This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
In testargs.t in Test::Harness, don't run a world-writable file.
[perl5.git] / pod / perllocale.pod
CommitLineData
5f05dabc
PP
1=head1 NAME
2
b0c42ed9 3perllocale - Perl locale handling (internationalization and localization)
5f05dabc
PP
4
5=head1 DESCRIPTION
6
5a964f20
TC
7Perl supports language-specific notions of data such as "is this
8a letter", "what is the uppercase equivalent of this letter", and
9"which of these letters comes first". These are important issues,
10especially for languages other than English--but also for English: it
11would be naE<iuml>ve to imagine that C<A-Za-z> defines all the "letters"
12needed to write in English. Perl is also aware that some character other
13than '.' may be preferred as a decimal point, and that output date
14representations may be language-specific. The process of making an
15application take account of its users' preferences in such matters is
16called B<internationalization> (often abbreviated as B<i18n>); telling
17such an application about a particular set of preferences is known as
18B<localization> (B<l10n>).
14280422
DD
19
20Perl can understand language-specific data via the standardized (ISO C,
21XPG4, POSIX 1.c) method called "the locale system". The locale system is
b0c42ed9 22controlled per application using one pragma, one function call, and
14280422
DD
23several environment variables.
24
25B<NOTE>: This feature is new in Perl 5.004, and does not apply unless an
5a964f20 26application specifically requests it--see L<Backward compatibility>.
e38874e2
DD
27The one exception is that write() now B<always> uses the current locale
28- see L<"NOTES">.
5f05dabc
PP
29
30=head1 PREPARING TO USE LOCALES
31
5a964f20 32If Perl applications are to understand and present your data
14280422 33correctly according a locale of your choice, B<all> of the following
5f05dabc
PP
34must be true:
35
36=over 4
37
38=item *
39
40B<Your operating system must support the locale system>. If it does,
14280422 41you should find that the setlocale() function is a documented part of
5f05dabc
PP
42its C library.
43
44=item *
45
5a964f20 46B<Definitions for locales that you use must be installed>. You, or
14280422
DD
47your system administrator, must make sure that this is the case. The
48available locales, the location in which they are kept, and the manner
5a964f20
TC
49in which they are installed all vary from system to system. Some systems
50provide only a few, hard-wired locales and do not allow more to be
51added. Others allow you to add "canned" locales provided by the system
52supplier. Still others allow you or the system administrator to define
14280422 53and add arbitrary locales. (You may have to ask your supplier to
5a964f20 54provide canned locales that are not delivered with your operating
14280422 55system.) Read your system documentation for further illumination.
5f05dabc
PP
56
57=item *
58
59B<Perl must believe that the locale system is supported>. If it does,
60C<perl -V:d_setlocale> will say that the value for C<d_setlocale> is
61C<define>.
62
63=back
64
65If you want a Perl application to process and present your data
66according to a particular locale, the application code should include
2ae324a7 67the S<C<use locale>> pragma (see L<The use locale pragma>) where
5f05dabc
PP
68appropriate, and B<at least one> of the following must be true:
69
70=over 4
71
72=item *
73
14280422 74B<The locale-determining environment variables (see L<"ENVIRONMENT">)
5a964f20
TC
75must be correctly set up> at the time the application is started, either
76by yourself or by whoever set up your system account.
5f05dabc
PP
77
78=item *
79
14280422
DD
80B<The application must set its own locale> using the method described in
81L<The setlocale function>.
5f05dabc
PP
82
83=back
84
85=head1 USING LOCALES
86
87=head2 The use locale pragma
88
14280422
DD
89By default, Perl ignores the current locale. The S<C<use locale>>
90pragma tells Perl to use the current locale for some operations:
5f05dabc
PP
91
92=over 4
93
94=item *
95
14280422
DD
96B<The comparison operators> (C<lt>, C<le>, C<cmp>, C<ge>, and C<gt>) and
97the POSIX string collation functions strcoll() and strxfrm() use
5a964f20
TC
98C<LC_COLLATE>. sort() is also affected if used without an
99explicit comparison function, because it uses C<cmp> by default.
14280422 100
5a964f20 101B<Note:> C<eq> and C<ne> are unaffected by locale: they always
de108802 102perform a char-by-char comparison of their scalar operands. What's
14280422
DD
103more, if C<cmp> finds that its operands are equal according to the
104collation sequence specified by the current locale, it goes on to
de108802
RGS
105perform a char-by-char comparison, and only returns I<0> (equal) if the
106operands are char-for-char identical. If you really want to know whether
5a964f20 107two strings--which C<eq> and C<cmp> may consider different--are equal
14280422
DD
108as far as collation in the locale is concerned, see the discussion in
109L<Category LC_COLLATE: Collation>.
5f05dabc
PP
110
111=item *
112
14280422
DD
113B<Regular expressions and case-modification functions> (uc(), lc(),
114ucfirst(), and lcfirst()) use C<LC_CTYPE>
5f05dabc
PP
115
116=item *
117
903eb63f 118B<Format declarations> (format()) use C<LC_NUMERIC>
5f05dabc
PP
119
120=item *
121
14280422 122B<The POSIX date formatting function> (strftime()) uses C<LC_TIME>.
5f05dabc
PP
123
124=back
125
13a2d996
SP
126C<LC_COLLATE>, C<LC_CTYPE>, and so on, are discussed further in
127L<LOCALE CATEGORIES>.
5f05dabc 128
5a964f20
TC
129The default behavior is restored with the S<C<no locale>> pragma, or
130upon reaching the end of block enclosing C<use locale>.
5f05dabc 131
5a964f20 132The string result of any operation that uses locale
14280422
DD
133information is tainted, as it is possible for a locale to be
134untrustworthy. See L<"SECURITY">.
5f05dabc
PP
135
136=head2 The setlocale function
137
14280422
DD
138You can switch locales as often as you wish at run time with the
139POSIX::setlocale() function:
5f05dabc
PP
140
141 # This functionality not usable prior to Perl 5.004
142 require 5.004;
143
144 # Import locale-handling tool set from POSIX module.
