This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Document all the control flow keywords in perlfunc
[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
14280422
DD
102perform a byte-by-byte comparison of their scalar operands. What's
103more, if C<cmp> finds that its operands are equal according to the
104collation sequence specified by the current locale, it goes on to
105perform a byte-by-byte comparison, and only returns I<0> (equal) if the
106operands are bit-for-bit 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
14280422 118B<The formatting functions> (printf(), sprintf() and write()) use
5f05dabc
PP
119C<LC_NUMERIC>
120
121=item *
122
14280422 123B<The POSIX date formatting function> (strftime()) uses C<LC_TIME>.
5f05dabc
PP
124
125=back
126
14280422
DD
127C<LC_COLLATE>, C<LC_CTYPE>, and so on, are discussed further in L<LOCALE
128CATEGORIES>.
5f05dabc 129
5a964f20
TC
130The default behavior is restored with the S<C<no locale>> pragma, or
131upon reaching the end of block enclosing C<use locale>.
5f05dabc 132
5a964f20 133The string result of any operation that uses locale
14280422
DD
134information is tainted, as it is possible for a locale to be
135untrustworthy. See L<"SECURITY">.
5f05dabc
PP
136
137=head2 The setlocale function
138
14280422
DD
139You can switch locales as often as you wish at run time with the
140POSIX::setlocale() function:
5f05dabc
PP
141
142 # This functionality not usable prior to Perl 5.004
143 require 5.004;
144
145 # Import locale-handling tool set from POSIX module.
146 # This example uses: setlocale -- the function call
147 # LC_CTYPE -- explained below
148 use POSIX qw(locale_h);
149
14280422 150 # query and save the old locale
5f05dabc
PP
151 $old_locale = setlocale(LC_CTYPE);
152
153 setlocale(LC_CTYPE, "fr_CA.ISO8859-1");
154 # LC_CTYPE now in locale "French, Canada, codeset ISO 8859-1"
155
156 setlocale(LC_CTYPE, "");
157 # LC_CTYPE now reset to default defined by LC_ALL/LC_CTYPE/LANG
158 # environment variables. See below for documentation.
159
160 # restore the old locale
161 setlocale(LC_CTYPE, $old_locale);
162
14280422
DD
163The first argument of setlocale() gives the B<category>, the second the
164B<locale>. The category tells in what aspect of data processing you
165want to apply locale-specific rules. Category names are discussed in
166L<LOCALE CATEGORIES> and L<"ENVIRONMENT">. The locale is the name of a
167collection of customization information corresponding to a particular
168combination of language, country or territory, and codeset. Read on for
169hints on the naming of locales: not all systems name locales as in the
170example.
171
502a173a
JH
172If no second argument is provided and the category is something else
173than LC_ALL, the function returns a string naming the current locale
174for the category. You can use this value as the second argument in a
175subsequent call to setlocale().
176
177If no second argument is provided and the category is LC_ALL, the
178result is implementation-dependent. It may be a string of
179concatenated locales names (separator also implementation-dependent)
180or a single locale name. Please consult your L<setlocale(3)> for
181details.
182
183If a second argument is given and it corresponds to a valid locale,
184the locale for the category is set to that value, and the function
185returns the now-current locale value. You can then use this in yet
186another call to setlocale(). (In some implementations, the return
187value may sometimes differ from the value you gave as the second
188argument--think of it as an alias for the value you gave.)
5f05dabc
PP
189
190As the example shows, if the second argument is an empty string, the
191category's locale is returned to the default specified by the
192corresponding environment variables. Generally, this results in a
5a964f20 193return to the default that was in force when Perl started up: changes
54310121 194to the environment made by the application after startup may or may not
5a964f20 195be noticed, depending on your system's C library.
5f05dabc 196
14280422
DD
197If the second argument does not correspond to a valid locale, the locale
198for the category is not changed, and the function returns I<undef>.
5f05dabc 199
14280422 200For further information about the categories, consult L<setlocale(3)>.
3e6e419a
JH
201
202=head2 Finding locales
203
5a964f20
TC
204For locales available in your system, consult also L<setlocale(3)> to
205see whether it leads to the list of available locales (search for the
206I<SEE ALSO> section). If that fails, try the following command lines:
5f05dabc
PP
207
208 locale -a
209
210 nlsinfo
211
212 ls /usr/lib/nls/loc
213
214 ls /usr/lib/locale
215
216 ls /usr/lib/nls
217
218and see whether they list something resembling these
219
2bdf8add 220 en_US.ISO8859-1 de_DE.ISO8859-1 ru_RU.ISO8859-5
502a173a 221 en_US.iso88591 de_DE.iso88591 ru_RU.iso88595
2bdf8add 222 en_US de_DE ru_RU
14280422 223 en de ru
2bdf8add
JH
224 english german russian
225 english.iso88591 german.iso88591 russian.iso88595
502a173a 226 english.roman8 russian.koi8r
5f05dabc 227
5a964f20
TC
228Sadly, even though the calling interface for setlocale() has
229been standardized, names of locales and the directories where the
230configuration resides have not been. The basic form of the name is
231I<language_country/territory>B<.>I<codeset>, but the latter parts after
232I<language> are not always present. The I<language> and I<country> are
233usually from the standards B<ISO 3166> and B<ISO 639>, the two-letter
234abbreviations for the countries and the languages of the world,
235respectively. The I<codeset> part often mentions some B<ISO 8859>
236character set, the Latin codesets. For example, C<ISO 8859-1> is the
237so-called "Western codeset" that can be used to encode most Western
238European languages. Again, there are several ways to write even the
239name of that one standard. Lamentably.
