This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
More portability defines, now mostly type-related
[perl5.git] / pod / perluniintro.pod
CommitLineData
ba62762e
JH
1=head1 NAME
2
07fcf8ff 3perluniintro - Perl Unicode introduction
ba62762e
JH
4
5=head1 DESCRIPTION
6
7This document gives a general idea of Unicode and how to use Unicode
8in Perl.
9
10=head2 Unicode
11
376d9008 12Unicode is a character set standard which plans to codify all of the
ba62762e
JH
13writing systems of the world, plus many other symbols.
14
15Unicode and ISO/IEC 10646 are coordinated standards that provide code
376d9008 16points for characters in almost all modern character set standards,
ba62762e 17covering more than 30 writing systems and hundreds of languages,
376d9008 18including all commercially-important modern languages. All characters
ba62762e
JH
19in the largest Chinese, Japanese, and Korean dictionaries are also
20encoded. The standards will eventually cover almost all characters in
21more than 250 writing systems and thousands of languages.
22
23A Unicode I<character> is an abstract entity. It is not bound to any
376d9008
JB
24particular integer width, especially not to the C language C<char>.
25Unicode is language-neutral and display-neutral: it does not encode the
26language of the text and it does not define fonts or other graphical
ba62762e
JH
27layout details. Unicode operates on characters and on text built from
28those characters.
29
30Unicode defines characters like C<LATIN CAPITAL LETTER A> or C<GREEK
376d9008
JB
31SMALL LETTER ALPHA> and unique numbers for the characters, in this
32case 0x0041 and 0x03B1, respectively. These unique numbers are called
33I<code points>.
ba62762e
JH
34
35The Unicode standard prefers using hexadecimal notation for the code
376d9008
JB
36points. If numbers like "C<0x0041" are unfamiliar to
37you, take a peek at a later section, L</"Hexadecimal Notation">.
ba62762e 38The Unicode standard uses the notation C<U+0041 LATIN CAPITAL LETTER A>,
376d9008 39to give the hexadecimal code point and the normative name of
ba62762e
JH
40the character.
41
42Unicode also defines various I<properties> for the characters, like
376d9008 43"uppercase" or "lowercase", "decimal digit", or "punctuation";
ba62762e
JH
44these properties are independent of the names of the characters.
45Furthermore, various operations on the characters like uppercasing,
376d9008 46lowercasing, and collating (sorting) are defined.
ba62762e
JH
47
48A Unicode character consists either of a single code point, or a
49I<base character> (like C<LATIN CAPITAL LETTER A>), followed by one or
50more I<modifiers> (like C<COMBINING ACUTE ACCENT>). This sequence of
376d9008 51base character and modifiers is called a I<combining character
ba62762e
JH
52sequence>.
53
376d9008
JB
54Whether to call these combining character sequences "characters"
55depends on your point of view. If you are a programmer, you probably
56would tend towards seeing each element in the sequences as one unit,
57or "character". The whole sequence could be seen as one "character",
58however, from the user's point of view, since that's probably what it
59looks like in the context of the user's language.
60
61With this "whole sequence" view of characters, the total number of
62characters is open-ended. But in the programmer's "one unit is one
63character" point of view, the concept of "characters" is more
64deterministic. In this document, we take that second point of view: one
65"character" is one Unicode code point, be it a base character or a
66combining character.
67
68For some combinations, there are I<precomposed> characters.
69C<LATIN CAPITAL LETTER A WITH ACUTE>, for example, is defined as
ba62762e 70a single code point. These precomposed characters are, however,
376d9008 71only available for some combinations, and are mainly
ba62762e 72meant to support round-trip conversions between Unicode and legacy
376d9008
JB
73standards (like the ISO 8859). In the general case, the composing
74method is more extensible. To support conversion between
ba62762e 75different compositions of the characters, various I<normalization
376d9008 76forms> to standardize representations are also defined.
ba62762e
JH
77
78Because of backward compatibility with legacy encodings, the "a unique
376d9008
JB
79number for every character" idea breaks down a bit: instead, there is
80"at least one number for every character". The same character could
81be represented differently in several legacy encodings. The
82converse is also not true: some code points do not have an assigned
83character. Firstly, there are unallocated code points within
84otherwise used blocks. Secondly, there are special Unicode control
85characters that do not represent true characters.
ba62762e
JH
86
87A common myth about Unicode is that it would be "16-bit", that is,
376d9008
JB
88Unicode is only represented as C<0x10000> (or 65536) characters from
89C<0x0000> to C<0xFFFF>. B<This is untrue.> Since Unicode 2.0, Unicode
90has been defined all the way up to 21 bits (C<0x10FFFF>), and since
91Unicode 3.1, characters have been defined beyond C<0xFFFF>. The first
92C<0x10000> characters are called the I<Plane 0>, or the I<Basic
93Multilingual Plane> (BMP). With Unicode 3.1, 17 planes in all are
94defined--but nowhere near full of defined characters, yet.
ba62762e
JH
95
96Another myth is that the 256-character blocks have something to do
376d9008
JB
97with languages--that each languages is specified inside a block.
