5 perllocale - Perl locale handling (internationalization and localization)
9 In the beginning there was ASCII, the "American Standard Code for
10 Information Interchange", which works quite well for Americans with
11 their English alphabet and dollar-denominated currency. But it doesn't
12 work so well even for other English speakers, who may use different
13 currencies, such as the pound sterling (as the symbol for that currency
14 is not in ASCII); and it's hopelessly inadequate for many of the
15 thousands of the world's other languages.
17 To address these deficiencies, the concept of locales was invented
18 (formally the ISO C, XPG4, POSIX 1.c "locale system"). And applications
19 were and are being written that use the locale mechanism. The process of
20 making such an application take account of its users' preferences in
21 these kinds of matters is called B<internationalization> (often
22 abbreviated as B<i18n>); telling such an application about a particular
23 set of preferences is known as B<localization> (B<l10n>).
25 Perl was extended to support the locale system. This
26 is controlled per application by using one pragma, one function call,
27 and several environment variables.
29 Unfortunately, there are quite a few deficiencies with the design (and
30 often, the implementations) of locales, and their use for character sets
31 has mostly been supplanted by Unicode (see L<perlunitut> for an
32 introduction to that, and keep on reading here for how Unicode interacts
33 with locales in Perl).
35 Perl continues to support the old locale system, and starting in v5.16,
36 provides a hybrid way to use the Unicode character set, along with the
37 other portions of locales that may not be so problematic.
38 (Unicode is also creating C<CLDR>, the "Common Locale Data Repository",
39 L<http://cldr.unicode.org/> which includes more types of information than
40 are available in the POSIX locale system. At the time of this writing,
41 there was no CPAN module that provides access to this XML-encoded data.
42 However, many of its locales have the POSIX-only data extracted, and are
43 available at L<http://unicode.org/Public/cldr/latest/>.)
45 =head1 WHAT IS A LOCALE
47 A locale is a set of data that describes various aspects of how various
48 communities in the world categorize their world. These categories are
49 broken down into the following types (some of which include a brief
54 =item Category LC_NUMERIC: Numeric formatting
56 This indicates how numbers should be formatted for human readability,
57 for example the character used as the decimal point.
59 =item Category LC_MONETARY: Formatting of monetary amounts
62 The nbsp below makes this look better
66 =item Category LC_TIME: Date/Time formatting
69 The nbsp below makes this look better
73 =item Category LC_MESSAGES: Error and other messages
75 This for the most part is beyond the scope of Perl
77 =item Category LC_COLLATE: Collation
79 This indicates the ordering of letters for comparision and sorting.
80 In Latin alphabets, for example, "b", generally follows "a".
82 =item Category LC_CTYPE: Character Types
84 This indicates, for example if a character is an uppercase letter.
88 More details on the categories are given below in L</LOCALE CATEGORIES>.
90 Together, these categories go a long way towards being able to customize
91 a single program to run in many different locations. But there are
92 deficiencies, so keep reading.
94 =head1 PREPARING TO USE LOCALES
96 Perl will not use locales unless specifically requested to (see L</NOTES> below
97 for the partial exception of C<write()>). But even if there is such a
98 request, B<all> of the following must be true for it to work properly:
104 B<Your operating system must support the locale system>. If it does,
105 you should find that the setlocale() function is a documented part of
110 B<Definitions for locales that you use must be installed>. You, or
111 your system administrator, must make sure that this is the case. The
112 available locales, the location in which they are kept, and the manner
113 in which they are installed all vary from system to system. Some systems
114 provide only a few, hard-wired locales and do not allow more to be
115 added. Others allow you to add "canned" locales provided by the system
116 supplier. Still others allow you or the system administrator to define
117 and add arbitrary locales. (You may have to ask your supplier to
118 provide canned locales that are not delivered with your operating
119 system.) Read your system documentation for further illumination.
123 B<Perl must believe that the locale system is supported>. If it does,
124 C<perl -V:d_setlocale> will say that the value for C<d_setlocale> is
129 If you want a Perl application to process and present your data
130 according to a particular locale, the application code should include
131 the S<C<use locale>> pragma (see L<The use locale pragma>) where
132 appropriate, and B<at least one> of the following must be true:
138 B<The locale-determining environment variables (see L</"ENVIRONMENT">)
139 must be correctly set up> at the time the application is started, either
140 by yourself or by whomever set up your system account; or
144 B<The application must set its own locale> using the method described in
145 L<The setlocale function>.
151 =head2 The use locale pragma
153 By default, Perl ignores the current locale. The S<C<use locale>>
154 pragma tells Perl to use the current locale for some operations.