5f05dabc 240
14280422
DD
241Two special locales are worth particular mention: "C" and "POSIX".
242Currently these are effectively the same locale: the difference is
5a964f20
TC
243mainly that the first one is defined by the C standard, the second by
244the POSIX standard. They define the B<default locale> in which
14280422 245every program starts in the absence of locale information in its
5a964f20 246environment. (The I<default> default locale, if you will.) Its language
14280422 247is (American) English and its character codeset ASCII.
5f05dabc 248
14280422
DD
249B<NOTE>: Not all systems have the "POSIX" locale (not all systems are
250POSIX-conformant), so use "C" when you need explicitly to specify this
251default locale.
5f05dabc 252
3e6e419a
JH
253=head2 LOCALE PROBLEMS
254
5a964f20 255You may encounter the following warning message at Perl startup:
3e6e419a
JH
256
257 perl: warning: Setting locale failed.
258 perl: warning: Please check that your locale settings:
259 LC_ALL = "En_US",
260 LANG = (unset)
261 are supported and installed on your system.
262 perl: warning: Falling back to the standard locale ("C").
263
5a964f20
TC
264This means that your locale settings had LC_ALL set to "En_US" and
265LANG exists but has no value. Perl tried to believe you but could not.
266Instead, Perl gave up and fell back to the "C" locale, the default locale
267that is supposed to work no matter what. This usually means your locale
268settings were wrong, they mention locales your system has never heard
269of, or the locale installation in your system has problems (for example,
270some system files are broken or missing). There are quick and temporary
271fixes to these problems, as well as more thorough and lasting fixes.
3e6e419a
JH
272
273=head2 Temporarily fixing locale problems
274
5a964f20 275The two quickest fixes are either to render Perl silent about any
3e6e419a
JH
276locale inconsistencies or to run Perl under the default locale "C".
277
278Perl's moaning about locale problems can be silenced by setting the
279environment variable PERL_BADLANG to a non-zero value, for example
280"1". This method really just sweeps the problem under the carpet: you
281tell Perl to shut up even when Perl sees that something is wrong. Do
5a964f20 282not be surprised if later something locale-dependent misbehaves.
3e6e419a
JH
283
284Perl can be run under the "C" locale by setting the environment
5a964f20
TC
285variable LC_ALL to "C". This method is perhaps a bit more civilized
286than the PERL_BADLANG approach, but setting LC_ALL (or
287other locale variables) may affect other programs as well, not just
288Perl. In particular, external programs run from within Perl will see
3e6e419a 289these changes. If you make the new settings permanent (read on), all
5a964f20
TC
290programs you run see the changes. See L<ENVIRONMENT> for for
291the full list of relevant environment variables and L<USING LOCALES>
292for their effects in Perl. Effects in other programs are
293easily deducible. For example, the variable LC_COLLATE may well affect
294your B<sort> program (or whatever the program that arranges `records'
3e6e419a
JH
295alphabetically in your system is called).
296
5a964f20
TC
297You can test out changing these variables temporarily, and if the
298new settings seem to help, put those settings into your shell startup
299files. Consult your local documentation for the exact details. For in
300Bourne-like shells (B<sh>, B<ksh>, B<bash>, B<zsh>):
3e6e419a
JH
301
302 LC_ALL=en_US.ISO8859-1
303 export LC_ALL
304
5a964f20
TC
305This assumes that we saw the locale "en_US.ISO8859-1" using the commands
306discussed above. We decided to try that instead of the above faulty
307locale "En_US"--and in Cshish shells (B<csh>, B<tcsh>)
3e6e419a
JH
308
309 setenv LC_ALL en_US.ISO8859-1
310
5a964f20 311If you do not know what shell you have, consult your local
3e6e419a
JH
312helpdesk or the equivalent.
313
314=head2 Permanently fixing locale problems
315
5a964f20
TC
316The slower but superior fixes are when you may be able to yourself
317fix the misconfiguration of your own environment variables. The
3e6e419a
JH
318mis(sing)configuration of the whole system's locales usually requires
319the help of your friendly system administrator.