98B<This is also untrue.> The division into blocks exists, but it is
99almost completely accidental--an artifact of how the characters have
100been historically allocated. Instead, there is a concept called
101I<scripts>, which is more useful: there is C<Latin> script,
102C<Greek> script, and so on. Scripts usually span varied parts of
103several blocks. For further information see L<Unicode::UCD>.
ba62762e
JH
104
105The Unicode code points are just abstract numbers. To input and
106output these abstract numbers, the numbers must be I<encoded> somehow.
107Unicode defines several I<character encoding forms>, of which I<UTF-8>
108is perhaps the most popular. UTF-8 is a variable length encoding that
109encodes Unicode characters as 1 to 6 bytes (only 4 with the currently
8baee566 110defined characters). Other encodings include UTF-16 and UTF-32 and their
376d9008 111big- and little-endian variants (UTF-8 is byte-order independent).
ba62762e
JH
112The ISO/IEC 10646 defines the UCS-2 and UCS-4 encoding forms.
113
376d9008
JB
114For more information about encodings--for instance, to learn what
115I<surrogates> and I<byte order marks> (BOMs) are--see L<perlunicode>.
ba62762e
JH
116
117=head2 Perl's Unicode Support
118
376d9008
JB
119Starting from Perl 5.6.0, Perl has had the capacity to handle Unicode
120natively. Perl 5.8.0, however, is the first recommended release for
121serious Unicode work. The maintenance release 5.6.1 fixed many of the
122problems of the initial Unicode implementation, but for example
123regular expressions did not really work with Unicode.
ba62762e
JH
124
125B<Starting from Perl 5.8.0, the use of C<use utf8> is no longer
126necessary.> In earlier releases the C<utf8> pragma was used to declare
127that operations in the current block or file would be Unicode-aware.
376d9008
JB
128This model was found to be wrong, or at least clumsy: the "Unicodeness"
129is now carried with the data instead of being attached to the
130operations. Only one case remains where an explicit C<use utf8> is
131needed: if your Perl script itself is encoded in UTF-8, you can use
132UTF-8 in your identifier names, and in string and regular expression
133literals, by saying C<use utf8>. This is not the default because
134scripts with legacy 8-bit data in them would break.
ba62762e
JH
135
136=head2 Perl's Unicode Model
137
376d9008
JB
138Perl supports both pre-5.6 strings of eight-bit native bytes, and
139strings of Unicode characters. The principle is that Perl tries to
140keep its data as eight-bit bytes for as long as possible, but as soon
141as Unicodeness cannot be avoided, the data is transparently upgraded
142to Unicode.
ba62762e 143
4192de81 144Internally, Perl currently uses either whatever the native eight-bit
376d9008
JB
145character set of the platform (for example Latin-1) is, defaulting to
146UTF-8, to encode Unicode strings. Specifically, if all code points in
147the string are C<0xFF> or less, Perl uses the native eight-bit
148character set. Otherwise, it uses UTF-8.
4192de81 149
7ca610e8 150A user of Perl does not normally need to know nor care how Perl
20ba30f4 151happens to encode its internal strings, but it becomes relevant when
376d9008
JB
152outputting Unicode strings to a stream without a discipline--one with
153the "default" encoding. In such a case, the raw bytes used internally
7ca610e8
JH
154(the native character set or UTF-8, as appropriate for each string)
155will be used, and a "Wide character" warning will be issued if those
156strings contain a character beyond 0x00FF.
4192de81
JH
157
158For example,
159
7ca610e8 160 perl -e 'print "\x{DF}\n", "\x{0100}\x{DF}\n"'
4192de81
JH
161
162produces a fairly useless mixture of native bytes and UTF-8, as well
163as a warning.
164
376d9008 165To output UTF-8 always, use the "C<:utf8>" output discipline. Prepending
4192de81
JH
166
167 binmode(STDOUT, ":utf8");
168
376d9008
JB
169to this sample program ensures that the output is completely UTF-8,
170and helpfully removes the program's warning.
ba62762e 171
376d9008
JB
172If your locale environment variables (C<LANGUAGE>, C<LC_ALL>,
173C<LC_CTYPE>, C<LANG>) contain the strings 'UTF-8' or 'UTF8',
174regardless of case, then the default encoding of your STDIN, STDOUT,
175and STDERR and of B<any subsequent file open>, is UTF-8. Note that
176this means that Perl expects other software to work, too: if STDIN
177coming in from another command is not UTF-8, Perl will complain about
ac730995 178malformed UTF-8.
b310b053 179
64c66fb6
JH
180=head2 Unicode and EBCDIC
181
182Perl 5.8.0 also supports Unicode on EBCDIC platforms. There,
376d9008 183Unicode support is somewhat more complex to implement since
64c66fb6 184additional conversions are needed at every step. Some problems
dc4af4bb 185remain, see L<perlebcdic> for details.
64c66fb6
JH
186
187In any case, the Unicode support on EBCDIC platforms is better than
188in the 5.6 series, which didn't work much at all for EBCDIC platform.