145 # This example uses: setlocale -- the function call
146 # LC_CTYPE -- explained below
147 use POSIX qw(locale_h);
148
14280422 149 # query and save the old locale
5f05dabc
PP
150 $old_locale = setlocale(LC_CTYPE);
151
152 setlocale(LC_CTYPE, "fr_CA.ISO8859-1");
153 # LC_CTYPE now in locale "French, Canada, codeset ISO 8859-1"
154
155 setlocale(LC_CTYPE, "");
156 # LC_CTYPE now reset to default defined by LC_ALL/LC_CTYPE/LANG
157 # environment variables. See below for documentation.
158
159 # restore the old locale
160 setlocale(LC_CTYPE, $old_locale);
161
14280422
DD
162The first argument of setlocale() gives the B<category>, the second the
163B<locale>. The category tells in what aspect of data processing you
164want to apply locale-specific rules. Category names are discussed in
165L<LOCALE CATEGORIES> and L<"ENVIRONMENT">. The locale is the name of a
166collection of customization information corresponding to a particular
167combination of language, country or territory, and codeset. Read on for
168hints on the naming of locales: not all systems name locales as in the
169example.
170
502a173a
JH
171If no second argument is provided and the category is something else
172than LC_ALL, the function returns a string naming the current locale
173for the category. You can use this value as the second argument in a
174subsequent call to setlocale().
175
176If no second argument is provided and the category is LC_ALL, the
177result is implementation-dependent. It may be a string of
178concatenated locales names (separator also implementation-dependent)
f979aebc 179or a single locale name. Please consult your setlocale(3) man page for
502a173a
JH
180details.
181
182If a second argument is given and it corresponds to a valid locale,
183the locale for the category is set to that value, and the function
184returns the now-current locale value. You can then use this in yet
185another call to setlocale(). (In some implementations, the return
186value may sometimes differ from the value you gave as the second
187argument--think of it as an alias for the value you gave.)
5f05dabc
PP
188
189As the example shows, if the second argument is an empty string, the
190category's locale is returned to the default specified by the
191corresponding environment variables. Generally, this results in a
5a964f20 192return to the default that was in force when Perl started up: changes
54310121 193to the environment made by the application after startup may or may not
5a964f20 194be noticed, depending on your system's C library.
5f05dabc 195
14280422
DD
196If the second argument does not correspond to a valid locale, the locale
197for the category is not changed, and the function returns I<undef>.
5f05dabc 198
f979aebc 199For further information about the categories, consult setlocale(3).
3e6e419a
JH
200
201=head2 Finding locales
202
f979aebc 203For locales available in your system, consult also setlocale(3) to
5a964f20
TC
204see whether it leads to the list of available locales (search for the
205I<SEE ALSO> section). If that fails, try the following command lines:
5f05dabc
PP
206
207 locale -a
208
209 nlsinfo
210
211 ls /usr/lib/nls/loc
212
213 ls /usr/lib/locale
214
215 ls /usr/lib/nls
216
b478f28d
JH
217 ls /usr/share/locale
218
5f05dabc
PP
219and see whether they list something resembling these
220
2bdf8add 221 en_US.ISO8859-1 de_DE.ISO8859-1 ru_RU.ISO8859-5
502a173a 222 en_US.iso88591 de_DE.iso88591 ru_RU.iso88595
2bdf8add 223 en_US de_DE ru_RU
14280422 224 en de ru
2bdf8add
JH
225 english german russian
226 english.iso88591 german.iso88591 russian.iso88595
502a173a 227 english.roman8 russian.koi8r
5f05dabc 228
528d65ad
JH
229Sadly, even though the calling interface for setlocale() has been
230standardized, names of locales and the directories where the
5a964f20 231configuration resides have not been. The basic form of the name is
528d65ad
JH
232I<language_territory>B<.>I<codeset>, but the latter parts after
233I<language> are not always present. The I<language> and I<country>
234are usually from the standards B<ISO 3166> and B<ISO 639>, the
235two-letter abbreviations for the countries and the languages of the
236world, respectively. The I<codeset> part often mentions some B<ISO
2378859> character set, the Latin codesets. For example, C<ISO 8859-1>
238is the so-called "Western European codeset" that can be used to encode
239most Western European languages adequately. Again, there are several
240ways to write even the name of that one standard. Lamentably.
5f05dabc 241
14280422
DD
242Two special locales are worth particular mention: "C" and "POSIX".
243Currently these are effectively the same locale: the difference is
5a964f20
TC
244mainly that the first one is defined by the C standard, the second by
245the POSIX standard. They define the B<default locale> in which
14280422 246every program starts in the absence of locale information in its
5a964f20 247environment. (The I<default> default locale, if you will.) Its language
14280422 248is (American) English and its character codeset ASCII.
5f05dabc 249
14280422
DD
250B<NOTE>: Not all systems have the "POSIX" locale (not all systems are
251POSIX-conformant), so use "C" when you need explicitly to specify this
252default locale.
5f05dabc 253
3e6e419a
JH
254=head2 LOCALE PROBLEMS
255
5a964f20 256You may encounter the following warning message at Perl startup:
3e6e419a
JH
257
258 perl: warning: Setting locale failed.
259 perl: warning: Please check that your locale settings:
260 LC_ALL = "En_US",
261 LANG = (unset)
262 are supported and installed on your system.
263 perl: warning: Falling back to the standard locale ("C").
264
5a964f20
TC
265This means that your locale settings had LC_ALL set to "En_US" and
266LANG exists but has no value. Perl tried to believe you but could not.
267Instead, Perl gave up and fell back to the "C" locale, the default locale
268that is supposed to work no matter what. This usually means your locale
269settings were wrong, they mention locales your system has never heard
270of, or the locale installation in your system has problems (for example,
271some system files are broken or missing). There are quick and temporary
272fixes to these problems, as well as more thorough and lasting fixes.
3e6e419a
JH
273
274=head2 Temporarily fixing locale problems
275
5a964f20 276The two quickest fixes are either to render Perl silent about any
3e6e419a
JH
277locale inconsistencies or to run Perl under the default locale "C".