155 Starting in v5.16, there is an optional parameter to this pragma:
157 use locale ':not_characters';
159 This parameter allows better mixing of locales and Unicode, and is
160 described fully in L</Unicode and UTF-8>, but briefly, it tells Perl to
161 not use the character portions of the locale definition, that is
162 the C<LC_CTYPE> and C<LC_COLLATE> categories. Instead it will use the
163 native (extended by Unicode) character set. When using this parameter,
164 you are responsible for getting the external character set translated
165 into the native/Unicode one (which it already will be if it is one of
166 the increasingly popular UTF-8 locales). There are convenient ways of
167 doing this, as described in L</Unicode and UTF-8>.
169 The current locale is set at execution time by
170 L<setlocale()|/The setlocale function> described below. If that function
171 hasn't yet been called in the course of the program's execution, the
172 current locale is that which was determined by the L</"ENVIRONMENT"> in
173 effect at the start of the program, except that
174 C<L<LC_NUMERIC|/Category LC_NUMERIC: Numeric Formatting>> is always
175 initialized to the C locale (mentioned under L<Finding locales>).
176 If there is no valid environment, the current locale is undefined. It
177 is likely, but not necessarily, the "C" locale.
179 The operations that are affected by locale are:
183 =item B<Under C<use locale ':not_characters';>>
189 B<Format declarations> (format()) use C<LC_NUMERIC>
193 B<The POSIX date formatting function> (strftime()) uses C<LC_TIME>.
198 The nbsp below makes this look better
202 =item B<Under just plain C<use locale;>>
204 The above operations are affected, as well as the following:
210 B<The comparison operators> (C<lt>, C<le>, C<cmp>, C<ge>, and C<gt>) and
211 the POSIX string collation functions strcoll() and strxfrm() use
212 C<LC_COLLATE>. sort() is also affected if used without an
213 explicit comparison function, because it uses C<cmp> by default.
215 B<Note:> C<eq> and C<ne> are unaffected by locale: they always
216 perform a char-by-char comparison of their scalar operands. What's
217 more, if C<cmp> finds that its operands are equal according to the
218 collation sequence specified by the current locale, it goes on to
219 perform a char-by-char comparison, and only returns I<0> (equal) if the
220 operands are char-for-char identical. If you really want to know whether
221 two strings--which C<eq> and C<cmp> may consider different--are equal
222 as far as collation in the locale is concerned, see the discussion in
223 L<Category LC_COLLATE: Collation>.
227 B<Regular expressions and case-modification functions> (uc(), lc(),
228 ucfirst(), and lcfirst()) use C<LC_CTYPE>
234 The default behavior is restored with the S<C<no locale>> pragma, or
235 upon reaching the end of the block enclosing C<use locale>.
236 Note that C<use locale> and C<use locale ':not_characters'> may be
237 nested, and that what is in effect within an inner scope will revert to
238 the outer scope's rules at the end of the inner scope.
240 The string result of any operation that uses locale
241 information is tainted, as it is possible for a locale to be
242 untrustworthy. See L<"SECURITY">.
244 =head2 The setlocale function
246 You can switch locales as often as you wish at run time with the
247 POSIX::setlocale() function:
249 # Import locale-handling tool set from POSIX module.
250 # This example uses: setlocale -- the function call
251 # LC_CTYPE -- explained below
252 use POSIX qw(locale_h);
254 # query and save the old locale
255 $old_locale = setlocale(LC_CTYPE);
257 setlocale(LC_CTYPE, "fr_CA.ISO8859-1");
258 # LC_CTYPE now in locale "French, Canada, codeset ISO 8859-1"
260 setlocale(LC_CTYPE, "");
261 # LC_CTYPE now reset to default defined by LC_ALL/LC_CTYPE/LANG
262 # environment variables. See below for documentation.
264 # restore the old locale
265 setlocale(LC_CTYPE, $old_locale);
267 The first argument of setlocale() gives the B<category>, the second the
268 B<locale>. The category tells in what aspect of data processing you
269 want to apply locale-specific rules. Category names are discussed in
270 L</LOCALE CATEGORIES> and L</"ENVIRONMENT">. The locale is the name of a
271 collection of customization information corresponding to a particular
272 combination of language, country or territory, and codeset. Read on for
273 hints on the naming of locales: not all systems name locales as in the
276 If no second argument is provided and the category is something else
277 than LC_ALL, the function returns a string naming the current locale
278 for the category. You can use this value as the second argument in a
279 subsequent call to setlocale().