189On EBCDIC platforms, the internal Unicode encoding form is UTF-EBCDIC
376d9008
JB
190instead of UTF-8. The difference is that as UTF-8 is "ASCII-safe" in
191that ASCII characters encode to UTF-8 as-is, while UTF-EBCDIC is
192"EBCDIC-safe".
ba62762e
JH
193
194=head2 Creating Unicode
195
376d9008
JB
196To create Unicode characters in literals for code points above C<0xFF>,
197use the C<\x{...}> notation in double-quoted strings:
ba62762e
JH
198
199 my $smiley = "\x{263a}";
200
376d9008 201Similarly, it can be used in regular expression literals
ba62762e
JH
202
203 $smiley =~ /\x{263a}/;
204
205At run-time you can use C<chr()>:
206
207 my $hebrew_alef = chr(0x05d0);
208
376d9008 209See L</"Further Resources"> for how to find all these numeric codes.
ba62762e 210
376d9008
JB
211Naturally, C<ord()> will do the reverse: it turns a character into
212a code point.
ba62762e 213
a5f0baef 214Note that C<\x..> (no C<{}> and only two hexadecimal digits),
376d9008 215C<\x{...}>, and C<chr(...)> for arguments less than C<0x100> (decimal
a5f0baef 216256) generate an eight-bit character for backward compatibility with
376d9008 217older Perls. For arguments of C<0x100> or more, Unicode characters are
a5f0baef
JH
218always produced. If you want to force the production of Unicode
219characters regardless of the numeric value, use C<pack("U", ...)>
220instead of C<\x..>, C<\x{...}>, or C<chr()>.
ba62762e
JH
221
222You can also use the C<charnames> pragma to invoke characters
376d9008 223by name in double-quoted strings:
ba62762e
JH
224
225 use charnames ':full';
226 my $arabic_alef = "\N{ARABIC LETTER ALEF}";
227
228And, as mentioned above, you can also C<pack()> numbers into Unicode
229characters:
230
231 my $georgian_an = pack("U", 0x10a0);
232
8a5e5dd5
JH
233Note that both C<\x{...}> and C<\N{...}> are compile-time string
234constants: you cannot use variables in them. if you want similar
235run-time functionality, use C<chr()> and C<charnames::vianame()>.
236
1eda90df
JH
237Also note that if all the code points for pack "U" are below 0x100,
238bytes will be generated, just like if you were using C<chr()>.
239
240 my $bytes = pack("U*", 0x80, 0xFF);
241
242If you want to force the result to Unicode characters, use the special
243C<"U0"> prefix. It consumes no arguments but forces the result to be
244in Unicode characters, instead of bytes.
245
246 my $chars = pack("U0U*", 0x80, 0xFF);
247
ba62762e
JH
248=head2 Handling Unicode
249
250Handling Unicode is for the most part transparent: just use the
251strings as usual. Functions like C<index()>, C<length()>, and
252C<substr()> will work on the Unicode characters; regular expressions
253will work on the Unicode characters (see L<perlunicode> and L<perlretut>).
254
376d9008
JB
255Note that Perl considers combining character sequences
256to be characters, so for example
ba62762e
JH
257
258 use charnames ':full';
259 print length("\N{LATIN CAPITAL LETTER A}\N{COMBINING ACUTE ACCENT}"), "\n";
260
261will print 2, not 1. The only exception is that regular expressions
262have C<\X> for matching a combining character sequence.
263
376d9008
JB
264Life is not quite so transparent, however, when working with legacy
265encodings, I/O, and certain special cases:
ba62762e
JH
266
267=head2 Legacy Encodings
268
269When you combine legacy data and Unicode the legacy data needs
270to be upgraded to Unicode. Normally ISO 8859-1 (or EBCDIC, if
271applicable) is assumed. You can override this assumption by
272using the C<encoding> pragma, for example
273
274 use encoding 'latin2'; # ISO 8859-2
275
376d9008
JB
276in which case literals (string or regular expressions), C<chr()>, and
277C<ord()> in your whole script are assumed to produce Unicode
278characters from ISO 8859-2 code points. Note that the matching for
279encoding names is forgiving: instead of C<latin2> you could have
280said C<Latin 2>, or C<iso8859-2>, or other variations. With just
ba62762e
JH
281
282 use encoding;
283
376d9008
JB
284the environment variable C<PERL_ENCODING> will be consulted.
285If that variable isn't set, the encoding pragma will fail.