320
5a964f20
TC
321First, see earlier in this document about L<Finding locales>. That tells
322how to find which locales are really supported--and more importantly,
323installed--on your system. In our example error message, environment
324variables affecting the locale are listed in the order of decreasing
325importance (and unset variables do not matter). Therefore, having
326LC_ALL set to "En_US" must have been the bad choice, as shown by the
327error message. First try fixing locale settings listed first.
3e6e419a 328
5a964f20
TC
329Second, if using the listed commands you see something B<exactly>
330(prefix matches do not count and case usually counts) like "En_US"
331without the quotes, then you should be okay because you are using a
332locale name that should be installed and available in your system.
333In this case, see L<Fixing system locale configuration>.
3e6e419a 334
5a964f20 335=head2 Permanently fixing your locale configuration
3e6e419a 336
5a964f20 337This is when you see something like:
3e6e419a
JH
338
339 perl: warning: Please check that your locale settings:
340 LC_ALL = "En_US",
341 LANG = (unset)
342 are supported and installed on your system.
343
344but then cannot see that "En_US" listed by the above-mentioned
5a964f20
TC
345commands. You may see things like "en_US.ISO8859-1", but that isn't
346the same. In this case, try running under a locale
347that you can list and which somehow matches what you tried. The
3e6e419a 348rules for matching locale names are a bit vague because
5a964f20
TC
349standardization is weak in this area. See again the L<Finding
350locales> about general rules.
3e6e419a 351
5a964f20 352=head2 Permanently fixing system locale configuration
3e6e419a 353
5a964f20
TC
354Contact a system administrator (preferably your own) and report the exact
355error message you get, and ask them to read this same documentation you
356are now reading. They should be able to check whether there is something
357wrong with the locale configuration of the system. The L<Finding locales>
358section is unfortunately a bit vague about the exact commands and places
359because these things are not that standardized.
3e6e419a 360
5f05dabc
PP
361=head2 The localeconv function
362
14280422
DD
363The POSIX::localeconv() function allows you to get particulars of the
364locale-dependent numeric formatting information specified by the current
365C<LC_NUMERIC> and C<LC_MONETARY> locales. (If you just want the name of
366the current locale for a particular category, use POSIX::setlocale()
5a964f20 367with a single parameter--see L<The setlocale function>.)
5f05dabc
PP
368
369 use POSIX qw(locale_h);
5f05dabc
PP
370
371 # Get a reference to a hash of locale-dependent info
372 $locale_values = localeconv();
373
374 # Output sorted list of the values
375 for (sort keys %$locale_values) {
14280422 376 printf "%-20s = %s\n", $_, $locale_values->{$_}
5f05dabc
PP
377 }
378
14280422 379localeconv() takes no arguments, and returns B<a reference to> a hash.
5a964f20 380The keys of this hash are variable names for formatting, such as
502a173a
JH
381C<decimal_point> and C<thousands_sep>. The values are the
382corresponding, er, values. See L<POSIX (3)/localeconv> for a longer
383example listing the categories an implementation might be expected to
384provide; some provide more and others fewer. You don't need an
385explicit C<use locale>, because localeconv() always observes the
386current locale.
5f05dabc 387
5a964f20
TC
388Here's a simple-minded example program that rewrites its command-line
389parameters as integers correctly formatted in the current locale:
5f05dabc
PP
390
391 # See comments in previous example
392 require 5.004;
393 use POSIX qw(locale_h);
5f05dabc
PP
394
395 # Get some of locale's numeric formatting parameters
396 my ($thousands_sep, $grouping) =
14280422 397 @{localeconv()}{'thousands_sep', 'grouping'};
5f05dabc
PP
398
399 # Apply defaults if values are missing
400 $thousands_sep = ',' unless $thousands_sep;
502a173a
JH
401
402 # grouping and mon_grouping are packed lists
403 # of small integers (characters) telling the
404 # grouping (thousand_seps and mon_thousand_seps
405 # being the group dividers) of numbers and
406 # monetary quantities. The integers' meanings:
407 # 255 means no more grouping, 0 means repeat
408 # the previous grouping, 1-254 means use that
409 # as the current grouping. Grouping goes from
410 # right to left (low to high digits). In the
411 # below we cheat slightly by never using anything
412 # else than the first grouping (whatever that is).
413 if ($grouping) {
414 @grouping = unpack("C*", $grouping);
415 } else {
416 @grouping = (3);
417 }
5f05dabc
PP
418
419 # Format command line params for current locale
14280422
DD
420 for (@ARGV) {
421 $_ = int; # Chop non-integer part
5f05dabc 422 1 while
502a173a 423 s/(\d)(\d{$grouping[0]}($|$thousands_sep))/$1$thousands_sep$2/;
14280422 424 print "$_";
5f05dabc
PP
425 }
426 print "\n";
427
5f05dabc
PP
428=head1 LOCALE CATEGORIES
429
5a964f20
TC
430The following subsections describe basic locale categories. Beyond these,
431some combination categories allow manipulation of more than one
432basic category at a time. See L<"ENVIRONMENT"> for a discussion of these.
