Debian Codegraph @

upstream/fedora-44/man1/perlunifaq.1

raw
1 .\" -*- mode: troff; coding: utf-8 -*-
2 .\" Automatically generated by Pod::Man v6.0.2 (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 .\" Required to disable full justification in groff 1.23.0.
57 .if n .ds AD l
58 .\" ========================================================================
59 .\"
60 .IX Title "PERLUNIFAQ 1"
61 .TH PERLUNIFAQ 1 2026-03-09 "perl v5.42.1" "Perl Programmers Reference Guide"
62 .\" For nroff, turn off justification. Always turn off hyphenation; it makes
63 .\" way too many mistakes in technical documents.
64 .if n .ad l
65 .nh
66 .SH NAME
67 perlunifaq \- Perl Unicode FAQ
68 .SH "Q and A"
69 .IX Header "Q and A"
70 This is a list of questions and answers about Unicode in Perl, intended to be
71 read after perlunitut.
72 .SS "perlunitut isn\*(Aqt really a Unicode tutorial, is it?"
73 .IX Subsection "perlunitut isn't really a Unicode tutorial, is it?"
74 No, and this isn\*(Aqt really a Unicode FAQ.
75 .PP
76 Perl has an abstracted interface for all supported character encodings, so this
77 is actually a generic \f(CW\*(C`Encode\*(C'\fR tutorial and \f(CW\*(C`Encode\*(C'\fR FAQ. But many people
78 think that Unicode is special and magical, and I didn\*(Aqt want to disappoint
79 them, so I decided to call the document a Unicode tutorial.
80 .SS "What character encodings does Perl support?"
81 .IX Subsection "What character encodings does Perl support?"
82 To find out which character encodings your Perl supports, run:
83 .PP
84 .Vb 1
85 \& perl \-MEncode \-le "print for Encode\->encodings(\*(Aq:all\*(Aq)"
86 .Ve
87 .SS "Which version of perl should I use?"
88 .IX Subsection "Which version of perl should I use?"
89 Well, if you can, upgrade to the most recent, but certainly \f(CW5.8.1\fR or newer.
90 The tutorial and FAQ assume the latest release.
91 .PP
92 You should also check your modules, and upgrade them if necessary. For example,
93 HTML::Entities requires version >= 1.32 to function correctly, even though the
94 changelog is silent about this.
95 .SS "What about binary data, like images?"
96 .IX Subsection "What about binary data, like images?"
97 Well, apart from a bare \f(CW\*(C`binmode $fh\*(C'\fR, you shouldn\*(Aqt treat them specially.
98 (The binmode is needed because otherwise Perl may convert line endings on Win32
99 systems.)
100 .PP
101 Be careful, though, to never combine text strings with binary strings. If you
102 need text in a binary stream, encode your text strings first using the
103 appropriate encoding, then join them with binary strings. See also: "What if I
104 don\*(Aqt encode?".
105 .SS "When should I decode or encode?"
106 .IX Subsection "When should I decode or encode?"
107 Whenever you\*(Aqre communicating text with anything that is external to your perl
108 process, like a database, a text file, a socket, or another program. Even if
109 the thing you\*(Aqre communicating with is also written in Perl.
110 .SS "What if I don\*(Aqt decode?"
111 .IX Subsection "What if I don't decode?"
112 Whenever your encoded, binary string is used together with a text string, Perl
113 will assume that your binary string was encoded with ISO\-8859\-1, also known as
114 latin\-1. If it wasn\*(Aqt latin\-1, then your data is unpleasantly converted. For
115 example, if it was UTF\-8, the individual bytes of multibyte characters are seen
116 as separate characters, and then again converted to UTF\-8. Such double encoding
117 can be compared to double HTML encoding (\f(CW\*(C`&amp;gt;\*(C'\fR), or double URI encoding
118 (\f(CW%253E\fR).
119 .PP
120 This silent implicit decoding is known as "upgrading". That may sound
121 positive, but it\*(Aqs best to avoid it.
122 .SS "What if I don\*(Aqt encode?"
123 .IX Subsection "What if I don't encode?"