ba62762e 286
376d9008 287The C<Encode> module knows about many encodings and has interfaces
ba62762e
JH
288for doing conversions between those encodings:
289
290 use Encode 'from_to';
291 from_to($data, "iso-8859-3", "utf-8"); # from legacy to utf-8
292
293=head2 Unicode I/O
294
8baee566 295Normally, writing out Unicode data
ba62762e 296
8baee566 297 print FH $some_string_with_unicode, "\n";
ba62762e 298
8baee566 299produces raw bytes that Perl happens to use to internally encode the
376d9008
JB
300Unicode string. Perl's internal encoding depends on the system as
301well as what characters happen to be in the string at the time. If
302any of the characters are at code points C<0x100> or above, you will get
303a warning. To ensure that the output is explicitly rendered in the
304encoding you desire--and to avoid the warning--open the stream with
305the desired encoding. Some examples:
ba62762e 306
ec90690f
ST
307 open FH, ">:utf8", "file";
308
309 open FH, ">:encoding(ucs2)", "file";
310 open FH, ">:encoding(UTF-8)", "file";
311 open FH, ">:encoding(shift_jis)", "file";
1d7919c5 312
376d9008 313and on already open streams, use C<binmode()>:
1d7919c5
JH
314
315 binmode(STDOUT, ":utf8");
316
ec90690f
ST
317 binmode(STDOUT, ":encoding(ucs2)");
318 binmode(STDOUT, ":encoding(UTF-8)");
319 binmode(STDOUT, ":encoding(shift_jis)");
320
b5d8778e
JH
321The matching of encoding names is loose: case does not matter, and
322many encodings have several aliases. Note that C<:utf8> discipline
376d9008 323must always be specified exactly like that; it is I<not> subject to the
b5d8778e
JH
324loose matching of encoding names.
325
376d9008
JB
326See L<PerlIO> for the C<:utf8> layer, L<PerlIO::encoding> and
327L<Encode::PerlIO> for the C<:encoding()> layer, and
328L<Encode::Supported> for many encodings supported by the C<Encode>
329module.
ba62762e 330
a5f0baef
JH
331Reading in a file that you know happens to be encoded in one of the
332Unicode encodings does not magically turn the data into Unicode in
333Perl's eyes. To do that, specify the appropriate discipline when
334opening files
ba62762e
JH
335
336 open(my $fh,'<:utf8', 'anything');
8baee566
JF
337 my $line_of_unicode = <$fh>;
338
ec90690f 339 open(my $fh,'<:encoding(Big5)', 'anything');
8baee566 340 my $line_of_unicode = <$fh>;
ba62762e
JH
341
342The I/O disciplines can also be specified more flexibly with
376d9008 343the C<open> pragma. See L<open>, or look at the following example.
ba62762e 344
1d7919c5
JH
345 use open ':utf8'; # input and output default discipline will be UTF-8
346 open X, ">file";
347 print X chr(0x100), "\n";
ba62762e 348 close X;
1d7919c5 349 open Y, "<file";
ba62762e
JH
350 printf "%#x\n", ord(<Y>); # this should print 0x100
351 close Y;
352
353With the C<open> pragma you can use the C<:locale> discipline
354
1ecefa54
JH
355 $ENV{LC_ALL} = $ENV{LANG} = 'ru_RU.KOI8-R';
356 # the :locale will probe the locale environment variables like LC_ALL
ba62762e
JH
357 use open OUT => ':locale'; # russki parusski
358 open(O, ">koi8");
359 print O chr(0x430); # Unicode CYRILLIC SMALL LETTER A = KOI8-R 0xc1
360 close O;
361 open(I, "<koi8");
362 printf "%#x\n", ord(<I>), "\n"; # this should print 0xc1
363 close I;
364
365or you can also use the C<':encoding(...)'> discipline
366
367 open(my $epic,'<:encoding(iso-8859-7)','iliad.greek');
8baee566 368 my $line_of_unicode = <$epic>;
ba62762e 369
8baee566
JF
370These methods install a transparent filter on the I/O stream that
371converts data from the specified encoding when it is read in from the
a5f0baef 372stream. The result is always Unicode.
ba62762e
JH
373
374The L<open> pragma affects all the C<open()> calls after the pragma by
375setting default disciplines. If you want to affect only certain
376streams, use explicit disciplines directly in the C<open()> call.
377
378You can switch encodings on an already opened stream by using
8baee566 379C<binmode()>; see L<perlfunc/binmode>.
ba62762e 380
1ecefa54
JH
381The C<:locale> does not currently (as of Perl 5.8.0) work with
382C<open()> and C<binmode()>, only with the C<open> pragma. The
8baee566 383C<:utf8> and C<:encoding(...)> methods do work with all of C<open()>,
1ecefa54 384C<binmode()>, and the C<open> pragma.
ba62762e 385
8baee566 386Similarly, you may use these I/O disciplines on output streams to
a5f0baef
JH
387automatically convert Unicode to the specified encoding when it is
388written to the stream. For example, the following snippet copies the
389contents of the file "text.jis" (encoded as ISO-2022-JP, aka JIS) to
390the file "text.utf8", encoded as UTF-8:
ba62762e 391
8baee566
JF
392 open(my $nihongo, '<:encoding(iso2022-jp)', 'text.jis');
393 open(my $unicode, '>:utf8', 'text.utf8');
394 while (<$nihongo>) { print $unicode }
ba62762e
JH
395
396The naming of encodings, both by the C<open()> and by the C<open>
376d9008
JB
397pragma, is similar to the C<encoding> pragma in that it allows for
398flexible names: C<koi8-r> and C<KOI8R> will both be understood.
ba62762e
JH
399
400Common encodings recognized by ISO, MIME, IANA, and various other
8baee566 401standardisation organisations are recognised; for a more detailed
ba62762e
JH
402list see L<Encode>.
403
404C<read()> reads characters and returns the number of characters.
405C<seek()> and C<tell()> operate on byte counts, as do C<sysread()>
406and C<sysseek()>.