5f05dabc
PP
433
434=head2 Category LC_COLLATE: Collation
435
5a964f20
TC
436In the scope of S<C<use locale>>, Perl looks to the C<LC_COLLATE>
437environment variable to determine the application's notions on collation
438(ordering) of characters. For example, 'b' follows 'a' in Latin
439alphabets, but where do 'E<aacute>' and 'E<aring>' belong? And while
440'color' follows 'chocolate' in English, what about in Spanish?
5f05dabc 441
60f0fa02
JH
442The following collations all make sense and you may meet any of them
443if you "use locale".
444
445 A B C D E a b c d e
446 A a B b C c D d D e
447 a A b B c C d D e E
448 a b c d e A B C D E
449
5a964f20
TC
450Here is a code snippet to tell what alphanumeric
451characters are in the current locale, in that locale's order:
5f05dabc
PP
452
453 use locale;
454 print +(sort grep /\w/, map { chr() } 0..255), "\n";
455
14280422
DD
456Compare this with the characters that you see and their order if you
457state explicitly that the locale should be ignored:
5f05dabc
PP
458
459 no locale;
460 print +(sort grep /\w/, map { chr() } 0..255), "\n";
461
462This machine-native collation (which is what you get unless S<C<use
463locale>> has appeared earlier in the same block) must be used for
464sorting raw binary data, whereas the locale-dependent collation of the
b0c42ed9 465first example is useful for natural text.
5f05dabc 466
14280422
DD
467As noted in L<USING LOCALES>, C<cmp> compares according to the current
468collation locale when C<use locale> is in effect, but falls back to a
5a964f20 469byte-by-byte comparison for strings that the locale says are equal. You
14280422
DD
470can use POSIX::strcoll() if you don't want this fall-back:
471
472 use POSIX qw(strcoll);
473 $equal_in_locale =
474 !strcoll("space and case ignored", "SpaceAndCaseIgnored");
475
476$equal_in_locale will be true if the collation locale specifies a
5a964f20 477dictionary-like ordering that ignores space characters completely and
9e3a2af8 478which folds case.
14280422 479
5a964f20 480If you have a single string that you want to check for "equality in
14280422
DD
481locale" against several others, you might think you could gain a little
482efficiency by using POSIX::strxfrm() in conjunction with C<eq>:
483
484 use POSIX qw(strxfrm);
485 $xfrm_string = strxfrm("Mixed-case string");
486 print "locale collation ignores spaces\n"
487 if $xfrm_string eq strxfrm("Mixed-casestring");
488 print "locale collation ignores hyphens\n"
489 if $xfrm_string eq strxfrm("Mixedcase string");
490 print "locale collation ignores case\n"
491 if $xfrm_string eq strxfrm("mixed-case string");
492
493strxfrm() takes a string and maps it into a transformed string for use
494in byte-by-byte comparisons against other transformed strings during
495collation. "Under the hood", locale-affected Perl comparison operators
5a964f20
TC
496call strxfrm() for both operands, then do a byte-by-byte
497comparison of the transformed strings. By calling strxfrm() explicitly
14280422 498and using a non locale-affected comparison, the example attempts to save
5a964f20 499a couple of transformations. But in fact, it doesn't save anything: Perl
2ae324a7 500magic (see L<perlguts/Magic Variables>) creates the transformed version of a
5a964f20 501string the first time it's needed in a comparison, then keeps this version around
14280422 502in case it's needed again. An example rewritten the easy way with
e38874e2 503C<cmp> runs just about as fast. It also copes with null characters
14280422 504embedded in strings; if you call strxfrm() directly, it treats the first
5a964f20
TC
505null it finds as a terminator. don't expect the transformed strings
506it produces to be portable across systems--or even from one revision
e38874e2
DD
507of your operating system to the next. In short, don't call strxfrm()
508directly: let Perl do it for you.
14280422 509
5a964f20 510Note: C<use locale> isn't shown in some of these examples because it isn't
14280422
DD
511needed: strcoll() and strxfrm() exist only to generate locale-dependent
512results, and so always obey the current C<LC_COLLATE> locale.
5f05dabc
PP
513
514=head2 Category LC_CTYPE: Character Types
515
5a964f20 516In the scope of S<C<use locale>>, Perl obeys the C<LC_CTYPE> locale
14280422
DD
517setting. This controls the application's notion of which characters are
518alphabetic. This affects Perl's C<\w> regular expression metanotation,
5a964f20 519which stands for alphanumeric characters--that is, alphabetic and
14280422
DD
520numeric characters. (Consult L<perlre> for more information about
521regular expressions.) Thanks to C<LC_CTYPE>, depending on your locale
522setting, characters like 'E<aelig>', 'E<eth>', 'E<szlig>', and
523'E<oslash>' may be understood as C<\w> characters.