124 It depends on what you output and how you output it.
125 .PP
126 \fIOutput via a filehandle\fR
127 .IX Subsection "Output via a filehandle"
128 .IP \(bu 4
129 If the string\*(Aqs characters are all code point 255 or lower, Perl
130 outputs bytes that match those code points. This is what happens with encoded
131 strings. It can also, though, happen with unencoded strings that happen to be
132 all code point 255 or lower.
133 .IP \(bu 4
134 Otherwise, Perl outputs the string encoded as UTF\-8. This only happens
135 with strings you neglected to encode. Since that should not happen, Perl also
136 throws a "wide character" warning in this case.
137 .PP
138 \fIOther output mechanisms (e.g., \fR\f(CI\*(C`exec\*(C'\fR\fI, \fR\f(CI\*(C`chdir\*(C'\fR\fI, ..)\fR
139 .IX Subsection "Other output mechanisms (e.g., exec, chdir, ..)"
140 .PP
141 Your text string will be sent using the bytes in Perl\*(Aqs internal format.
142 .PP
143 Because the internal format is often UTF\-8, these bugs are hard to spot,
144 because UTF\-8 is usually the encoding you wanted! But don\*(Aqt be lazy, and don\*(Aqt
145 use the fact that Perl\*(Aqs internal format is UTF\-8 to your advantage. Encode
146 explicitly to avoid weird bugs, and to show to maintenance programmers that you
147 thought this through.
148 .SS "Is there a way to automatically decode or encode?"
149 .IX Subsection "Is there a way to automatically decode or encode?"
150 If all data that comes from a certain handle is encoded in exactly the same
151 way, you can tell the PerlIO system to automatically decode everything, with
152 the \f(CW\*(C`encoding\*(C'\fR layer. If you do this, you can\*(Aqt accidentally forget to decode
153 or encode anymore, on things that use the layered handle.
154 .PP
155 You can provide this layer when \f(CW\*(C`open\*(C'\fRing the file:
156 .PP
157 .Vb 2
158 \& open my $fh, \*(Aq>:encoding(UTF\-8)\*(Aq, $filename; # auto encoding on write
159 \& open my $fh, \*(Aq<:encoding(UTF\-8)\*(Aq, $filename; # auto decoding on read
160 .Ve
161 .PP
162 Or if you already have an open filehandle:
163 .PP
164 .Vb 1
165 \& binmode $fh, \*(Aq:encoding(UTF\-8)\*(Aq;
166 .Ve
167 .PP
168 Some database drivers for DBI can also automatically encode and decode, but
169 that is sometimes limited to the UTF\-8 encoding.
170 .SS "What if I don\*(Aqt know which encoding was used?"
171 .IX Subsection "What if I don't know which encoding was used?"
172 Do whatever you can to find out, and if you have to: guess. (Don\*(Aqt forget to
173 document your guess with a comment.)
174 .PP
175 You could open the document in a web browser, and change the character set or
176 character encoding until you can visually confirm that all characters look the
177 way they should.
178 .PP
179 There is no way to reliably detect the encoding automatically, so if people
180 keep sending you data without charset indication, you may have to educate them.
181 .SS "Can I use Unicode in my Perl sources?"
182 .IX Subsection "Can I use Unicode in my Perl sources?"
183 Yes, you can! If your sources are UTF\-8 encoded, you can indicate that with the
184 \&\f(CW\*(C`use utf8\*(C'\fR pragma.
185 .PP
186 .Vb 1
187 \& use utf8;
188 .Ve
189 .PP
190 This doesn\*(Aqt do anything to your input, or to your output. It only influences
191 the way your sources are read. You can use Unicode in string literals, in
192 identifiers (but they still have to be "word characters" according to \f(CW\*(C`\ew\*(C'\fR),
193 and even in custom delimiters.
194 .SS "Data::Dumper doesn\*(Aqt restore the UTF8 flag; is it broken?"
195 .IX Subsection "Data::Dumper doesn't restore the UTF8 flag; is it broken?"
196 No, Data::Dumper\*(Aqs Unicode abilities are as they should be. There have been
197 some complaints that it should restore the UTF8 flag when the data is read
198 again with \f(CW\*(C`eval\*(C'\fR. However, you should really not look at the flag, and
199 nothing indicates that Data::Dumper should break this rule.
