOPSIS" (though this simple approach may require manual adjustments
111 afterwards).
112 .PP
113 If UTF\-8 is not an option, it is recommended that one use a simple source
114 filter, such as that provided by Filter::Encoding on CPAN or this
115 pragma's own \f(CW\*(C`Filter\*(C'\fR option (see below).
116 .PP
117 The only legitimate use of this pragma is almost certainly just one per file,
118 near the top, with file scope, as the file is likely going to only be written
119 in one encoding. Further restrictions apply in Perls before v5.22 (see
120 "Prior to Perl v5.22").
121 .PP
122 There are two basic modes of operation (plus turning if off):
123 .ie n .IP """use encoding [\*(Aq\fIENCNAME\fR\*(Aq] ;""" 4
124 .el .IP "\f(CWuse encoding [\*(Aq\fR\f(CIENCNAME\fR\f(CW\*(Aq] ;\fR" 4
125 .IX Item "use encoding [ENCNAME] ;"
126 Please note: This mode of operation is no longer supported as of Perl
127 v5.26.
128 .Sp
129 This is the normal operation. It translates various literals encountered in
130 the Perl source file from the encoding \fIENCNAME\fR into UTF\-8, and similarly
131 converts character code points. This is used when the script is a combination
132 of ASCII (for the variable names and punctuation, \fIetc\fR), but the literal
133 data is in the specified encoding.
134 .Sp
135 \&\fIENCNAME\fR is optional. If omitted, the encoding specified in the environment
136 variable \f(CW\*(C`PERL_ENCODING\*(C'\fR is used. If this isn't
137 set, or the resolved-to encoding is not known to \f(CW\*(C`Encode\*(C'\fR, the error
138 \&\f(CW\*(C`Unknown encoding \*(Aq\fR\f(CIENCNAME\fR\f(CW\*(Aq\*(C'\fR will be thrown.
139 .Sp
140 Starting in Perl v5.8.6 (\f(CW\*(C`Encode\*(C'\fR version 2.0.1), \fIENCNAME\fR may be the
141 name \f(CW\*(C`:locale\*(C'\fR. This is for very specialized applications, and is documented
142 in "The \f(CW\*(C`:locale\*(C'\fR sub-pragma" below.
143 .Sp
144 The literals that are converted are \f(CW\*(C`q//, qq//, qr//, qw///, qx//\*(C'\fR, and
145 starting in v5.8.1, \f(CW\*(C`tr///\*(C'\fR. Operations that do conversions include \f(CW\*(C`chr\*(C'\fR,
146 \&\f(CW\*(C`ord\*(C'\fR, \f(CW\*(C`utf8::upgrade\*(C'\fR (but not \f(CW\*(C`utf8::downgrade\*(C'\fR), and \f(CW\*(C`chomp\*(C'\fR.
147 .Sp
148 Also starting in v5.8.1, the \f(CW\*(C`DATA\*(C'\fR pseudo-filehandle is translated from the
149 encoding into UTF\-8.
150 .Sp
151 For example, you can write code in EUC-JP as follows:
152 .Sp
153 .Vb 3
154 \& my $Rakuda = "\exF1\exD1\exF1\exCC"; # Camel in Kanji
155 \& #<\-char\-><\-char\-> # 4 octets
156 \& s/\ebCamel\eb/$Rakuda/;
157 .Ve
158 .Sp
159 And with \f(CW\*(C`use encoding "euc\-jp"\*(C'\fR in effect, it is the same thing as
160 that code in UTF\-8:
161 .Sp
162 .Vb 2
163 \& my $Rakuda = "\ex{99F1}\ex{99DD}"; # two Unicode Characters
164 \& s/\ebCamel\eb/$Rakuda/;
165 .Ve
166 .Sp
167 See "EXAMPLE" below for a more complete example.
168 .Sp
169 Unless \f(CW\*(C`${^UNICODE}\*(C'\fR (available starting in v5.8.2) exists and is non-zero, the
170 PerlIO layers of \f(CW\*(C`STDIN\*(C'\fR and \f(CW\*(C`STDOUT\*(C'\fR are set to "\f(CW:encoding(\fR\f(CIENCNAME\fR\f(CW)\fR".