281 If no second argument is provided and the category is LC_ALL, the
282 result is implementation-dependent. It may be a string of
283 concatenated locale names (separator also implementation-dependent)
284 or a single locale name. Please consult your setlocale(3) man page for
287 If a second argument is given and it corresponds to a valid locale,
288 the locale for the category is set to that value, and the function
289 returns the now-current locale value. You can then use this in yet
290 another call to setlocale(). (In some implementations, the return
291 value may sometimes differ from the value you gave as the second
292 argument--think of it as an alias for the value you gave.)
294 As the example shows, if the second argument is an empty string, the
295 category's locale is returned to the default specified by the
296 corresponding environment variables. Generally, this results in a
297 return to the default that was in force when Perl started up: changes
298 to the environment made by the application after startup may or may not
299 be noticed, depending on your system's C library.
301 If the second argument does not correspond to a valid locale, the locale
302 for the category is not changed, and the function returns I<undef>.
304 Note that Perl ignores the current C<LC_CTYPE> and C<LC_COLLATE> locales
305 within the scope of a C<use locale ':not_characters'>.
307 For further information about the categories, consult setlocale(3).
309 =head2 Finding locales
311 For locales available in your system, consult also setlocale(3) to
312 see whether it leads to the list of available locales (search for the
313 I<SEE ALSO> section). If that fails, try the following command lines:
327 and see whether they list something resembling these
329 en_US.ISO8859-1 de_DE.ISO8859-1 ru_RU.ISO8859-5
330 en_US.iso88591 de_DE.iso88591 ru_RU.iso88595
333 english german russian
334 english.iso88591 german.iso88591 russian.iso88595
335 english.roman8 russian.koi8r
337 Sadly, even though the calling interface for setlocale() has been
338 standardized, names of locales and the directories where the
339 configuration resides have not been. The basic form of the name is
340 I<language_territory>B<.>I<codeset>, but the latter parts after
341 I<language> are not always present. The I<language> and I<country>
342 are usually from the standards B<ISO 3166> and B<ISO 639>, the
343 two-letter abbreviations for the countries and the languages of the
344 world, respectively. The I<codeset> part often mentions some B<ISO
345 8859> character set, the Latin codesets. For example, C<ISO 8859-1>
346 is the so-called "Western European codeset" that can be used to encode
347 most Western European languages adequately. Again, there are several
348 ways to write even the name of that one standard. Lamentably.
350 Two special locales are worth particular mention: "C" and "POSIX".
351 Currently these are effectively the same locale: the difference is
352 mainly that the first one is defined by the C standard, the second by
353 the POSIX standard. They define the B<default locale> in which
354 every program starts in the absence of locale information in its
355 environment. (The I<default> default locale, if you will.) Its language
356 is (American) English and its character codeset ASCII.
357 B<Warning>. The C locale delivered by some vendors may not
358 actually exactly match what the C standard calls for. So beware.
360 B<NOTE>: Not all systems have the "POSIX" locale (not all systems are
361 POSIX-conformant), so use "C" when you need explicitly to specify this
364 =head2 LOCALE PROBLEMS
366 You may encounter the following warning message at Perl startup:
368 perl: warning: Setting locale failed.
369 perl: warning: Please check that your locale settings:
372 are supported and installed on your system.
373 perl: warning: Falling back to the standard locale ("C").
375 This means that your locale settings had LC_ALL set to "En_US" and
376 LANG exists but has no value. Perl tried to believe you but could not.
377 Instead, Perl gave up and fell back to the "C" locale, the default locale
378 that is supposed to work no matter what. This usually means your locale
379 settings were wrong, they mention locales your system has never heard
380 of, or the locale installation in your system has problems (for example,
381 some system files are broken or missing). There are quick and temporary
382 fixes to these problems, as well as more thorough and lasting fixes.
384 =head2 Temporarily fixing locale problems
386 The two quickest fixes are either to render Perl silent about any
387 locale inconsistencies or to run Perl under the default locale "C".
389 Perl's moaning about locale problems can be silenced by setting the
390 environment variable PERL_BADLANG to a zero value, for example "0".
391 This method really just sweeps the problem under the carpet: you tell
392 Perl to shut up even when Perl sees that something is wrong. Do not
393 be surprised if later something locale-dependent misbehaves.