278
279Perl's moaning about locale problems can be silenced by setting the
900bd440
JH
280environment variable PERL_BADLANG to a zero value, for example "0".
281This method really just sweeps the problem under the carpet: you tell
282Perl to shut up even when Perl sees that something is wrong. Do not
283be surprised if later something locale-dependent misbehaves.
3e6e419a
JH
284
285Perl can be run under the "C" locale by setting the environment
5a964f20
TC
286variable LC_ALL to "C". This method is perhaps a bit more civilized
287than the PERL_BADLANG approach, but setting LC_ALL (or
288other locale variables) may affect other programs as well, not just
289Perl. In particular, external programs run from within Perl will see
3e6e419a 290these changes. If you make the new settings permanent (read on), all
f979aebc 291programs you run see the changes. See L<"ENVIRONMENT"> for
5a964f20
TC
292the full list of relevant environment variables and L<USING LOCALES>
293for their effects in Perl. Effects in other programs are
294easily deducible. For example, the variable LC_COLLATE may well affect
b432a672 295your B<sort> program (or whatever the program that arranges "records"
3e6e419a
JH
296alphabetically in your system is called).
297
5a964f20
TC
298You can test out changing these variables temporarily, and if the
299new settings seem to help, put those settings into your shell startup
300files. Consult your local documentation for the exact details. For in
301Bourne-like shells (B<sh>, B<ksh>, B<bash>, B<zsh>):
3e6e419a
JH
302
303 LC_ALL=en_US.ISO8859-1
304 export LC_ALL
305
5a964f20
TC
306This assumes that we saw the locale "en_US.ISO8859-1" using the commands
307discussed above. We decided to try that instead of the above faulty
308locale "En_US"--and in Cshish shells (B<csh>, B<tcsh>)
3e6e419a
JH
309
310 setenv LC_ALL en_US.ISO8859-1
c47ff5f1 311
c406981e
JH
312or if you have the "env" application you can do in any shell
313
314 env LC_ALL=en_US.ISO8859-1 perl ...
315
5a964f20 316If you do not know what shell you have, consult your local
3e6e419a
JH
317helpdesk or the equivalent.
318
319=head2 Permanently fixing locale problems
320
5a964f20
TC
321The slower but superior fixes are when you may be able to yourself
322fix the misconfiguration of your own environment variables. The
3e6e419a
JH
323mis(sing)configuration of the whole system's locales usually requires
324the help of your friendly system administrator.
325
5a964f20
TC
326First, see earlier in this document about L<Finding locales>. That tells
327how to find which locales are really supported--and more importantly,
328installed--on your system. In our example error message, environment
329variables affecting the locale are listed in the order of decreasing
330importance (and unset variables do not matter). Therefore, having
331LC_ALL set to "En_US" must have been the bad choice, as shown by the
332error message. First try fixing locale settings listed first.
3e6e419a 333
5a964f20
TC
334Second, if using the listed commands you see something B<exactly>
335(prefix matches do not count and case usually counts) like "En_US"
336without the quotes, then you should be okay because you are using a
337locale name that should be installed and available in your system.
4a4eefd0 338In this case, see L<Permanently fixing your system's locale configuration>.
3e6e419a 339
4a4eefd0 340=head2 Permanently fixing your system's locale configuration
3e6e419a 341
5a964f20 342This is when you see something like:
3e6e419a
JH
343
344 perl: warning: Please check that your locale settings:
345 LC_ALL = "En_US",
346 LANG = (unset)
347 are supported and installed on your system.
348
349but then cannot see that "En_US" listed by the above-mentioned
5a964f20
TC
350commands. You may see things like "en_US.ISO8859-1", but that isn't
351the same. In this case, try running under a locale
352that you can list and which somehow matches what you tried. The
3e6e419a 353rules for matching locale names are a bit vague because
13a2d996
SP
354standardization is weak in this area. See again the
355L<Finding locales> about general rules.
3e6e419a 356
b687b08b 357=head2 Fixing system locale configuration
3e6e419a 358
5a964f20
TC
359Contact a system administrator (preferably your own) and report the exact
360error message you get, and ask them to read this same documentation you
361are now reading. They should be able to check whether there is something
362wrong with the locale configuration of the system. The L<Finding locales>
363section is unfortunately a bit vague about the exact commands and places
364because these things are not that standardized.
3e6e419a 365
5f05dabc
PP
366=head2 The localeconv function
367
14280422
DD
368The POSIX::localeconv() function allows you to get particulars of the
369locale-dependent numeric formatting information specified by the current
370C<LC_NUMERIC> and C<LC_MONETARY> locales. (If you just want the name of
371the current locale for a particular category, use POSIX::setlocale()
5a964f20 372with a single parameter--see L<The setlocale function>.)
5f05dabc
PP
373
374 use POSIX qw(locale_h);
5f05dabc
PP
375
376 # Get a reference to a hash of locale-dependent info
377 $locale_values = localeconv();
378
379 # Output sorted list of the values
380 for (sort keys %$locale_values) {
14280422 381 printf "%-20s = %s\n", $_, $locale_values->{$_}
5f05dabc
PP
382 }
383
14280422 384localeconv() takes no arguments, and returns B<a reference to> a hash.
5a964f20 385The keys of this hash are variable names for formatting, such as
502a173a 386C<decimal_point> and C<thousands_sep>. The values are the
cea6626f 387corresponding, er, values. See L<POSIX/localeconv> for a longer
502a173a
JH
388example listing the categories an implementation might be expected to
389provide; some provide more and others fewer. You don't need an
390explicit C<use locale>, because localeconv() always observes the
391current locale.