171 Therefore,
172 .Sp
173 .Vb 5
174 \& use encoding "euc\-jp";
175 \& my $message = "Camel is the symbol of perl.\en";
176 \& my $Rakuda = "\exF1\exD1\exF1\exCC"; # Camel in Kanji
177 \& $message =~ s/\ebCamel\eb/$Rakuda/;
178 \& print $message;
179 .Ve
180 .Sp
181 will print
182 .Sp
183 .Vb 1
184 \& "\exF1\exD1\exF1\exCC is the symbol of perl.\en"
185 .Ve
186 .Sp
187 not
188 .Sp
189 .Vb 1
190 \& "\ex{99F1}\ex{99DD} is the symbol of perl.\en"
191 .Ve
192 .Sp
193 You can override this by giving extra arguments; see below.
194 .Sp
195 Note that \f(CW\*(C`STDERR\*(C'\fR WILL NOT be changed, regardless.
196 .Sp
197 Also note that non-STD file handles remain unaffected. Use \f(CW\*(C`use
198 open\*(C'\fR or \f(CW\*(C`binmode\*(C'\fR to change the layers of those.
199 .ie n .IP """use encoding \fIENCNAME\fR, Filter=>1;""" 4
200 .el .IP "\f(CWuse encoding \fR\f(CIENCNAME\fR\f(CW, Filter=>1;\fR" 4
201 .IX Item "use encoding ENCNAME, Filter=>1;"
202 This operates as above, but the \f(CW\*(C`Filter\*(C'\fR argument with a non-zero
203 value causes the entire script, and not just literals, to be translated from
204 the encoding into UTF\-8. This allows identifiers in the source to be in that
205 encoding as well. (Problems may occur if the encoding is not a superset of
206 ASCII; imagine all your semi-colons being translated into something
207 different.) One can use this form to make
208 .Sp
209 .Vb 1
210 \& ${"\ex{4eba}"}++
211 .Ve
212 .Sp
213 work. (This is equivalent to \f(CW\*(C`$\fR\f(CIhuman\fR\f(CW++\*(C'\fR, where \fIhuman\fR is a single Han
214 ideograph).
215 .Sp
216 This effectively means that your source code behaves as if it were written in
217 UTF\-8 with \f(CW\*(C`\*(Aquse utf8\*(C'\fR' in effect. So even if your editor only supports
218 Shift_JIS, for example, you can still try examples in Chapter 15 of
219 \&\f(CW\*(C`Programming Perl, 3rd Ed.\*(C'\fR.
220 .Sp
221 This option is significantly slower than the other one.
222 .ie n .IP """no encoding;""" 4
223 .el .IP "\f(CWno encoding;\fR" 4
224 .IX Item "no encoding;"
225 Unsets the script encoding. The layers of \f(CW\*(C`STDIN\*(C'\fR, \f(CW\*(C`STDOUT\*(C'\fR are
226 reset to "\f(CW\*(C`:raw\*(C'\fR" (the default unprocessed raw stream of bytes).
227 .SH OPTIONS
228 .IX Header "OPTIONS"
229 .ie n .SS "Setting ""STDIN"" and/or ""STDOUT"" individually"
230 .el .SS "Setting \f(CWSTDIN\fP and/or \f(CWSTDOUT\fP individually"
231 .IX Subsection "Setting STDIN and/or STDOUT individually"
232 The encodings of \f(CW\*(C`STDIN\*(C'\fR and \f(CW\*(C`STDOUT\*(C'\fR are individually settable by parameters to
233 the pragma:
234 .PP
235 .Vb 1
236 \& use encoding \*(Aqeuc\-tw\*(Aq, STDIN => \*(Aqgreek\*(Aq ...;
237 .Ve
238 .PP
239 In this case, you cannot omit the first \fIENCNAME\fR. \f(CW\*(C`STDIN => undef\*(C'\fR
240 turns the I/O transcoding completely off for that filehandle.
241 .PP
242 When \f(CW\*(C`${^UNICODE}\*(C'\fR (available starting in v5.8.2) exists and is non-zero,
243 these options will be completely ignored. See "\f(CW\*(C`${^UNICODE}\*(C'\fR" in perlvar and
244 "\f(CW\*(C`\-C\*(C'\fR" in perlrun for details.