407
8baee566
JF
408Notice that because of the default behaviour of not doing any
409conversion upon input if there is no default discipline,
ba62762e 410it is easy to mistakenly write code that keeps on expanding a file
8baee566 411by repeatedly encoding:
ba62762e
JH
412
413 # BAD CODE WARNING
414 open F, "file";
8baee566 415 local $/; ## read in the whole file of 8-bit characters
ba62762e
JH
416 $t = <F>;
417 close F;
418 open F, ">:utf8", "file";
8baee566 419 print F $t; ## convert to UTF-8 on output
ba62762e
JH
420 close F;
421
422If you run this code twice, the contents of the F<file> will be twice
1d7919c5
JH
423UTF-8 encoded. A C<use open ':utf8'> would have avoided the bug, or
424explicitly opening also the F<file> for input as UTF-8.
ba62762e 425
0c901d84
JH
426B<NOTE>: the C<:utf8> and C<:encoding> features work only if your
427Perl has been built with the new "perlio" feature. Almost all
428Perl 5.8 platforms do use "perlio", though: you can see whether
429yours is by running "perl -V" and looking for C<useperlio=define>.
430
1ecefa54
JH
431=head2 Displaying Unicode As Text
432
433Sometimes you might want to display Perl scalars containing Unicode as
8baee566 434simple ASCII (or EBCDIC) text. The following subroutine converts
1ecefa54 435its argument so that Unicode characters with code points greater than
376d9008
JB
436255 are displayed as "C<\x{...}>", control characters (like "C<\n>") are
437displayed as "C<\x..>", and the rest of the characters as themselves:
1ecefa54 438
58c274a1
JF
439 sub nice_string {
440 join("",
441 map { $_ > 255 ? # if wide character...
8baee566 442 sprintf("\\x{%04X}", $_) : # \x{...}
58c274a1 443 chr($_) =~ /[[:cntrl:]]/ ? # else if control character ...
8baee566 444 sprintf("\\x%02X", $_) : # \x..
58c274a1
JF
445 chr($_) # else as themselves
446 } unpack("U*", $_[0])); # unpack Unicode characters
447 }
448
449For example,
450
451 nice_string("foo\x{100}bar\n")
452
8baee566 453returns:
58c274a1 454
8baee566 455 "foo\x{0100}bar\x0A"
1ecefa54 456
ba62762e
JH
457=head2 Special Cases
458
459=over 4
460
461=item *
462
463Bit Complement Operator ~ And vec()
464
1d9bf7bb
JF
465The bit complement operator C<~> may produce surprising results if used on
466strings containing characters with ordinal values above 255. In such a
467case, the results are consistent with the internal encoding of the
376d9008
JB
468characters, but not with much else. So don't do that. Similarly for C<vec()>:
469you will be operating on the internally-encoded bit patterns of the Unicode
1d9bf7bb
JF
470characters, not on the code point values, which is very probably not what
471you want.
ba62762e
JH
472
473=item *
474
8baee566
JF
475Peeking At Perl's Internal Encoding
476
477Normal users of Perl should never care how Perl encodes any particular
a5f0baef 478Unicode string (because the normal ways to get at the contents of a
376d9008 479string with Unicode--via input and output--should always be via
a5f0baef
JH
480explicitly-defined I/O disciplines). But if you must, there are two
481ways of looking behind the scenes.
ba62762e
JH
482
483One way of peeking inside the internal encoding of Unicode characters
376d9008 484is to use C<unpack("C*", ...> to get the bytes or C<unpack("H*", ...)>
ba62762e
JH
485to display the bytes:
486
8baee566 487 # this prints c4 80 for the UTF-8 bytes 0xc4 0x80
ba62762e
JH
488 print join(" ", unpack("H*", pack("U", 0x100))), "\n";
489
490Yet another way would be to use the Devel::Peek module:
491
492 perl -MDevel::Peek -e 'Dump(chr(0x100))'
493
8baee566 494That shows the UTF8 flag in FLAGS and both the UTF-8 bytes
376d9008 495and Unicode characters in C<PV>. See also later in this document
ba62762e
JH
496the discussion about the C<is_utf8> function of the C<Encode> module.
497
498=back
499
500=head2 Advanced Topics
501
502=over 4
503
504=item *
505
506String Equivalence
507
508The question of string equivalence turns somewhat complicated
376d9008 509in Unicode: what do you mean by "equal"?
ba62762e 510
07698885
RGS
511(Is C<LATIN CAPITAL LETTER A WITH ACUTE> equal to
512C<LATIN CAPITAL LETTER A>?)
ba62762e 513
a5f0baef
JH
514The short answer is that by default Perl compares equivalence (C<eq>,
515C<ne>) based only on code points of the characters. In the above
376d9008
JB
516case, the answer is no (because 0x00C1 != 0x0041). But sometimes, any
517CAPITAL LETTER As should be considered equal, or even As of any case.
ba62762e
JH
518
519The long answer is that you need to consider character normalization
376d9008 520and casing issues: see L<Unicode::Normalize>, Unicode Technical
ba62762e 521Reports #15 and #21, I<Unicode Normalization Forms> and I<Case
376d9008
JB
522Mappings>, http://www.unicode.org/unicode/reports/tr15/ and
523http://www.unicode.org/unicode/reports/tr21/
ba62762e 524
58c274a1 525As of Perl 5.8.0, regular expression case-ignoring matching
ba62762e
JH
526implements only 1:1 semantics: one character matches one character.