5f05dabc 392
5a964f20
TC
393Here's a simple-minded example program that rewrites its command-line
394parameters as integers correctly formatted in the current locale:
5f05dabc
PP
395
396 # See comments in previous example
397 require 5.004;
398 use POSIX qw(locale_h);
5f05dabc
PP
399
400 # Get some of locale's numeric formatting parameters
401 my ($thousands_sep, $grouping) =
14280422 402 @{localeconv()}{'thousands_sep', 'grouping'};
5f05dabc
PP
403
404 # Apply defaults if values are missing
405 $thousands_sep = ',' unless $thousands_sep;
502a173a
JH
406
407 # grouping and mon_grouping are packed lists
408 # of small integers (characters) telling the
409 # grouping (thousand_seps and mon_thousand_seps
410 # being the group dividers) of numbers and
411 # monetary quantities. The integers' meanings:
412 # 255 means no more grouping, 0 means repeat
413 # the previous grouping, 1-254 means use that
414 # as the current grouping. Grouping goes from
415 # right to left (low to high digits). In the
416 # below we cheat slightly by never using anything
417 # else than the first grouping (whatever that is).
418 if ($grouping) {
419 @grouping = unpack("C*", $grouping);
420 } else {
421 @grouping = (3);
422 }
5f05dabc
PP
423
424 # Format command line params for current locale
14280422
DD
425 for (@ARGV) {
426 $_ = int; # Chop non-integer part
5f05dabc 427 1 while
502a173a 428 s/(\d)(\d{$grouping[0]}($|$thousands_sep))/$1$thousands_sep$2/;
14280422 429 print "$_";
5f05dabc
PP
430 }
431 print "\n";
432
74c76037 433=head2 I18N::Langinfo
4bbcc6e8
JH
434
435Another interface for querying locale-dependent information is the
e1020413 436I18N::Langinfo::langinfo() function, available at least in Unix-like
4bbcc6e8
JH
437systems and VMS.
438
74c76037
JH
439The following example will import the langinfo() function itself and
440three constants to be used as arguments to langinfo(): a constant for
441the abbreviated first day of the week (the numbering starts from
442Sunday = 1) and two more constants for the affirmative and negative
443answers for a yes/no question in the current locale.
4bbcc6e8 444
74c76037 445 use I18N::Langinfo qw(langinfo ABDAY_1 YESSTR NOSTR);
4bbcc6e8 446
74c76037 447 my ($abday_1, $yesstr, $nostr) = map { langinfo } qw(ABDAY_1 YESSTR NOSTR);
4bbcc6e8 448
74c76037 449 print "$abday_1? [$yesstr/$nostr] ";
4bbcc6e8 450
74c76037
JH
451In other words, in the "C" (or English) locale the above will probably
452print something like:
453
454 Sun? [yes/no]
4bbcc6e8
JH
455
456See L<I18N::Langinfo> for more information.
457
5f05dabc
PP
458=head1 LOCALE CATEGORIES
459
5a964f20
TC
460The following subsections describe basic locale categories. Beyond these,
461some combination categories allow manipulation of more than one
462basic category at a time. See L<"ENVIRONMENT"> for a discussion of these.
5f05dabc
PP
463
464=head2 Category LC_COLLATE: Collation
465
5a964f20
TC
466In the scope of S<C<use locale>>, Perl looks to the C<LC_COLLATE>
467environment variable to determine the application's notions on collation
468(ordering) of characters. For example, 'b' follows 'a' in Latin
469alphabets, but where do 'E<aacute>' and 'E<aring>' belong? And while
470'color' follows 'chocolate' in English, what about in Spanish?
5f05dabc 471
60f0fa02
JH
472The following collations all make sense and you may meet any of them
473if you "use locale".
474
475 A B C D E a b c d e
35316ca3 476 A a B b C c D d E e
60f0fa02
JH
477 a A b B c C d D e E
478 a b c d e A B C D E
479
f1cbbd6e 480Here is a code snippet to tell what "word"
5a964f20 481characters are in the current locale, in that locale's order:
5f05dabc
PP
482
483 use locale;
35316ca3 484 print +(sort grep /\w/, map { chr } 0..255), "\n";
5f05dabc 485
14280422
DD
486Compare this with the characters that you see and their order if you
487state explicitly that the locale should be ignored:
5f05dabc
PP
488
489 no locale;
35316ca3 490 print +(sort grep /\w/, map { chr } 0..255), "\n";
5f05dabc
PP
491
492This machine-native collation (which is what you get unless S<C<use
493locale>> has appeared earlier in the same block) must be used for
494sorting raw binary data, whereas the locale-dependent collation of the
b0c42ed9 495first example is useful for natural text.
5f05dabc 496
14280422
DD
497As noted in L<USING LOCALES>, C<cmp> compares according to the current
498collation locale when C<use locale> is in effect, but falls back to a
de108802 499char-by-char comparison for strings that the locale says are equal. You
14280422
DD
500can use POSIX::strcoll() if you don't want this fall-back:
501
502 use POSIX qw(strcoll);
503 $equal_in_locale =
504 !strcoll("space and case ignored", "SpaceAndCaseIgnored");
505
506$equal_in_locale will be true if the collation locale specifies a
5a964f20 507dictionary-like ordering that ignores space characters completely and
9e3a2af8 508which folds case.
14280422 509
5a964f20 510If you have a single string that you want to check for "equality in
14280422
DD
511locale" against several others, you might think you could gain a little
512efficiency by using POSIX::strxfrm() in conjunction with C<eq>:
513
514 use POSIX qw(strxfrm);
515 $xfrm_string = strxfrm("Mixed-case string");
516 print "locale collation ignores spaces\n"
517 if $xfrm_string eq strxfrm("Mixed-casestring");
518 print "locale collation ignores hyphens\n"
519 if $xfrm_string eq strxfrm("Mixedcase string");
520 print "locale collation ignores case\n"
521 if $xfrm_string eq strxfrm("mixed-case string");
522
523strxfrm() takes a string and maps it into a transformed string for use
de108802 524in char-by-char comparisons against other transformed strings during
14280422 525collation. "Under the hood", locale-affected Perl comparison operators
de108802 526call strxfrm() for both operands, then do a char-by-char
5a964f20 527comparison of the transformed strings. By calling strxfrm() explicitly
14280422 528and using a non locale-affected comparison, the example attempts to save
5a964f20 529a couple of transformations. But in fact, it doesn't save anything: Perl
2ae324a7 530magic (see L<perlguts/Magic Variables>) creates the transformed version of a
5a964f20 531string the first time it's needed in a comparison, then keeps this version around
14280422 532in case it's needed again. An example rewritten the easy way with
e38874e2 533C<cmp> runs just about as fast. It also copes with null characters
14280422 534embedded in strings; if you call strxfrm() directly, it treats the first
5a964f20
TC
535null it finds as a terminator. don't expect the transformed strings
536it produces to be portable across systems--or even from one revision
e38874e2
DD
537of your operating system to the next. In short, don't call strxfrm()
538directly: let Perl do it for you.