245 .ie n .SS "The "":locale"" sub-pragma"
246 .el .SS "The \f(CW:locale\fP sub-pragma"
247 .IX Subsection "The :locale sub-pragma"
248 Starting in v5.8.6, the encoding name may be \f(CW\*(C`:locale\*(C'\fR. This means that the
249 encoding is taken from the current locale, and not hard-coded by the pragma.
250 Since a script really can only be encoded in exactly one encoding, this option
251 is dangerous. It makes sense only if the script itself is written in ASCII,
252 and all the possible locales that will be in use when the script is executed
253 are supersets of ASCII. That means that the script itself doesn't get
254 changed, but the I/O handles have the specified encoding added, and the
255 operations like \f(CW\*(C`chr\*(C'\fR and \f(CW\*(C`ord\*(C'\fR use that encoding.
256 .PP
257 The logic of finding which locale \f(CW\*(C`:locale\*(C'\fR uses is as follows:
258 .IP 1. 4
259 If the platform supports the \f(CWlanginfo(CODESET)\fR interface, the codeset
260 returned is used as the default encoding for the open pragma.
261 .IP 2. 4
262 If 1. didn't work but we are under the locale pragma, the environment
263 variables \f(CW\*(C`LC_ALL\*(C'\fR and \f(CW\*(C`LANG\*(C'\fR (in that order) are matched for encodings
264 (the part after "\f(CW\*(C`.\*(C'\fR", if any), and if any found, that is used
265 as the default encoding for the open pragma.
266 .IP 3. 4
267 If 1. and 2. didn't work, the environment variables \f(CW\*(C`LC_ALL\*(C'\fR and \f(CW\*(C`LANG\*(C'\fR
268 (in that order) are matched for anything looking like UTF\-8, and if
269 any found, \f(CW\*(C`:utf8\*(C'\fR is used as the default encoding for the open
270 pragma.
271 .PP
272 If your locale environment variables (\f(CW\*(C`LC_ALL\*(C'\fR, \f(CW\*(C`LC_CTYPE\*(C'\fR, \f(CW\*(C`LANG\*(C'\fR)
273 contain the strings 'UTF\-8' or 'UTF8' (case-insensitive matching),
274 the default encoding of your \f(CW\*(C`STDIN\*(C'\fR, \f(CW\*(C`STDOUT\*(C'\fR, and \f(CW\*(C`STDERR\*(C'\fR, and of
275 \&\fBany subsequent file open\fR, is UTF\-8.
276 .SH CAVEATS
277 .IX Header "CAVEATS"
278 .SS "SIDE EFFECTS"
279 .IX Subsection "SIDE EFFECTS"
280 .IP \(bu 4
281 If the \f(CW\*(C`encoding\*(C'\fR pragma is in scope then the lengths returned are
282 calculated from the length of \f(CW$/\fR in Unicode characters, which is not
283 always the same as the length of \f(CW$/\fR in the native encoding.
284 .IP \(bu 4
285 Without this pragma, if strings operating under byte semantics and strings
286 with Unicode character data are concatenated, the new string will
287 be created by decoding the byte strings as \fIISO 8859\-1 (Latin\-1)\fR.
288 .Sp
289 The \fBencoding\fR pragma changes this to use the specified encoding
290 instead. For example:
291 .Sp
292 .Vb 5
293 \& use encoding \*(Aqutf8\*(Aq;
294 \& my $string = chr(20000); # a Unicode string
295 \& utf8::encode($string); # now it\*(Aqs a UTF\-8 encoded byte string
296 \& # concatenate with another Unicode string
297 \& print length($string . chr(20000));
298 .Ve
299 .Sp
300 Will print \f(CW2\fR, because \f(CW$string\fR is upgraded as UTF\-8. Without
301 \&\f(CW\*(C`use encoding \*(Aqutf8\*(Aq;\*(C'\fR, it will print \f(CW4\fR instead, since \f(CW$string\fR
302 is three octets when interpreted as Latin\-1.