395 Perl can be run under the "C" locale by setting the environment
396 variable LC_ALL to "C". This method is perhaps a bit more civilized
397 than the PERL_BADLANG approach, but setting LC_ALL (or
398 other locale variables) may affect other programs as well, not just
399 Perl. In particular, external programs run from within Perl will see
400 these changes. If you make the new settings permanent (read on), all
401 programs you run see the changes. See L<"ENVIRONMENT"> for
402 the full list of relevant environment variables and L<USING LOCALES>
403 for their effects in Perl. Effects in other programs are
404 easily deducible. For example, the variable LC_COLLATE may well affect
405 your B<sort> program (or whatever the program that arranges "records"
406 alphabetically in your system is called).
408 You can test out changing these variables temporarily, and if the
409 new settings seem to help, put those settings into your shell startup
410 files. Consult your local documentation for the exact details. For in
411 Bourne-like shells (B<sh>, B<ksh>, B<bash>, B<zsh>):
413 LC_ALL=en_US.ISO8859-1
416 This assumes that we saw the locale "en_US.ISO8859-1" using the commands
417 discussed above. We decided to try that instead of the above faulty
418 locale "En_US"--and in Cshish shells (B<csh>, B<tcsh>)
420 setenv LC_ALL en_US.ISO8859-1
422 or if you have the "env" application you can do in any shell
424 env LC_ALL=en_US.ISO8859-1 perl ...
426 If you do not know what shell you have, consult your local
427 helpdesk or the equivalent.
429 =head2 Permanently fixing locale problems
431 The slower but superior fixes are when you may be able to yourself
432 fix the misconfiguration of your own environment variables. The
433 mis(sing)configuration of the whole system's locales usually requires
434 the help of your friendly system administrator.
436 First, see earlier in this document about L<Finding locales>. That tells
437 how to find which locales are really supported--and more importantly,
438 installed--on your system. In our example error message, environment
439 variables affecting the locale are listed in the order of decreasing
440 importance (and unset variables do not matter). Therefore, having
441 LC_ALL set to "En_US" must have been the bad choice, as shown by the
442 error message. First try fixing locale settings listed first.
444 Second, if using the listed commands you see something B<exactly>
445 (prefix matches do not count and case usually counts) like "En_US"
446 without the quotes, then you should be okay because you are using a
447 locale name that should be installed and available in your system.
448 In this case, see L<Permanently fixing your system's locale configuration>.
450 =head2 Permanently fixing your system's locale configuration
452 This is when you see something like:
454 perl: warning: Please check that your locale settings:
457 are supported and installed on your system.
459 but then cannot see that "En_US" listed by the above-mentioned
460 commands. You may see things like "en_US.ISO8859-1", but that isn't
461 the same. In this case, try running under a locale
462 that you can list and which somehow matches what you tried. The
463 rules for matching locale names are a bit vague because
464 standardization is weak in this area. See again the
465 L<Finding locales> about general rules.
467 =head2 Fixing system locale configuration
469 Contact a system administrator (preferably your own) and report the exact
470 error message you get, and ask them to read this same documentation you
471 are now reading. They should be able to check whether there is something
472 wrong with the locale configuration of the system. The L<Finding locales>
473 section is unfortunately a bit vague about the exact commands and places
474 because these things are not that standardized.
476 =head2 The localeconv function
478 The POSIX::localeconv() function allows you to get particulars of the
479 locale-dependent numeric formatting information specified by the current
480 C<LC_NUMERIC> and C<LC_MONETARY> locales. (If you just want the name of
481 the current locale for a particular category, use POSIX::setlocale()
482 with a single parameter--see L<The setlocale function>.)
484 use POSIX qw(locale_h);
486 # Get a reference to a hash of locale-dependent info
487 $locale_values = localeconv();
489 # Output sorted list of the values
490 for (sort keys %$locale_values) {
491 printf "%-20s = %s\n", $_, $locale_values->{$_}
494 localeconv() takes no arguments, and returns B<a reference to> a hash.
495 The keys of this hash are variable names for formatting, such as
496 C<decimal_point> and C<thousands_sep>. The values are the
497 corresponding, er, values. See L<POSIX/localeconv> for a longer
498 example listing the categories an implementation might be expected to
499 provide; some provide more and others fewer. You don't need an
500 explicit C<use locale>, because localeconv() always observes the
503 Here's a simple-minded example program that rewrites its command-line
504 parameters as integers correctly formatted in the current locale:
506 use POSIX qw(locale_h);
508 # Get some of locale's numeric formatting parameters
509 my ($thousands_sep, $grouping) =
510 @{localeconv()}{'thousands_sep', 'grouping'};
512 # Apply defaults if values are missing
513 $thousands_sep = ',' unless $thousands_sep;
515 # grouping and mon_grouping are packed lists
516 # of small integers (characters) telling the
517 # grouping (thousand_seps and mon_thousand_seps
518 # being the group dividers) of numbers and
519 # monetary quantities. The integers' meanings:
520 # 255 means no more grouping, 0 means repeat
521 # the previous grouping, 1-254 means use that
522 # as the current grouping. Grouping goes from
523 # right to left (low to high digits). In the
524 # below we cheat slightly by never using anything
525 # else than the first grouping (whatever that is).