200 .PP
201 Here\*(Aqs what happens: when Perl reads in a string literal, it sticks to 8 bit
202 encoding as long as it can. (But perhaps originally it was internally encoded
203 as UTF\-8, when you dumped it.) When it has to give that up because other
204 characters are added to the text string, it silently upgrades the string to
205 UTF\-8.
206 .PP
207 If you properly encode your strings for output, none of this is of your
208 concern, and you can just \f(CW\*(C`eval\*(C'\fR dumped data as always.
209 .SS "Why do regex character classes sometimes match only in the ASCII range?"
210 .IX Subsection "Why do regex character classes sometimes match only in the ASCII range?"
211 Starting in Perl 5.14 (and partially in Perl 5.12), just put a
212 \&\f(CW\*(C`use feature \*(Aqunicode_strings\*(Aq\*(C'\fR near the beginning of your program.
213 Within its lexical scope you shouldn\*(Aqt have this problem. It also is
214 automatically enabled under \f(CW\*(C`use feature \*(Aq:5.12\*(Aq\*(C'\fR or \f(CW\*(C`use v5.12\*(C'\fR or
215 using \f(CW\*(C`\-E\*(C'\fR on the command line for Perl 5.12 or higher.
216 .PP
217 The rationale for requiring this is to not break older programs that
218 rely on the way things worked before Unicode came along. Those older
219 programs knew only about the ASCII character set, and so may not work
220 properly for additional characters. When a string is encoded in UTF\-8,
221 Perl assumes that the program is prepared to deal with Unicode, but when
222 the string isn\*(Aqt, Perl assumes that only ASCII
223 is wanted, and so those characters that are not ASCII
224 characters aren\*(Aqt recognized as to what they would be in Unicode.
225 \&\f(CW\*(C`use feature \*(Aqunicode_strings\*(Aq\*(C'\fR tells Perl to treat all characters as
226 Unicode, whether the string is encoded in UTF\-8 or not, thus avoiding
227 the problem.
228 .PP
229 However, on earlier Perls, or if you pass strings to subroutines outside
230 the feature\*(Aqs scope, you can force Unicode rules by changing the
231 encoding to UTF\-8 by doing \f(CWutf8::upgrade($string)\fR. This can be used
232 safely on any string, as it checks and does not change strings that have
233 already been upgraded.
234 .PP
235 For a more detailed discussion, see Unicode::Semantics on CPAN.
236 .SS "Why do some characters not uppercase or lowercase correctly?"
237 .IX Subsection "Why do some characters not uppercase or lowercase correctly?"
238 See the answer to the previous question.
239 .SS "How can I determine if a string is a text string or a binary string?"
240 .IX Subsection "How can I determine if a string is a text string or a binary string?"
241 You can\*(Aqt. Some use the UTF8 flag for this, but that\*(Aqs misuse, and makes well
242 behaved modules like Data::Dumper look bad. The flag is useless for this
243 purpose, because it\*(Aqs off when an 8 bit encoding (by default ISO\-8859\-1) is
244 used to store the string.
245 .PP
246 This is something you, the programmer, has to keep track of; sorry. You could
247 consider adopting a kind of "Hungarian notation" to help with this.
248 .SS "How do I convert from encoding FOO to encoding BAR?"
249 .IX Subsection "How do I convert from encoding FOO to encoding BAR?"
250 By first converting the FOO\-encoded byte string to a text string, and then the
251 text string to a BAR\-encoded byte string:
252 .PP
253 .Vb 2
254 \& my $text_string = decode(\*(AqFOO\*(Aq, $foo_string);
255 \& my $bar_string = encode(\*(AqBAR\*(Aq, $text_string);
256 .Ve
257 .PP
258 or by skipping the text string part, and going directly from one binary
259 encoding to the other:
260 .PP
261 .Vb 2
262 \& use Encode qw(from_to);
263 \& from_to($string, \*(AqFOO\*(Aq, \*(AqBAR\*(Aq); # changes contents of $string
264 .Ve
265 .PP
266 or by letting automatic decoding and encoding do all the work:
267 .PP
268 .Vb 3
269 \& open my $foofh, \*(Aq<:encoding(FOO)\*(Aq, \*(Aqexample.foo.txt\*(Aq;
270 \& open my $barfh, \*(Aq>:encoding(BAR)\*(Aq, \*(Aqexample.bar.txt\*(Aq;
271 \& print { $barfh } $_ while <$foofh>;
272 .Ve
273 .ie n .SS "What are ""decode_utf8"" and ""encode_utf8""?"