527In I<Case Mappings> both 1:N and N:1 matches are defined.
528
529=item *
530
531String Collation
532
376d9008 533People like to see their strings nicely sorted--or as Unicode
ba62762e
JH
534parlance goes, collated. But again, what do you mean by collate?
535
07698885
RGS
536(Does C<LATIN CAPITAL LETTER A WITH ACUTE> come before or after
537C<LATIN CAPITAL LETTER A WITH GRAVE>?)
ba62762e 538
58c274a1 539The short answer is that by default, Perl compares strings (C<lt>,
ba62762e 540C<le>, C<cmp>, C<ge>, C<gt>) based only on the code points of the
376d9008
JB
541characters. In the above case, the answer is "after", since C<0x00C1>
542> C<0x00C0>.
ba62762e
JH
543
544The long answer is that "it depends", and a good answer cannot be
545given without knowing (at the very least) the language context.
546See L<Unicode::Collate>, and I<Unicode Collation Algorithm>
547http://www.unicode.org/unicode/reports/tr10/
548
549=back
550
551=head2 Miscellaneous
552
553=over 4
554
555=item *
556
3ff56b75 557Character Ranges and Classes
ba62762e
JH
558
559Character ranges in regular expression character classes (C</[a-z]/>)
560and in the C<tr///> (also known as C<y///>) operator are not magically
58c274a1 561Unicode-aware. What this means that C<[A-Za-z]> will not magically start
376d9008
JB
562to mean "all alphabetic letters"; not that it does mean that even for
5638-bit characters, you should be using C</[[:alpha:]]/> in that case.
ba62762e 564
376d9008
JB
565For specifying character classes like that in regular expressions, you can use
566the various Unicode properties--C<\pL>, or perhaps C<\p{Alphabetic}>,
a5f0baef 567in this particular case. You can use Unicode code points as the end
376d9008
JB
568points of character ranges, but there is no magic associated with
569specifying a certain range. For further information--there are dozens
570of Unicode character classes--see L<perlunicode>.
ba62762e
JH
571
572=item *
573
574String-To-Number Conversions
575
376d9008
JB
576Unicode does define several other decimal--and numeric--characters
577besides the familiar 0 to 9, such as the Arabic and Indic digits.
ba62762e 578Perl does not support string-to-number conversion for digits other
58c274a1 579than ASCII 0 to 9 (and ASCII a to f for hexadecimal).
ba62762e
JH
580
581=back
582
583=head2 Questions With Answers
584
585=over 4
586
5cb3728c
RB
587=item
588
589Will My Old Scripts Break?
ba62762e
JH
590
591Very probably not. Unless you are generating Unicode characters
376d9008 592somehow, old behaviour should be preserved. About the only
ba62762e
JH
593behaviour that has changed and which could start generating Unicode
594is the old behaviour of C<chr()> where supplying an argument more
376d9008
JB
595than 255 produced a character modulo 255. C<chr(300)>, for example,
596was equal to C<chr(45)>.
ba62762e 597
5cb3728c
RB
598=item
599
600How Do I Make My Scripts Work With Unicode?
ba62762e
JH
601
602Very little work should be needed since nothing changes until you
376d9008
JB
603generate Unicode data. The trick is getting input as Unicode; for
604that, see the earlier I/O discussion.
ba62762e 605
5cb3728c
RB
606=item
607
608How Do I Know Whether My String Is In Unicode?
ba62762e 609
376d9008
JB
610You shouldn't care. No, you really shouldn't. No, really. If you have
611to care--beyond the cases described above--it means that we
ba62762e
JH
612didn't get the transparency of Unicode quite right.
613
614Okay, if you insist:
615
616 use Encode 'is_utf8';
617 print is_utf8($string) ? 1 : 0, "\n";
618
619But note that this doesn't mean that any of the characters in the
620string are necessary UTF-8 encoded, or that any of the characters have
621code points greater than 0xFF (255) or even 0x80 (128), or that the
622string has any characters at all. All the C<is_utf8()> does is to
623return the value of the internal "utf8ness" flag attached to the
376d9008 624C<$string>. If the flag is off, the bytes in the scalar are interpreted
3c1c8017 625as a single byte encoding. If the flag is on, the bytes in the scalar
376d9008 626are interpreted as the (multi-byte, variable-length) UTF-8 encoded code
3c1c8017
AT
627points of the characters. Bytes added to an UTF-8 encoded string are
628automatically upgraded to UTF-8. If mixed non-UTF8 and UTF-8 scalars
376d9008 629are merged (double-quoted interpolation, explicit concatenation, and
3c1c8017
AT
630printf/sprintf parameter substitution), the result will be UTF-8 encoded
631as if copies of the byte strings were upgraded to UTF-8: for example,
632
633 $a = "ab\x80c";
634 $b = "\x{100}";
635 print "$a = $b\n";
636
376d9008
JB
637the output string will be UTF-8-encoded "C<ab\x80c\x{100}\n>", but note
638that C<$a> will stay byte-encoded.