5f05dabc 524
2c268ad5 525The C<LC_CTYPE> locale also provides the map used in transliterating
68dc0745 526characters between lower and uppercase. This affects the case-mapping
5a964f20
TC
527functions--lc(), lcfirst, uc(), and ucfirst(); case-mapping
528interpolation with C<\l>, C<\L>, C<\u>, or C<\U> in double-quoted strings
529and C<s///> substitutions; and case-independent regular expression
e38874e2
DD
530pattern matching using the C<i> modifier.
531
5a964f20
TC
532Finally, C<LC_CTYPE> affects the POSIX character-class test
533functions--isalpha(), islower(), and so on. For example, if you move
534from the "C" locale to a 7-bit Scandinavian one, you may find--possibly
535to your surprise--that "|" moves from the ispunct() class to isalpha().
5f05dabc 536
14280422
DD
537B<Note:> A broken or malicious C<LC_CTYPE> locale definition may result
538in clearly ineligible characters being considered to be alphanumeric by
5a964f20
TC
539your application. For strict matching of (mundane) letters and
540digits--for example, in command strings--locale-aware applications
14280422 541should use C<\w> inside a C<no locale> block. See L<"SECURITY">.
5f05dabc
PP
542
543=head2 Category LC_NUMERIC: Numeric Formatting
544
5a964f20
TC
545In the scope of S<C<use locale>>, Perl obeys the C<LC_NUMERIC> locale
546information, which controls an application's idea of how numbers should
547be formatted for human readability by the printf(), sprintf(), and
548write() functions. String-to-numeric conversion by the POSIX::strtod()
549function is also affected. In most implementations the only effect is to
550change the character used for the decimal point--perhaps from '.' to ','.
551These functions aren't aware of such niceties as thousands separation and
552so on. (See L<The localeconv function> if you care about these things.)
553
554Output produced by print() is B<never> affected by the
5f05dabc 555current locale: it is independent of whether C<use locale> or C<no
14280422 556locale> is in effect, and corresponds to what you'd get from printf()
5f05dabc
PP
557in the "C" locale. The same is true for Perl's internal conversions
558between numeric and string formats:
559
560 use POSIX qw(strtod);
561 use locale;
14280422 562
5f05dabc
PP
563 $n = 5/2; # Assign numeric 2.5 to $n
564
565 $a = " $n"; # Locale-independent conversion to string
566
567 print "half five is $n\n"; # Locale-independent output
568
569 printf "half five is %g\n", $n; # Locale-dependent output
570
14280422
DD
571 print "DECIMAL POINT IS COMMA\n"
572 if $n == (strtod("2,5"))[0]; # Locale-dependent conversion
5f05dabc
PP
573
574=head2 Category LC_MONETARY: Formatting of monetary amounts
575
5a964f20
TC
576The C standard defines the C<LC_MONETARY> category, but no function
577that is affected by its contents. (Those with experience of standards
b0c42ed9 578committees will recognize that the working group decided to punt on the
14280422 579issue.) Consequently, Perl takes no notice of it. If you really want
5a964f20
TC
580to use C<LC_MONETARY>, you can query its contents--see L<The localeconv
581function>--and use the information that it returns in your application's
582own formatting of currency amounts. However, you may well find that
583the information, voluminous and complex though it may be, still does not
584quite meet your requirements: currency formatting is a hard nut to crack.
5f05dabc
PP
585
586=head2 LC_TIME
587
5a964f20 588Output produced by POSIX::strftime(), which builds a formatted
5f05dabc
PP
589human-readable date/time string, is affected by the current C<LC_TIME>
590locale. Thus, in a French locale, the output produced by the C<%B>
591format element (full month name) for the first month of the year would
5a964f20 592be "janvier". Here's how to get a list of long month names in the
5f05dabc
PP
593current locale:
594
595 use POSIX qw(strftime);
14280422
DD
596 for (0..11) {
597 $long_month_name[$_] =
598 strftime("%B", 0, 0, 0, 1, $_, 96);
5f05dabc
PP
599 }
600
5a964f20 601Note: C<use locale> isn't needed in this example: as a function that
14280422
DD
602exists only to generate locale-dependent results, strftime() always
603obeys the current C<LC_TIME> locale.
5f05dabc
PP
604
605=head2 Other categories
606
5a964f20
TC
607The remaining locale category, C<LC_MESSAGES> (possibly supplemented
608by others in particular implementations) is not currently used by
609Perl--except possibly to affect the behavior of library functions called
610by extensions outside the standard Perl distribution.
14280422
DD
611
612=head1 SECURITY
613
5a964f20 614Although the main discussion of Perl security issues can be found in
14280422
DD
615L<perlsec>, a discussion of Perl's locale handling would be incomplete
616if it did not draw your attention to locale-dependent security issues.
5a964f20
TC
617Locales--particularly on systems that allow unprivileged users to
618build their own locales--are untrustworthy. A malicious (or just plain
14280422
DD
619broken) locale can make a locale-aware application give unexpected
620results. Here are a few possibilities:
621
622=over 4
623
624=item *
625
626Regular expression checks for safe file names or mail addresses using
5a964f20 627C<\w> may be spoofed by an C<LC_CTYPE> locale that claims that
14280422
DD
628characters such as "E<gt>" and "|" are alphanumeric.