14280422 539
5a964f20 540Note: C<use locale> isn't shown in some of these examples because it isn't
14280422
DD
541needed: strcoll() and strxfrm() exist only to generate locale-dependent
542results, and so always obey the current C<LC_COLLATE> locale.
5f05dabc
PP
543
544=head2 Category LC_CTYPE: Character Types
545
5a964f20 546In the scope of S<C<use locale>>, Perl obeys the C<LC_CTYPE> locale
14280422
DD
547setting. This controls the application's notion of which characters are
548alphabetic. This affects Perl's C<\w> regular expression metanotation,
f1cbbd6e
GS
549which stands for alphanumeric characters--that is, alphabetic,
550numeric, and including other special characters such as the underscore or
551hyphen. (Consult L<perlre> for more information about
14280422
DD
552regular expressions.) Thanks to C<LC_CTYPE>, depending on your locale
553setting, characters like 'E<aelig>', 'E<eth>', 'E<szlig>', and
554'E<oslash>' may be understood as C<\w> characters.
5f05dabc 555
2c268ad5 556The C<LC_CTYPE> locale also provides the map used in transliterating
68dc0745 557characters between lower and uppercase. This affects the case-mapping
5a964f20
TC
558functions--lc(), lcfirst, uc(), and ucfirst(); case-mapping
559interpolation with C<\l>, C<\L>, C<\u>, or C<\U> in double-quoted strings
560and C<s///> substitutions; and case-independent regular expression
e38874e2
DD
561pattern matching using the C<i> modifier.
562
5a964f20
TC
563Finally, C<LC_CTYPE> affects the POSIX character-class test
564functions--isalpha(), islower(), and so on. For example, if you move
565from the "C" locale to a 7-bit Scandinavian one, you may find--possibly
566to your surprise--that "|" moves from the ispunct() class to isalpha().
5f05dabc 567
14280422
DD
568B<Note:> A broken or malicious C<LC_CTYPE> locale definition may result
569in clearly ineligible characters being considered to be alphanumeric by
5a964f20
TC
570your application. For strict matching of (mundane) letters and
571digits--for example, in command strings--locale-aware applications
14280422 572should use C<\w> inside a C<no locale> block. See L<"SECURITY">.
5f05dabc
PP
573
574=head2 Category LC_NUMERIC: Numeric Formatting
575
2095dafa
RGS
576After a proper POSIX::setlocale() call, Perl obeys the C<LC_NUMERIC>
577locale information, which controls an application's idea of how numbers
578should be formatted for human readability by the printf(), sprintf(), and
579write() functions. String-to-numeric conversion by the POSIX::strtod()
5a964f20
TC
580function is also affected. In most implementations the only effect is to
581change the character used for the decimal point--perhaps from '.' to ','.
582These functions aren't aware of such niceties as thousands separation and
2095dafa 583so on. (See L<The localeconv function> if you care about these things.)
5a964f20 584
3cf03d68 585Output produced by print() is also affected by the current locale: it
3cf03d68
JH
586corresponds to what you'd get from printf() in the "C" locale. The
587same is true for Perl's internal conversions between numeric and
588string formats:
5f05dabc 589
2095dafa
RGS
590 use POSIX qw(strtod setlocale LC_NUMERIC);
591
592 setlocale LC_NUMERIC, "";
14280422 593
5f05dabc
PP
594 $n = 5/2; # Assign numeric 2.5 to $n
595
35316ca3 596 $a = " $n"; # Locale-dependent conversion to string
5f05dabc 597
35316ca3 598 print "half five is $n\n"; # Locale-dependent output
5f05dabc
PP
599
600 printf "half five is %g\n", $n; # Locale-dependent output
601
14280422
DD
602 print "DECIMAL POINT IS COMMA\n"
603 if $n == (strtod("2,5"))[0]; # Locale-dependent conversion
5f05dabc 604
4bbcc6e8
JH
605See also L<I18N::Langinfo> and C<RADIXCHAR>.
606
5f05dabc
PP
607=head2 Category LC_MONETARY: Formatting of monetary amounts
608
5a964f20
TC
609The C standard defines the C<LC_MONETARY> category, but no function
610that is affected by its contents. (Those with experience of standards
b0c42ed9 611committees will recognize that the working group decided to punt on the
14280422 612issue.) Consequently, Perl takes no notice of it. If you really want
13a2d996
SP
613to use C<LC_MONETARY>, you can query its contents--see
614L<The localeconv function>--and use the information that it returns in your
615application's own formatting of currency amounts. However, you may well
616find that the information, voluminous and complex though it may be, still
617does not quite meet your requirements: currency formatting is a hard nut
618to crack.
5f05dabc 619
4bbcc6e8
JH
620See also L<I18N::Langinfo> and C<CRNCYSTR>.
621
5f05dabc
PP
622=head2 LC_TIME
623
5a964f20 624Output produced by POSIX::strftime(), which builds a formatted
5f05dabc
PP
625human-readable date/time string, is affected by the current C<LC_TIME>
626locale. Thus, in a French locale, the output produced by the C<%B>
627format element (full month name) for the first month of the year would
5a964f20 628be "janvier". Here's how to get a list of long month names in the
5f05dabc
PP
629current locale:
630
631 use POSIX qw(strftime);
14280422
DD
632 for (0..11) {
633 $long_month_name[$_] =
634 strftime("%B", 0, 0, 0, 1, $_, 96);
5f05dabc
PP
635 }
636
5a964f20 637Note: C<use locale> isn't needed in this example: as a function that
14280422
DD
638exists only to generate locale-dependent results, strftime() always
639obeys the current C<LC_TIME> locale.