ba62762e
JH
639
640Sometimes you might really need to know the byte length of a string
ce7675db
A
641instead of the character length. For that use either the
642C<Encode::encode_utf8()> function or the C<bytes> pragma and its only
643defined function C<length()>:
ba62762e
JH
644
645 my $unicode = chr(0x100);
646 print length($unicode), "\n"; # will print 1
ce7675db
A
647 require Encode;
648 print length(Encode::encode_utf8($unicode)), "\n"; # will print 2
ba62762e 649 use bytes;
ce7675db 650 print length($unicode), "\n"; # will also print 2 (the 0xC4 0x80 of the UTF-8)
ba62762e 651
5cb3728c
RB
652=item
653
654How Do I Detect Data That's Not Valid In a Particular Encoding?
ba62762e 655
8baee566
JF
656Use the C<Encode> package to try converting it.
657For example,
ba62762e
JH
658
659 use Encode 'encode_utf8';
8baee566 660 if (encode_utf8($string_of_bytes_that_I_think_is_utf8)) {
ba62762e
JH
661 # valid
662 } else {
663 # invalid
664 }
665
8baee566 666For UTF-8 only, you can use:
ba62762e
JH
667
668 use warnings;
8baee566 669 @chars = unpack("U0U*", $string_of_bytes_that_I_think_is_utf8);
ba62762e 670
8baee566 671If invalid, a C<Malformed UTF-8 character (byte 0x##) in
376d9008 672unpack> warning is produced. The "U0" means "expect strictly UTF-8
1eda90df
JH
673encoded Unicode". Without that the C<unpack("U*", ...)>
674would accept also data like C<chr(0xFF>), similarly to the
675C<pack> as we saw earlier.
ba62762e 676
5cb3728c
RB
677=item
678
679How Do I Convert Binary Data Into a Particular Encoding, Or Vice Versa?
ba62762e 680
8baee566
JF
681This probably isn't as useful as you might think.
682Normally, you shouldn't need to.
ba62762e 683
a5f0baef 684In one sense, what you are asking doesn't make much sense: Encodings
376d9008 685are for characters, and binary data are not "characters", so converting
a5f0baef
JH
686"data" into some encoding isn't meaningful unless you know in what
687character set and encoding the binary data is in, in which case it's
376d9008 688not just binary data, now is it?
8baee566
JF
689
690If you have a raw sequence of bytes that you know should be interpreted via
691a particular encoding, you can use C<Encode>:
ba62762e
JH
692
693 use Encode 'from_to';
694 from_to($data, "iso-8859-1", "utf-8"); # from latin-1 to utf-8
695
376d9008 696The call to C<from_to()> changes the bytes in C<$data>, but nothing material
8baee566 697about the nature of the string has changed as far as Perl is concerned.
376d9008
JB
698Both before and after the call, the string C<$data> contains just a bunch of
6998-bit bytes. As far as Perl is concerned, the encoding of the string
700remains as "system-native 8-bit bytes".
8baee566
JF
701
702You might relate this to a fictional 'Translate' module:
703
704 use Translate;
705 my $phrase = "Yes";
706 Translate::from_to($phrase, 'english', 'deutsch');
707 ## phrase now contains "Ja"
ba62762e 708
8baee566
JF
709The contents of the string changes, but not the nature of the string.
710Perl doesn't know any more after the call than before that the contents
711of the string indicates the affirmative.
ba62762e 712
376d9008 713Back to converting data. If you have (or want) data in your system's
a5f0baef
JH
714native 8-bit encoding (e.g. Latin-1, EBCDIC, etc.), you can use
715pack/unpack to convert to/from Unicode.
ba62762e 716
8baee566
JF
717 $native_string = pack("C*", unpack("U*", $Unicode_string));
718 $Unicode_string = pack("U*", unpack("C*", $native_string));
ba62762e
JH
719
720If you have a sequence of bytes you B<know> is valid UTF-8,
721but Perl doesn't know it yet, you can make Perl a believer, too:
722
723 use Encode 'decode_utf8';
8baee566 724 $Unicode = decode_utf8($bytes);
ba62762e
JH
725
726You can convert well-formed UTF-8 to a sequence of bytes, but if
727you just want to convert random binary data into UTF-8, you can't.
728Any random collection of bytes isn't well-formed UTF-8. You can
729use C<unpack("C*", $string)> for the former, and you can create
8baee566 730well-formed Unicode data by C<pack("U*", 0xff, ...)>.
ba62762e 731
5cb3728c
RB
732=item
733
734How Do I Display Unicode? How Do I Input Unicode?
ba62762e 735
076d825e 736See http://www.alanwood.net/unicode/ and
ba62762e
JH
737http://www.cl.cam.ac.uk/~mgk25/unicode.html
738
5cb3728c
RB
739=item
740
741How Does Unicode Work With Traditional Locales?
ba62762e
JH
742
743In Perl, not very well. Avoid using locales through the C<locale>
744pragma. Use only one or the other.