274 .el .SS "What are \f(CWdecode_utf8\fP and \f(CWencode_utf8\fP?"
275 .IX Subsection "What are decode_utf8 and encode_utf8?"
276 These are alternate syntaxes for \f(CW\*(C`decode(\*(Aqutf8\*(Aq, ...)\*(C'\fR and \f(CW\*(C`encode(\*(Aqutf8\*(Aq,
277 \&...)\*(C'\fR. Do not use these functions for data exchange. Instead use
278 \&\f(CW\*(C`decode(\*(AqUTF\-8\*(Aq, ...)\*(C'\fR and \f(CW\*(C`encode(\*(AqUTF\-8\*(Aq, ...)\*(C'\fR; see
279 "What\*(Aqs the difference between UTF\-8 and utf8?" below.
280 .SS "What is a ""wide character""?"
281 .IX Subsection "What is a ""wide character""?"
282 This is a term used for characters occupying more than one byte.
283 .PP
284 The Perl warning "Wide character in ..." is caused by such a character.
285 With no specified encoding layer, Perl tries to
286 fit things into a single byte. When it can\*(Aqt, it
287 emits this warning (if warnings are enabled), and uses UTF\-8 encoded data
288 instead.
289 .PP
290 To avoid this warning and to avoid having different output encodings in a single
291 stream, always specify an encoding explicitly, for example with a PerlIO layer:
292 .PP
293 .Vb 1
294 \& binmode STDOUT, ":encoding(UTF\-8)";
295 .Ve
296 .SH INTERNALS
297 .IX Header "INTERNALS"
298 .SS "What is ""the UTF8 flag""?"
299 .IX Subsection "What is ""the UTF8 flag""?"
300 Please, unless you\*(Aqre hacking the internals, or debugging weirdness, don\*(Aqt
301 think about the UTF8 flag at all. That means that you very probably shouldn\*(Aqt
302 use \f(CW\*(C`is_utf8\*(C'\fR, \f(CW\*(C`_utf8_on\*(C'\fR or \f(CW\*(C`_utf8_off\*(C'\fR at all.
303 .PP
304 The UTF8 flag, also called SvUTF8, is an internal flag that indicates that the
305 current internal representation is UTF\-8. Without the flag, it is assumed to be
306 ISO\-8859\-1. Perl converts between these automatically. (Actually Perl usually
307 assumes the representation is ASCII; see "Why do regex character classes
308 sometimes match only in the ASCII range?" above.)
309 .PP
310 One of Perl\*(Aqs internal formats happens to be UTF\-8. Unfortunately, Perl can\*(Aqt
311 keep a secret, so everyone knows about this. That is the source of much
312 confusion. It\*(Aqs better to pretend that the internal format is some unknown
313 encoding, and that you always have to encode and decode explicitly.
314 .ie n .SS "What about the ""use bytes"" pragma?"
315 .el .SS "What about the \f(CWuse bytes\fP pragma?"
316 .IX Subsection "What about the use bytes pragma?"
317 Don\*(Aqt use it. It makes no sense to deal with bytes in a text string, and it
318 makes no sense to deal with characters in a byte string. Do the proper
319 conversions (by decoding/encoding), and things will work out well: you get
320 character counts for decoded data, and byte counts for encoded data.
321 .PP
322 \&\f(CW\*(C`use bytes\*(C'\fR is usually a failed attempt to do something useful. Just forget
323 about it.
324 .ie n .SS "What about the ""use encoding"" pragma?"
325 .el .SS "What about the \f(CWuse encoding\fP pragma?"
326 .IX Subsection "What about the use encoding pragma?"
327 Don\*(Aqt use it. Unfortunately, it assumes that the programmer\*(Aqs environment and
328 that of the user will use the same encoding. It will use the same encoding for
329 the source code and for STDIN and STDOUT. When a program is copied to another
330 machine, the source code does not change, but the STDIO environment might.
331 .PP
332 If you need non\-ASCII characters in your source code, make it a UTF\-8 encoded
333 file and \f(CW\*(C`use utf8\*(C'\fR.