629
630=item *
631
e38874e2
DD
632String interpolation with case-mapping, as in, say, C<$dest =
633"C:\U$name.$ext">, may produce dangerous results if a bogus LC_CTYPE
634case-mapping table is in effect.
635
636=item *
637
14280422
DD
638If the decimal point character in the C<LC_NUMERIC> locale is
639surreptitiously changed from a dot to a comma, C<sprintf("%g",
6400.123456e3)> produces a string result of "123,456". Many people would
641interpret this as one hundred and twenty-three thousand, four hundred
642and fifty-six.
643
644=item *
645
646A sneaky C<LC_COLLATE> locale could result in the names of students with
647"D" grades appearing ahead of those with "A"s.
648
649=item *
650
5a964f20 651An application that takes the trouble to use information in
14280422 652C<LC_MONETARY> may format debits as if they were credits and vice versa
5a964f20 653if that locale has been subverted. Or it might make payments in US
14280422
DD
654dollars instead of Hong Kong dollars.
655
656=item *
657
658The date and day names in dates formatted by strftime() could be
659manipulated to advantage by a malicious user able to subvert the
5a964f20 660C<LC_DATE> locale. ("Look--it says I wasn't in the building on
14280422
DD
661Sunday.")
662
663=back
664
665Such dangers are not peculiar to the locale system: any aspect of an
5a964f20 666application's environment which may be modified maliciously presents
14280422 667similar challenges. Similarly, they are not specific to Perl: any
5a964f20 668programming language that allows you to write programs that take
14280422
DD
669account of their environment exposes you to these issues.
670
5a964f20
TC
671Perl cannot protect you from all possibilities shown in the
672examples--there is no substitute for your own vigilance--but, when
14280422 673C<use locale> is in effect, Perl uses the tainting mechanism (see
5a964f20 674L<perlsec>) to mark string results that become locale-dependent, and
14280422 675which may be untrustworthy in consequence. Here is a summary of the
5a964f20 676tainting behavior of operators and functions that may be affected by
14280422
DD
677the locale:
678
679=over 4
680
681=item B<Comparison operators> (C<lt>, C<le>, C<ge>, C<gt> and C<cmp>):
682
683Scalar true/false (or less/equal/greater) result is never tainted.
684
7b8d334a 685=item B<Case-mapping interpolation> (with C<\l>, C<\L>, C<\u> or C<\U>)
e38874e2
DD
686
687Result string containing interpolated material is tainted if
688C<use locale> is in effect.
689
14280422
DD
690=item B<Matching operator> (C<m//>):
691
692Scalar true/false result never tainted.
693
5a964f20 694Subpatterns, either delivered as a list-context result or as $1 etc.
14280422 695are tainted if C<use locale> is in effect, and the subpattern regular
e38874e2
DD
696expression contains C<\w> (to match an alphanumeric character), C<\W>
697(non-alphanumeric character), C<\s> (white-space character), or C<\S>
5a964f20 698(non white-space character). The matched-pattern variable, $&, $`
e38874e2
DD
699(pre-match), $' (post-match), and $+ (last match) are also tainted if
700C<use locale> is in effect and the regular expression contains C<\w>,
701C<\W>, C<\s>, or C<\S>.
14280422
DD
702
703=item B<Substitution operator> (C<s///>):
704
e38874e2 705Has the same behavior as the match operator. Also, the left
5a964f20
TC
706operand of C<=~> becomes tainted when C<use locale> in effect
707if modified as a result of a substitution based on a regular
e38874e2 708expression match involving C<\w>, C<\W>, C<\s>, or C<\S>; or of
7b8d334a 709case-mapping with C<\l>, C<\L>,C<\u> or C<\U>.
14280422
DD
710
711=item B<In-memory formatting function> (sprintf()):
712
8ada0baa 713Result is tainted if C<use locale> is in effect.
14280422
DD
714
715=item B<Output formatting functions> (printf() and write()):
716
717Success/failure result is never tainted.
718
719=item B<Case-mapping functions> (lc(), lcfirst(), uc(), ucfirst()):
720
721Results are tainted if C<use locale> is in effect.
722
723=item B<POSIX locale-dependent functions> (localeconv(), strcoll(),
724strftime(), strxfrm()):
725
726Results are never tainted.
727
728=item B<POSIX character class tests> (isalnum(), isalpha(), isdigit(),
729isgraph(), islower(), isprint(), ispunct(), isspace(), isupper(),
730isxdigit()):
731
732True/false results are never tainted.
733
734=back
735
736Three examples illustrate locale-dependent tainting.
737The first program, which ignores its locale, won't run: a value taken
54310121 738directly from the command line may not be used to name an output file
14280422
DD
739when taint checks are enabled.