303 .SS "DO NOT MIX MULTIPLE ENCODINGS"
304 .IX Subsection "DO NOT MIX MULTIPLE ENCODINGS"
305 Notice that only literals (string or regular expression) having only
306 legacy code points are affected: if you mix data like this
307 .PP
308 .Vb 2
309 \& \ex{100}\exDF
310 \& \exDF\ex{100}
311 .Ve
312 .PP
313 the data is assumed to be in (Latin 1 and) Unicode, not in your native
314 encoding. In other words, this will match in "greek":
315 .PP
316 .Vb 1
317 \& "\exDF" =~ /\ex{3af}/
318 .Ve
319 .PP
320 but this will not
321 .PP
322 .Vb 1
323 \& "\exDF\ex{100}" =~ /\ex{3af}\ex{100}/
324 .Ve
325 .PP
326 since the \f(CW\*(C`\exDF\*(C'\fR (ISO 8859\-7 GREEK SMALL LETTER IOTA WITH TONOS) on
327 the left will \fBnot\fR be upgraded to \f(CW\*(C`\ex{3af}\*(C'\fR (Unicode GREEK SMALL
328 LETTER IOTA WITH TONOS) because of the \f(CW\*(C`\ex{100}\*(C'\fR on the left. You
329 should not be mixing your legacy data and Unicode in the same string.
330 .PP
331 This pragma also affects encoding of the 0x80..0xFF code point range:
332 normally characters in that range are left as eight-bit bytes (unless
333 they are combined with characters with code points 0x100 or larger,
334 in which case all characters need to become UTF\-8 encoded), but if
335 the \f(CW\*(C`encoding\*(C'\fR pragma is present, even the 0x80..0xFF range always
336 gets UTF\-8 encoded.
337 .PP
338 After all, the best thing about this pragma is that you don't have to
339 resort to \ex{....} just to spell your name in a native encoding.
340 So feel free to put your strings in your encoding in quotes and
341 regexes.
342 .SS "Prior to Perl v5.22"
343 .IX Subsection "Prior to Perl v5.22"
344 The pragma was a per script, not a per block lexical. Only the last
345 \&\f(CW\*(C`use encoding\*(C'\fR or \f(CW\*(C`no encoding\*(C'\fR mattered, and it affected
346 \&\fBthe whole script\fR. However, the \f(CW\*(C`no encoding\*(C'\fR pragma was supported and
347 \&\f(CW\*(C`use encoding\*(C'\fR could appear as many times as you want in a given script
348 (though only the last was effective).
349 .PP
350 Since the scope wasn't lexical, other modules' use of \f(CW\*(C`chr\*(C'\fR, \f(CW\*(C`ord\*(C'\fR, \fIetc.\fR
351 were affected. This leads to spooky, incorrect action at a distance that is
352 hard to debug.
353 .PP
354 This means you would have to be very careful of the load order:
355 .PP
356 .Vb 5
357 \& # called module
358 \& package Module_IN_BAR;
359 \& use encoding "bar";
360 \& # stuff in "bar" encoding here
361 \& 1;
362 \&
363 \& # caller script
364 \& use encoding "foo"
365 \& use Module_IN_BAR;
366 \& # surprise! use encoding "bar" is in effect.
367 .Ve
368 .PP
369 The best way to avoid this oddity is to use this pragma RIGHT AFTER
370 other modules are loaded. i.e.
371 .PP
372 .Vb 2
373 \& use Module_IN_BAR;
374 \& use encoding "foo";
375 .Ve
376 .SS "Prior to Encode version 1.87"
377 .IX Subsection "Prior to Encode version 1.87"
378 .IP \(bu 4
379 \&\f(CW\*(C`STDIN\*(C'\fR and \f(CW\*(C`STDOUT\*(C'\fR were not set under the filter option.
380 And \f(CW\*(C`STDIN=>\fR\f(CIENCODING\fR\f(CW\*(C'\fR and \f(CW\*(C`STDOUT=>\fR\f(CIENCODING\fR\f(CW\*(C'\fR didn't work like
381 non-filter version.