334 .PP
335 If you need to set the encoding for STDIN, STDOUT, and STDERR, for example
336 based on the user\*(Aqs locale, \f(CW\*(C`use open\*(C'\fR.
337 .ie n .SS "What is the difference between "":encoding"" and "":utf8""?"
338 .el .SS "What is the difference between \f(CW:encoding\fP and \f(CW:utf8\fP?"
339 .IX Subsection "What is the difference between :encoding and :utf8?"
340 Because UTF\-8 is one of Perl\*(Aqs internal formats, you can often just skip the
341 encoding or decoding step, and manipulate the UTF8 flag directly.
342 .PP
343 Instead of \f(CW:encoding(UTF\-8)\fR, you can simply use \f(CW\*(C`:utf8\*(C'\fR, which skips the
344 encoding step if the data was already represented as UTF8 internally. This is
345 widely accepted as good behavior when you\*(Aqre writing, but it can be dangerous
346 when reading, because it causes internal inconsistency when you have invalid
347 byte sequences. Using \f(CW\*(C`:utf8\*(C'\fR for input can sometimes result in security
348 breaches, so please use \f(CW:encoding(UTF\-8)\fR instead.
349 .PP
350 Instead of \f(CW\*(C`decode\*(C'\fR and \f(CW\*(C`encode\*(C'\fR, you could use \f(CW\*(C`_utf8_on\*(C'\fR and \f(CW\*(C`_utf8_off\*(C'\fR,
351 but this is considered bad style. Especially \f(CW\*(C`_utf8_on\*(C'\fR can be dangerous, for
352 the same reason that \f(CW\*(C`:utf8\*(C'\fR can.
353 .PP
354 There are some shortcuts for oneliners;
355 see \-C in perlrun.
356 .ie n .SS "What\*(Aqs the difference between ""UTF\-8"" and ""utf8""?"
357 .el .SS "What\*(Aqs the difference between \f(CWUTF\-8\fP and \f(CWutf8\fP?"
358 .IX Subsection "What's the difference between UTF-8 and utf8?"
359 \&\f(CW\*(C`UTF\-8\*(C'\fR is the official standard. \f(CW\*(C`utf8\*(C'\fR is Perl\*(Aqs way of being liberal in
360 what it accepts. If you have to communicate with things that aren\*(Aqt so liberal,
361 you may want to consider using \f(CW\*(C`UTF\-8\*(C'\fR. If you have to communicate with things
362 that are too liberal, you may have to use \f(CW\*(C`utf8\*(C'\fR. The full explanation is in
363 "UTF\-8 vs. utf8 vs. UTF8" in Encode.
364 .PP
365 \&\f(CW\*(C`UTF\-8\*(C'\fR is internally known as \f(CW\*(C`utf\-8\-strict\*(C'\fR. The tutorial uses UTF\-8
366 consistently, even where utf8 is actually used internally, because the
367 distinction can be hard to make, and is mostly irrelevant.
368 .PP
369 For example, utf8 can be used for code points that don\*(Aqt exist in Unicode, like
370 9999999, but if you encode that to UTF\-8, you get a substitution character (by
371 default; see "Handling Malformed Data" in Encode for more ways of dealing with
372 this.)
373 .PP
374 Okay, if you insist: the "internal format" is utf8, not UTF\-8. (When it\*(Aqs not
375 some other encoding.)
376 .SS "I lost track; what encoding is the internal format really?"
377 .IX Subsection "I lost track; what encoding is the internal format really?"
378 It\*(Aqs good that you lost track, because you shouldn\*(Aqt depend on the internal
379 format being any specific encoding. But since you asked: by default, the
380 internal format is either ISO\-8859\-1 (latin\-1), or utf8, depending on the
381 history of the string. On EBCDIC platforms, this may be different even.
382 .PP
383 Perl knows how it stored the string internally, and will use that knowledge
384 when you \f(CW\*(C`encode\*(C'\fR. In other words: don\*(Aqt try to find out what the internal
385 encoding for a certain string is, but instead just encode it into the encoding
386 that you want.
387 .SH AUTHOR
388 .IX Header "AUTHOR"
389 Juerd Waalboer <#####@juerd.nl>
390 .SH "SEE ALSO"
391 .IX Header "SEE ALSO"
392 perlunicode, perluniintro, Encode