740
741 #/usr/local/bin/perl -T
742 # Run with taint checking
743
54310121 744 # Command line sanity check omitted...
14280422
DD
745 $tainted_output_file = shift;
746
747 open(F, ">$tainted_output_file")
748 or warn "Open of $untainted_output_file failed: $!\n";
749
750The program can be made to run by "laundering" the tainted value through
5a964f20
TC
751a regular expression: the second example--which still ignores locale
752information--runs, creating the file named on its command line
14280422
DD
753if it can.
754
755 #/usr/local/bin/perl -T
756
757 $tainted_output_file = shift;
758 $tainted_output_file =~ m%[\w/]+%;
759 $untainted_output_file = $&;
760
761 open(F, ">$untainted_output_file")
762 or warn "Open of $untainted_output_file failed: $!\n";
763
5a964f20 764Compare this with a similar but locale-aware program:
14280422
DD
765
766 #/usr/local/bin/perl -T
767
768 $tainted_output_file = shift;
769 use locale;
770 $tainted_output_file =~ m%[\w/]+%;
771 $localized_output_file = $&;
772
773 open(F, ">$localized_output_file")
774 or warn "Open of $localized_output_file failed: $!\n";
775
776This third program fails to run because $& is tainted: it is the result
5a964f20 777of a match involving C<\w> while C<use locale> is in effect.
5f05dabc
PP
778
779=head1 ENVIRONMENT
780
781=over 12
782
783=item PERL_BADLANG
784
14280422 785A string that can suppress Perl's warning about failed locale settings
54310121 786at startup. Failure can occur if the locale support in the operating
5a964f20 787system is lacking (broken) in some way--or if you mistyped the name of
14280422 788a locale when you set up your environment. If this environment variable
5a964f20
TC
789is absent, or has a value that does not evaluate to integer zero--that
790is, "0" or ""--Perl will complain about locale setting failures.
5f05dabc 791
14280422
DD
792B<NOTE>: PERL_BADLANG only gives you a way to hide the warning message.
793The message tells about some problem in your system's locale support,
794and you should investigate what the problem is.
5f05dabc
PP
795
796=back
797
798The following environment variables are not specific to Perl: They are
14280422
DD
799part of the standardized (ISO C, XPG4, POSIX 1.c) setlocale() method
800for controlling an application's opinion on data.
5f05dabc
PP
801
802=over 12
803
804=item LC_ALL
805
5a964f20 806C<LC_ALL> is the "override-all" locale environment variable. If
5f05dabc
PP
807set, it overrides all the rest of the locale environment variables.
808
809=item LC_CTYPE
810
811In the absence of C<LC_ALL>, C<LC_CTYPE> chooses the character type
812locale. In the absence of both C<LC_ALL> and C<LC_CTYPE>, C<LANG>
813chooses the character type locale.
814
815=item LC_COLLATE
816
14280422
DD
817In the absence of C<LC_ALL>, C<LC_COLLATE> chooses the collation
818(sorting) locale. In the absence of both C<LC_ALL> and C<LC_COLLATE>,
819C<LANG> chooses the collation locale.
5f05dabc
PP
820
821=item LC_MONETARY
822
14280422
DD
823In the absence of C<LC_ALL>, C<LC_MONETARY> chooses the monetary
824formatting locale. In the absence of both C<LC_ALL> and C<LC_MONETARY>,
825C<LANG> chooses the monetary formatting locale.
5f05dabc
PP
826
827=item LC_NUMERIC
828
829In the absence of C<LC_ALL>, C<LC_NUMERIC> chooses the numeric format
830locale. In the absence of both C<LC_ALL> and C<LC_NUMERIC>, C<LANG>
831chooses the numeric format.
832
833=item LC_TIME
834
14280422
DD
835In the absence of C<LC_ALL>, C<LC_TIME> chooses the date and time
836formatting locale. In the absence of both C<LC_ALL> and C<LC_TIME>,
837C<LANG> chooses the date and time formatting locale.
5f05dabc
PP
838
839=item LANG
840
14280422
DD
841C<LANG> is the "catch-all" locale environment variable. If it is set, it
842is used as the last resort after the overall C<LC_ALL> and the
5f05dabc
PP
843category-specific C<LC_...>.
844
845=back
846
847=head1 NOTES
848
849=head2 Backward compatibility
850
b0c42ed9 851Versions of Perl prior to 5.004 B<mostly> ignored locale information,
5a964f20
TC
852generally behaving as if something similar to the C<"C"> locale were
853always in force, even if the program environment suggested otherwise
854(see L<The setlocale function>). By default, Perl still behaves this
855way for backward compatibility. If you want a Perl application to pay
856attention to locale information, you B<must> use the S<C<use locale>>
857pragma (see L<The use locale Pragma>) to instruct it to do so.
b0c42ed9
JH
858
859Versions of Perl from 5.002 to 5.003 did use the C<LC_CTYPE>
5a964f20
TC
860information if available; that is, C<\w> did understand what
861were the letters according to the locale environment variables.
b0c42ed9
JH
862The problem was that the user had no control over the feature:
863if the C library supported locales, Perl used them.