745
746=back
747
748=head2 Hexadecimal Notation
749
376d9008
JB
750The Unicode standard prefers using hexadecimal notation because
751that more clearly shows the division of Unicode into blocks of 256 characters.
ba62762e
JH
752Hexadecimal is also simply shorter than decimal. You can use decimal
753notation, too, but learning to use hexadecimal just makes life easier
376d9008 754with the Unicode standard. The "C<U+HHHH>" notation uses hexadecimal,
076d825e 755for example.
ba62762e
JH
756
757The C<0x> prefix means a hexadecimal number, the digits are 0-9 I<and>
758a-f (or A-F, case doesn't matter). Each hexadecimal digit represents
759four bits, or half a byte. C<print 0x..., "\n"> will show a
760hexadecimal number in decimal, and C<printf "%x\n", $decimal> will
761show a decimal number in hexadecimal. If you have just the
376d9008 762"hex digits" of a hexadecimal number, you can use the C<hex()> function.
ba62762e
JH
763
764 print 0x0009, "\n"; # 9
765 print 0x000a, "\n"; # 10
766 print 0x000f, "\n"; # 15
767 print 0x0010, "\n"; # 16
768 print 0x0011, "\n"; # 17
769 print 0x0100, "\n"; # 256
770
771 print 0x0041, "\n"; # 65
772
773 printf "%x\n", 65; # 41
774 printf "%#x\n", 65; # 0x41
775
776 print hex("41"), "\n"; # 65
777
778=head2 Further Resources
779
780=over 4
781
782=item *
783
784Unicode Consortium
785
786 http://www.unicode.org/
787
788=item *
789
790Unicode FAQ
791
792 http://www.unicode.org/unicode/faq/
793
794=item *
795
796Unicode Glossary
797
798 http://www.unicode.org/glossary/
799
800=item *
801
802Unicode Useful Resources
803
804 http://www.unicode.org/unicode/onlinedat/resources.html
805
806=item *
807
808Unicode and Multilingual Support in HTML, Fonts, Web Browsers and Other Applications
809
076d825e 810 http://www.alanwood.net/unicode/
ba62762e
JH
811
812=item *
813
814UTF-8 and Unicode FAQ for Unix/Linux
815
816 http://www.cl.cam.ac.uk/~mgk25/unicode.html
817
818=item *
819
820Legacy Character Sets
821
822 http://www.czyborra.com/
823 http://www.eki.ee/letter/
824
825=item *
826
827The Unicode support files live within the Perl installation in the
828directory
829
830 $Config{installprivlib}/unicore
831
832in Perl 5.8.0 or newer, and
833
834 $Config{installprivlib}/unicode
835
836in the Perl 5.6 series. (The renaming to F<lib/unicore> was done to
837avoid naming conflicts with lib/Unicode in case-insensitive filesystems.)
551b6b6f 838The main Unicode data file is F<UnicodeData.txt> (or F<Unicode.301> in
ba62762e
JH
839Perl 5.6.1.) You can find the C<$Config{installprivlib}> by
840
841 perl "-V:installprivlib"
842
ba62762e
JH
843You can explore various information from the Unicode data files using
844the C<Unicode::UCD> module.
845
846=back
847
f6edf83b
JH
848=head1 UNICODE IN OLDER PERLS
849
850If you cannot upgrade your Perl to 5.8.0 or later, you can still
851do some Unicode processing by using the modules C<Unicode::String>,
852C<Unicode::Map8>, and C<Unicode::Map>, available from CPAN.
853If you have the GNU recode installed, you can also use the
376d9008 854Perl front-end C<Convert::Recode> for character conversions.
f6edf83b 855
aaef10c5
JH
856The following are fast conversions from ISO 8859-1 (Latin-1) bytes
857to UTF-8 bytes, the code works even with older Perl 5 versions.
858
859 # ISO 8859-1 to UTF-8
860 s/([\x80-\xFF])/chr(0xC0|ord($1)>>6).chr(0x80|ord($1)&0x3F)/eg;
861
862 # UTF-8 to ISO 8859-1
863 s/([\xC2\xC3])([\x80-\xBF])/chr(ord($1)<<6&0xC0|ord($2)&0x3F)/eg;
864
ba62762e
JH
865=head1 SEE ALSO
866
867L<perlunicode>, L<Encode>, L<encoding>, L<open>, L<utf8>, L<bytes>,
868L<perlretut>, L<Unicode::Collate>, L<Unicode::Normalize>, L<Unicode::UCD>
869
376d9008 870=head1 ACKNOWLEDGMENTS
ba62762e
JH
871
872Thanks to the kind readers of the perl5-porters@perl.org,
873perl-unicode@perl.org, linux-utf8@nl.linux.org, and unicore@unicode.org
874mailing lists for their valuable feedback.
875
876=head1 AUTHOR, COPYRIGHT, AND LICENSE
877
be3c0a43 878Copyright 2001-2002 Jarkko Hietaniemi <jhi@iki.fi>
ba62762e
JH
879
880This document may be distributed under the same terms as Perl itself.