382 .IP \(bu 4
383 \&\f(CW\*(C`use utf8\*(C'\fR wasn't implicitly declared so you have to \f(CW\*(C`use utf8\*(C'\fR to do
384 .Sp
385 .Vb 1
386 \& ${"\ex{4eba}"}++
387 .Ve
388 .SS "Prior to Perl v5.8.1"
389 .IX Subsection "Prior to Perl v5.8.1"
390 .IP """NON-EUC"" doublebyte encodings" 4
391 .IX Item """NON-EUC"" doublebyte encodings"
392 Because perl needs to parse the script before applying this pragma, such
393 encodings as Shift_JIS and Big\-5 that may contain \f(CW\*(Aq\e\*(Aq\fR (BACKSLASH;
394 \&\f(CW\*(C`\ex5c\*(C'\fR) in the second byte fail because the second byte may
395 accidentally escape the quoting character that follows.
396 .ie n .IP """tr///""" 4
397 .el .IP \f(CWtr///\fR 4
398 .IX Item "tr///"
399 The \fBencoding\fR pragma works by decoding string literals in
400 \&\f(CW\*(C`q//,qq//,qr//,qw///, qx//\*(C'\fR and so forth. In perl v5.8.0, this
401 does not apply to \f(CW\*(C`tr///\*(C'\fR. Therefore,
402 .Sp
403 .Vb 4
404 \& use encoding \*(Aqeuc\-jp\*(Aq;
405 \& #....
406 \& $kana =~ tr/\exA4\exA1\-\exA4\exF3/\exA5\exA1\-\exA5\exF3/;
407 \& # \-\-\-\-\-\-\-\- \-\-\-\-\-\-\-\- \-\-\-\-\-\-\-\- \-\-\-\-\-\-\-\-
408 .Ve
409 .Sp
410 Does not work as
411 .Sp
412 .Vb 1
413 \& $kana =~ tr/\ex{3041}\-\ex{3093}/\ex{30a1}\-\ex{30f3}/;
414 .Ve
415 .RS 4
416 .IP "Legend of characters above" 4
417 .IX Item "Legend of characters above"
418 .Vb 6
419 \& utf8 euc\-jp charnames::viacode()
420 \& \-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-\-
421 \& \ex{3041} \exA4\exA1 HIRAGANA LETTER SMALL A
422 \& \ex{3093} \exA4\exF3 HIRAGANA LETTER N
423 \& \ex{30a1} \exA5\exA1 KATAKANA LETTER SMALL A
424 \& \ex{30f3} \exA5\exF3 KATAKANA LETTER N
425 .Ve
426 .RE
427 .RS 4
428 .Sp
429 This counterintuitive behavior has been fixed in perl v5.8.1.
430 .Sp
431 In perl v5.8.0, you can work around this as follows;
432 .Sp
433 .Vb 3
434 \& use encoding \*(Aqeuc\-jp\*(Aq;
435 \& # ....
436 \& eval qq{ \e$kana =~ tr/\exA4\exA1\-\exA4\exF3/\exA5\exA1\-\exA5\exF3/ };
437 .Ve
438 .Sp
439 Note the \f(CW\*(C`tr//\*(C'\fR expression is surrounded by \f(CW\*(C`qq{}\*(C'\fR. The idea behind
440 this is the same as the classic idiom that makes \f(CW\*(C`tr///\*(C'\fR 'interpolate':
441 .Sp
442 .Vb 2
443 \& tr/$from/$to/; # wrong!
444 \& eval qq{ tr/$from/$to/ }; # workaround.
445 .Ve
446 .RE
447 .SH "EXAMPLE \- Greekperl"
448 .IX Header "EXAMPLE - Greekperl"
449 .Vb 1
450 \& use encoding "iso 8859\-7";
451 \&
452 \& # \exDF in ISO 8859\-7 (Greek) is \ex{3af} in Unicode.
453 \&
454 \& $a = "\exDF";
455 \& $b = "\ex{100}";
456 \&
457 \& printf "%#x\en", ord($a); # will print 0x3af, not 0xdf
458 \&
459 \& $c = $a . $b;
460 \&
461 \& # $c will be "\ex{3af}\ex{100}", not "\ex{df}\ex{100}".