5f05dabc 640
4bbcc6e8 641See also L<I18N::Langinfo> and C<ABDAY_1>..C<ABDAY_7>, C<DAY_1>..C<DAY_7>,
2a2bf5f4 642C<ABMON_1>..C<ABMON_12>, and C<ABMON_1>..C<ABMON_12>.
4bbcc6e8 643
5f05dabc
PP
644=head2 Other categories
645
5a964f20
TC
646The remaining locale category, C<LC_MESSAGES> (possibly supplemented
647by others in particular implementations) is not currently used by
98a6f11e 648Perl--except possibly to affect the behavior of library functions
649called by extensions outside the standard Perl distribution and by the
650operating system and its utilities. Note especially that the string
651value of C<$!> and the error messages given by external utilities may
652be changed by C<LC_MESSAGES>. If you want to have portable error
265f5c4a 653codes, use C<%!>. See L<Errno>.
14280422
DD
654
655=head1 SECURITY
656
5a964f20 657Although the main discussion of Perl security issues can be found in
14280422
DD
658L<perlsec>, a discussion of Perl's locale handling would be incomplete
659if it did not draw your attention to locale-dependent security issues.
5a964f20
TC
660Locales--particularly on systems that allow unprivileged users to
661build their own locales--are untrustworthy. A malicious (or just plain
14280422
DD
662broken) locale can make a locale-aware application give unexpected
663results. Here are a few possibilities:
664
665=over 4
666
667=item *
668
669Regular expression checks for safe file names or mail addresses using
5a964f20 670C<\w> may be spoofed by an C<LC_CTYPE> locale that claims that
14280422
DD
671characters such as "E<gt>" and "|" are alphanumeric.
672
673=item *
674
e38874e2
DD
675String interpolation with case-mapping, as in, say, C<$dest =
676"C:\U$name.$ext">, may produce dangerous results if a bogus LC_CTYPE
677case-mapping table is in effect.
678
679=item *
680
14280422
DD
681A sneaky C<LC_COLLATE> locale could result in the names of students with
682"D" grades appearing ahead of those with "A"s.
683
684=item *
685
5a964f20 686An application that takes the trouble to use information in
14280422 687C<LC_MONETARY> may format debits as if they were credits and vice versa
5a964f20 688if that locale has been subverted. Or it might make payments in US
14280422
DD
689dollars instead of Hong Kong dollars.
690
691=item *
692
693The date and day names in dates formatted by strftime() could be
694manipulated to advantage by a malicious user able to subvert the
5a964f20 695C<LC_DATE> locale. ("Look--it says I wasn't in the building on
14280422
DD
696Sunday.")
697
698=back
699
700Such dangers are not peculiar to the locale system: any aspect of an
5a964f20 701application's environment which may be modified maliciously presents
14280422 702similar challenges. Similarly, they are not specific to Perl: any
5a964f20 703programming language that allows you to write programs that take
14280422
DD
704account of their environment exposes you to these issues.
705
5a964f20
TC
706Perl cannot protect you from all possibilities shown in the
707examples--there is no substitute for your own vigilance--but, when
14280422 708C<use locale> is in effect, Perl uses the tainting mechanism (see
5a964f20 709L<perlsec>) to mark string results that become locale-dependent, and
14280422 710which may be untrustworthy in consequence. Here is a summary of the
5a964f20 711tainting behavior of operators and functions that may be affected by
14280422
DD
712the locale:
713
714=over 4
715
551e1d92
RB
716=item *
717
718B<Comparison operators> (C<lt>, C<le>, C<ge>, C<gt> and C<cmp>):
14280422
DD
719
720Scalar true/false (or less/equal/greater) result is never tainted.
721
551e1d92
RB
722=item *
723
724B<Case-mapping interpolation> (with C<\l>, C<\L>, C<\u> or C<\U>)
e38874e2
DD
725
726Result string containing interpolated material is tainted if
727C<use locale> is in effect.
728
551e1d92
RB
729=item *
730
731B<Matching operator> (C<m//>):
14280422
DD
732
733Scalar true/false result never tainted.
734
5a964f20 735Subpatterns, either delivered as a list-context result or as $1 etc.
14280422 736are tainted if C<use locale> is in effect, and the subpattern regular
e38874e2 737expression contains C<\w> (to match an alphanumeric character), C<\W>
6b0ac556
OK
738(non-alphanumeric character), C<\s> (whitespace character), or C<\S>
739(non whitespace character). The matched-pattern variable, $&, $`
e38874e2
DD
740(pre-match), $' (post-match), and $+ (last match) are also tainted if
741C<use locale> is in effect and the regular expression contains C<\w>,
742C<\W>, C<\s>, or C<\S>.
14280422 743
551e1d92
RB
744=item *
745
746B<Substitution operator> (C<s///>):
14280422 747
e38874e2 748Has the same behavior as the match operator. Also, the left
5a964f20
TC
749operand of C<=~> becomes tainted when C<use locale> in effect
750if modified as a result of a substitution based on a regular
e38874e2 751expression match involving C<\w>, C<\W>, C<\s>, or C<\S>; or of
7b8d334a 752case-mapping with C<\l>, C<\L>,C<\u> or C<\U>.
14280422 753
551e1d92
RB
754=item *
755
756B<Output formatting functions> (printf() and write()):
14280422 757
3cf03d68
JH
758Results are never tainted because otherwise even output from print,
759for example C<print(1/7)>, should be tainted if C<use locale> is in
760effect.
14280422 761
551e1d92
RB
762=item *
763
764B<Case-mapping functions> (lc(), lcfirst(), uc(), ucfirst()):
14280422
DD
765
766Results are tainted if C<use locale> is in effect.
767
551e1d92
RB
768=item *
769
770B<POSIX locale-dependent functions> (localeconv(), strcoll(),
14280422
DD
771strftime(), strxfrm()):
772
773Results are never tainted.