864
865=head2 I18N:Collate obsolete
866
5a964f20 867In versions of Perl prior to 5.004, per-locale collation was possible
b0c42ed9
JH
868using the C<I18N::Collate> library module. This module is now mildly
869obsolete and should be avoided in new applications. The C<LC_COLLATE>
870functionality is now integrated into the Perl core language: One can
871use locale-specific scalar data completely normally with C<use locale>,
872so there is no longer any need to juggle with the scalar references of
873C<I18N::Collate>.
5f05dabc 874
14280422 875=head2 Sort speed and memory use impacts
5f05dabc
PP
876
877Comparing and sorting by locale is usually slower than the default
14280422
DD
878sorting; slow-downs of two to four times have been observed. It will
879also consume more memory: once a Perl scalar variable has participated
880in any string comparison or sorting operation obeying the locale
881collation rules, it will take 3-15 times more memory than before. (The
882exact multiplier depends on the string's contents, the operating system
883and the locale.) These downsides are dictated more by the operating
884system's implementation of the locale system than by Perl.
5f05dabc 885
e38874e2
DD
886=head2 write() and LC_NUMERIC
887
5a964f20 888Formats are the only part of Perl that unconditionally use information
e38874e2
DD
889from a program's locale; if a program's environment specifies an
890LC_NUMERIC locale, it is always used to specify the decimal point
891character in formatted output. Formatted output cannot be controlled by
892C<use locale> because the pragma is tied to the block structure of the
893program, and, for historical reasons, formats exist outside that block
894structure.
895
5f05dabc
PP
896=head2 Freely available locale definitions
897
898There is a large collection of locale definitions at
14280422
DD
899C<ftp://dkuug.dk/i18n/WG15-collection>. You should be aware that it is
900unsupported, and is not claimed to be fit for any purpose. If your
5a964f20 901system allows installation of arbitrary locales, you may find the
14280422
DD
902definitions useful as they are, or as a basis for the development of
903your own locales.
5f05dabc 904
14280422 905=head2 I18n and l10n
5f05dabc 906
b0c42ed9
JH
907"Internationalization" is often abbreviated as B<i18n> because its first
908and last letters are separated by eighteen others. (You may guess why
909the internalin ... internaliti ... i18n tends to get abbreviated.) In
910the same way, "localization" is often abbreviated to B<l10n>.
14280422
DD
911
912=head2 An imperfect standard
913
914Internationalization, as defined in the C and POSIX standards, can be
915criticized as incomplete, ungainly, and having too large a granularity.
916(Locales apply to a whole process, when it would arguably be more useful
917to have them apply to a single thread, window group, or whatever.) They
918also have a tendency, like standards groups, to divide the world into
919nations, when we all know that the world can equally well be divided
920into bankers, bikers, gamers, and so on. But, for now, it's the only
921standard we've got. This may be construed as a bug.
5f05dabc
PP
922
923=head1 BUGS
924
925=head2 Broken systems
926
5a964f20 927In certain systems, the operating system's locale support
2bdf8add
JH
928is broken and cannot be fixed or used by Perl. Such deficiencies can
929and will result in mysterious hangs and/or Perl core dumps when the
930C<use locale> is in effect. When confronted with such a system,
9607fc9c 931please report in excruciating detail to <F<perlbug@perl.com>>, and
5a964f20 932complain to your vendor: bug fixes may exist for these problems
2bdf8add
JH
933in your operating system. Sometimes such bug fixes are called an
934operating system upgrade.
5f05dabc
PP
935
936=head1 SEE ALSO
937
7b8d334a
GS
938L<POSIX (3)/isalnum>
939
940L<POSIX (3)/isalpha>
941
942L<POSIX (3)/isdigit>
943
944L<POSIX (3)/isgraph>
945
946L<POSIX (3)/islower>
947
948L<POSIX (3)/isprint>,
949
950L<POSIX (3)/ispunct>
951
952L<POSIX (3)/isspace>
953
954L<POSIX (3)/isupper>,
955
956L<POSIX (3)/isxdigit>
957
958L<POSIX (3)/localeconv>
959
960L<POSIX (3)/setlocale>,
961
962L<POSIX (3)/strcoll>
963
964L<POSIX (3)/strftime>
965
966L<POSIX (3)/strtod>,
967
14280422 968L<POSIX (3)/strxfrm>
5f05dabc
PP
969
970=head1 HISTORY
971
b0c42ed9 972Jarkko Hietaniemi's original F<perli18n.pod> heavily hacked by Dominic
5a964f20
TC
973Dunlop, assisted by the perl5-porters. Prose worked over a bit by
974Tom Christiansen.
5f05dabc 975
5a964f20 976Last update: Thu Jun 11 08:44:13 MDT 1998