462 \&
463 \& # chr() is affected, and ...
464 \&
465 \& print "mega\en" if ord(chr(0xdf)) == 0x3af;
466 \&
467 \& # ... ord() is affected by the encoding pragma ...
468 \&
469 \& print "tera\en" if ord(pack("C", 0xdf)) == 0x3af;
470 \&
471 \& # ... as are eq and cmp ...
472 \&
473 \& print "peta\en" if "\ex{3af}" eq pack("C", 0xdf);
474 \& print "exa\en" if "\ex{3af}" cmp pack("C", 0xdf) == 0;
475 \&
476 \& # ... but pack/unpack C are not affected, in case you still
477 \& # want to go back to your native encoding
478 \&
479 \& print "zetta\en" if unpack("C", (pack("C", 0xdf))) == 0xdf;
480 .Ve
481 .SH BUGS
482 .IX Header "BUGS"
483 .IP "Thread safety" 4
484 .IX Item "Thread safety"
485 \&\f(CW\*(C`use encoding ...\*(C'\fR is not thread-safe (i.e., do not use in threaded
486 applications).
487 .IP "Can't be used by more than one module in a single program." 4
488 .IX Item "Can't be used by more than one module in a single program."
489 Only one encoding is allowed. If you combine modules in a program that have
490 different encodings, only one will be actually used.
491 .ie n .IP "Other modules using ""STDIN"" and ""STDOUT"" get the encoded stream" 4
492 .el .IP "Other modules using \f(CWSTDIN\fR and \f(CWSTDOUT\fR get the encoded stream" 4
493 .IX Item "Other modules using STDIN and STDOUT get the encoded stream"
494 They may be expecting something completely different.
495 .IP "literals in regex that are longer than 127 bytes" 4
496 .IX Item "literals in regex that are longer than 127 bytes"
497 For native multibyte encodings (either fixed or variable length),
498 the current implementation of the regular expressions may introduce
499 recoding errors for regular expression literals longer than 127 bytes.
500 .IP EBCDIC 4
501 .IX Item "EBCDIC"
502 The encoding pragma is not supported on EBCDIC platforms.
503 .ie n .IP """format""" 4
504 .el .IP \f(CWformat\fR 4
505 .IX Item "format"
506 This pragma doesn't work well with \f(CW\*(C`format\*(C'\fR because PerlIO does not
507 get along very well with it. When \f(CW\*(C`format\*(C'\fR contains non-ASCII
508 characters it prints funny or gets "wide character warnings".
509 To understand it, try the code below.
510 .Sp
511 .Vb 11
512 \& # Save this one in utf8
513 \& # replace *non\-ascii* with a non\-ascii string
514 \& my $camel;
515 \& format STDOUT =
516 \& *non\-ascii*@>>>>>>>
517 \& $camel
518 \& .
519 \& $camel = "*non\-ascii*";
520 \& binmode(STDOUT=>\*(Aq:encoding(utf8)\*(Aq); # bang!
521 \& write; # funny
522 \& print $camel, "\en"; # fine
523 .Ve
524 .Sp
525 Without binmode this happens to work but without binmode, \fBprint()\fR
526 fails instead of \fBwrite()\fR.
527 .Sp
528 At any rate, the very use of \f(CW\*(C`format\*(C'\fR is questionable when it comes to
529 unicode characters since you have to consider such things as character
530 width (i.e. double-width for ideographs) and directions (i.e. BIDI for
531 Arabic and Hebrew).
532 .IP "See also ""CAVEATS""" 4
533 .IX Item "See also ""CAVEATS"""
534 .SH HISTORY
535 .IX Header "HISTORY"
536 This pragma first appeared in Perl v5.8.0. It has been enhanced in later
537 releases as specified above.
538 .SH "SEE ALSO"
539 .IX Header "SEE ALSO"
540 perlunicode, Encode, open, Filter::Util::Call,
541 .PP
542 Ch. 15 of \f(CW\*(C`Programming Perl (3rd Edition)\*(C'\fR
543 by Larry Wall, Tom Christiansen, Jon Orwant;
544 O'Reilly & Associates; ISBN 0\-596\-00027\-8