774
551e1d92
RB
775=item *
776
777B<POSIX character class tests> (isalnum(), isalpha(), isdigit(),
14280422
DD
778isgraph(), islower(), isprint(), ispunct(), isspace(), isupper(),
779isxdigit()):
780
781True/false results are never tainted.
782
783=back
784
785Three examples illustrate locale-dependent tainting.
786The first program, which ignores its locale, won't run: a value taken
54310121 787directly from the command line may not be used to name an output file
14280422
DD
788when taint checks are enabled.
789
790 #/usr/local/bin/perl -T
791 # Run with taint checking
792
54310121 793 # Command line sanity check omitted...
14280422
DD
794 $tainted_output_file = shift;
795
796 open(F, ">$tainted_output_file")
797 or warn "Open of $untainted_output_file failed: $!\n";
798
799The program can be made to run by "laundering" the tainted value through
5a964f20
TC
800a regular expression: the second example--which still ignores locale
801information--runs, creating the file named on its command line
14280422
DD
802if it can.
803
804 #/usr/local/bin/perl -T
805
806 $tainted_output_file = shift;
807 $tainted_output_file =~ m%[\w/]+%;
808 $untainted_output_file = $&;
809
810 open(F, ">$untainted_output_file")
811 or warn "Open of $untainted_output_file failed: $!\n";
812
5a964f20 813Compare this with a similar but locale-aware program:
14280422
DD
814
815 #/usr/local/bin/perl -T
816
817 $tainted_output_file = shift;
818 use locale;
819 $tainted_output_file =~ m%[\w/]+%;
820 $localized_output_file = $&;
821
822 open(F, ">$localized_output_file")
823 or warn "Open of $localized_output_file failed: $!\n";
824
825This third program fails to run because $& is tainted: it is the result
5a964f20 826of a match involving C<\w> while C<use locale> is in effect.
5f05dabc
PP
827
828=head1 ENVIRONMENT
829
830=over 12
831
832=item PERL_BADLANG
833
14280422 834A string that can suppress Perl's warning about failed locale settings
54310121 835at startup. Failure can occur if the locale support in the operating
5a964f20 836system is lacking (broken) in some way--or if you mistyped the name of
900bd440
JH
837a locale when you set up your environment. If this environment
838variable is absent, or has a value that does not evaluate to integer
839zero--that is, "0" or ""-- Perl will complain about locale setting
840failures.
5f05dabc 841
14280422
DD
842B<NOTE>: PERL_BADLANG only gives you a way to hide the warning message.
843The message tells about some problem in your system's locale support,
844and you should investigate what the problem is.
5f05dabc
PP
845
846=back
847
848The following environment variables are not specific to Perl: They are
14280422
DD
849part of the standardized (ISO C, XPG4, POSIX 1.c) setlocale() method
850for controlling an application's opinion on data.
5f05dabc
PP
851
852=over 12
853
854=item LC_ALL
855
5a964f20 856C<LC_ALL> is the "override-all" locale environment variable. If
5f05dabc
PP
857set, it overrides all the rest of the locale environment variables.
858
528d65ad
JH
859=item LANGUAGE
860
861B<NOTE>: C<LANGUAGE> is a GNU extension, it affects you only if you
862are using the GNU libc. This is the case if you are using e.g. Linux.
e1020413 863If you are using "commercial" Unixes you are most probably I<not>
22b6f60d
JH
864using GNU libc and you can ignore C<LANGUAGE>.
865
866However, in the case you are using C<LANGUAGE>: it affects the
867language of informational, warning, and error messages output by
868commands (in other words, it's like C<LC_MESSAGES>) but it has higher
96090e4f 869priority than C<LC_ALL>. Moreover, it's not a single value but
22b6f60d
JH
870instead a "path" (":"-separated list) of I<languages> (not locales).
871See the GNU C<gettext> library documentation for more information.
528d65ad 872
5f05dabc
PP
873=item LC_CTYPE
874
875In the absence of C<LC_ALL>, C<LC_CTYPE> chooses the character type
876locale. In the absence of both C<LC_ALL> and C<LC_CTYPE>, C<LANG>
877chooses the character type locale.
878
879=item LC_COLLATE
880
14280422
DD
881In the absence of C<LC_ALL>, C<LC_COLLATE> chooses the collation
882(sorting) locale. In the absence of both C<LC_ALL> and C<LC_COLLATE>,
883C<LANG> chooses the collation locale.
5f05dabc
PP
884
885=item LC_MONETARY
886
14280422
DD
887In the absence of C<LC_ALL>, C<LC_MONETARY> chooses the monetary
888formatting locale. In the absence of both C<LC_ALL> and C<LC_MONETARY>,
889C<LANG> chooses the monetary formatting locale.
5f05dabc
PP
890
891=item LC_NUMERIC
892
893In the absence of C<LC_ALL>, C<LC_NUMERIC> chooses the numeric format
894locale. In the absence of both C<LC_ALL> and C<LC_NUMERIC>, C<LANG>
895chooses the numeric format.
896
897=item LC_TIME
898
14280422
DD
899In the absence of C<LC_ALL>, C<LC_TIME> chooses the date and time
900formatting locale. In the absence of both C<LC_ALL> and C<LC_TIME>,
901C<LANG> chooses the date and time formatting locale.
5f05dabc
PP
902
903=item LANG
904
14280422
DD
905C<LANG> is the "catch-all" locale environment variable. If it is set, it
906is used as the last resort after the overall C<LC_ALL> and the
5f05dabc
PP
907category-specific C<LC_...>.
908
909=back
910
7e4353e9
RGS
911=head2 Examples
912
913The LC_NUMERIC controls the numeric output:
914
915 use locale;
916 use POSIX qw(locale_h); # Imports setlocale() and the LC_ constants.
917 setlocale(LC_NUMERIC, "fr_FR") or die "Pardon";
918 printf "%g\n", 1.23; # If the "fr_FR" succeeded, probably shows 1,23.
919
920and also how strings are parsed by POSIX::strtod() as numbers:
921
922 use locale;
923 use POSIX qw(locale_h strtod);
2095dafa 924 setlocale(LC_NUMERIC, "de_DE") or die "Entschuldigung";
7e4353e9
RGS
925 my $x = strtod("2,34") + 5;
926 print $x, "\n"; # Probably shows 7,34.
927
5f05dabc
PP
928=head1 NOTES
929
930=head2 Backward compatibility
931
b0c42ed9 932Versions of Perl prior to 5.004 B<mostly> ignored locale information,
5a964f20
TC
933generally behaving as if something similar to the C<"C"> locale were
934always in force, even if the program environment suggested otherwise
935(see L<The setlocale function>). By default, Perl still behaves this
936way for backward compatibility. If you want a Perl application to pay
937attention to locale information, you B<must> use the S<C<use locale>>
b687b08b 938pragma (see L<The use locale pragma>) to instruct it to do so.
b0c42ed9
JH
939
940Versions of Perl from 5.002 to 5.003 did use the C<LC_CTYPE>
5a964f20
TC
941information if available; that is, C<\w> did understand what
942were the letters according to the locale environment variables.
b0c42ed9
JH
943The problem was that the user had no control over the feature:
944if the C library supported locales, Perl used them.
945
946=head2 I18N:Collate obsolete
947
5a964f20 948In versions of Perl prior to 5.004, per-locale collation was possible
b0c42ed9
JH
949using the C<I18N::Collate> library module. This module is now mildly
950obsolete and should be avoided in new applications. The C<LC_COLLATE>
951functionality is now integrated into the Perl core language: One can
952use locale-specific scalar data completely normally with C<use locale>,
953so there is no longer any need to juggle with the scalar references of
954C<I18N::Collate>.
5f05dabc 955
14280422 956=head2 Sort speed and memory use impacts
5f05dabc
PP
957
958Comparing and sorting by locale is usually slower than the default
14280422
DD
959sorting; slow-downs of two to four times have been observed. It will
960also consume more memory: once a Perl scalar variable has participated
961in any string comparison or sorting operation obeying the locale
962collation rules, it will take 3-15 times more memory than before. (The
963exact multiplier depends on the string's contents, the operating system
964and the locale.) These downsides are dictated more by the operating
965system's implementation of the locale system than by Perl.
5f05dabc 966
e38874e2
DD
967=head2 write() and LC_NUMERIC
968
903eb63f
NT
969If a program's environment specifies an LC_NUMERIC locale and C<use
970locale> is in effect when the format is declared, the locale is used
971to specify the decimal point character in formatted output. Formatted
972output cannot be controlled by C<use locale> at the time when write()
973is called.
e38874e2 974
5f05dabc
PP
975=head2 Freely available locale definitions
976
08d7a6b2
LB
977There is a large collection of locale definitions at:
978
979 http://std.dkuug.dk/i18n/WG15-collection/locales/
980
981You should be aware that it is
14280422 982unsupported, and is not claimed to be fit for any purpose. If your
5a964f20 983system allows installation of arbitrary locales, you may find the
14280422
DD
984definitions useful as they are, or as a basis for the development of
985your own locales.
5f05dabc 986
14280422 987=head2 I18n and l10n
5f05dabc 988
b0c42ed9
JH
989"Internationalization" is often abbreviated as B<i18n> because its first
990and last letters are separated by eighteen others. (You may guess why
991the internalin ... internaliti ... i18n tends to get abbreviated.) In
992the same way, "localization" is often abbreviated to B<l10n>.
14280422
DD
993
994=head2 An imperfect standard
995
996Internationalization, as defined in the C and POSIX standards, can be
997criticized as incomplete, ungainly, and having too large a granularity.
998(Locales apply to a whole process, when it would arguably be more useful
999to have them apply to a single thread, window group, or whatever.) They
1000also have a tendency, like standards groups, to divide the world into
1001nations, when we all know that the world can equally well be divided
1002into bankers, bikers, gamers, and so on. But, for now, it's the only
1003standard we've got. This may be construed as a bug.
5f05dabc 1004
b310b053
JH
1005=head1 Unicode and UTF-8
1006
1007The support of Unicode is new starting from Perl version 5.6, and
1008more fully implemented in the version 5.8. See L<perluniintro> and
1009L<perlunicode> for more details.
1010
1011Usually locale settings and Unicode do not affect each other, but
1012there are exceptions, see L<perlunicode/Locales> for examples.
1013
5f05dabc
PP
1014=head1 BUGS
1015
1016=head2 Broken systems
1017
5a964f20 1018In certain systems, the operating system's locale support
2bdf8add
JH
1019is broken and cannot be fixed or used by Perl. Such deficiencies can
1020and will result in mysterious hangs and/or Perl core dumps when the
1021C<use locale> is in effect. When confronted with such a system,
7f2de2d2 1022please report in excruciating detail to <F<perlbug@perl.org>>, and
5a964f20 1023complain to your vendor: bug fixes may exist for these problems
2bdf8add
JH
1024in your operating system. Sometimes such bug fixes are called an
1025operating system upgrade.
5f05dabc
PP
1026
1027=head1 SEE ALSO
1028
b310b053
JH
1029L<I18N::Langinfo>, L<perluniintro>, L<perlunicode>, L<open>,
1030L<POSIX/isalnum>, L<POSIX/isalpha>,
4bbcc6e8
JH
1031L<POSIX/isdigit>, L<POSIX/isgraph>, L<POSIX/islower>,
1032L<POSIX/isprint>, L<POSIX/ispunct>, L<POSIX/isspace>,
1033L<POSIX/isupper>, L<POSIX/isxdigit>, L<POSIX/localeconv>,
1034L<POSIX/setlocale>, L<POSIX/strcoll>, L<POSIX/strftime>,
1035L<POSIX/strtod>, L<POSIX/strxfrm>.
5f05dabc
PP
1036
1037=head1 HISTORY
1038
b0c42ed9 1039Jarkko Hietaniemi's original F<perli18n.pod> heavily hacked by Dominic
5a964f20
TC
1040Dunlop, assisted by the perl5-porters. Prose worked over a bit by
1041Tom Christiansen.
5f05dabc 1042
5a964f20 1043Last update: Thu Jun 11 08:44:13 MDT 1998