527 @grouping = unpack("C*", $grouping);
532 # Format command line params for current locale
534 $_ = int; # Chop non-integer part
536 s/(\d)(\d{$grouping[0]}($|$thousands_sep))/$1$thousands_sep$2/;
541 =head2 I18N::Langinfo
543 Another interface for querying locale-dependent information is the
544 I18N::Langinfo::langinfo() function, available at least in Unix-like
547 The following example will import the langinfo() function itself and
548 three constants to be used as arguments to langinfo(): a constant for
549 the abbreviated first day of the week (the numbering starts from
550 Sunday = 1) and two more constants for the affirmative and negative
551 answers for a yes/no question in the current locale.
553 use I18N::Langinfo qw(langinfo ABDAY_1 YESSTR NOSTR);
555 my ($abday_1, $yesstr, $nostr)
556 = map { langinfo } qw(ABDAY_1 YESSTR NOSTR);
558 print "$abday_1? [$yesstr/$nostr] ";
560 In other words, in the "C" (or English) locale the above will probably
561 print something like:
565 See L<I18N::Langinfo> for more information.
567 =head1 LOCALE CATEGORIES
569 The following subsections describe basic locale categories. Beyond these,
570 some combination categories allow manipulation of more than one
571 basic category at a time. See L<"ENVIRONMENT"> for a discussion of these.
573 =head2 Category LC_COLLATE: Collation
575 In the scope of S<C<use locale>> (but not a
576 C<use locale ':not_characters'>), Perl looks to the C<LC_COLLATE>
577 environment variable to determine the application's notions on collation
578 (ordering) of characters. For example, "b" follows "a" in Latin
579 alphabets, but where do "E<aacute>" and "E<aring>" belong? And while
580 "color" follows "chocolate" in English, what about in traditional Spanish?
582 The following collations all make sense and you may meet any of them
590 Here is a code snippet to tell what "word"
591 characters are in the current locale, in that locale's order:
594 print +(sort grep /\w/, map { chr } 0..255), "\n";
596 Compare this with the characters that you see and their order if you
597 state explicitly that the locale should be ignored:
600 print +(sort grep /\w/, map { chr } 0..255), "\n";
602 This machine-native collation (which is what you get unless S<C<use
603 locale>> has appeared earlier in the same block) must be used for
604 sorting raw binary data, whereas the locale-dependent collation of the
605 first example is useful for natural text.
607 As noted in L<USING LOCALES>, C<cmp> compares according to the current
608 collation locale when C<use locale> is in effect, but falls back to a
609 char-by-char comparison for strings that the locale says are equal. You
610 can use POSIX::strcoll() if you don't want this fall-back:
612 use POSIX qw(strcoll);
614 !strcoll("space and case ignored", "SpaceAndCaseIgnored");
616 $equal_in_locale will be true if the collation locale specifies a
617 dictionary-like ordering that ignores space characters completely and
620 If you have a single string that you want to check for "equality in
621 locale" against several others, you might think you could gain a little
622 efficiency by using POSIX::strxfrm() in conjunction with C<eq>:
624 use POSIX qw(strxfrm);
625 $xfrm_string = strxfrm("Mixed-case string");
626 print "locale collation ignores spaces\n"
627 if $xfrm_string eq strxfrm("Mixed-casestring");
628 print "locale collation ignores hyphens\n"
629 if $xfrm_string eq strxfrm("Mixedcase string");
630 print "locale collation ignores case\n"
631 if $xfrm_string eq strxfrm("mixed-case string");
633 strxfrm() takes a string and maps it into a transformed string for use
634 in char-by-char comparisons against other transformed strings during
635 collation. "Under the hood", locale-affected Perl comparison operators
636 call strxfrm() for both operands, then do a char-by-char
637 comparison of the transformed strings. By calling strxfrm() explicitly
638 and using a non locale-affected comparison, the example attempts to save
639 a couple of transformations. But in fact, it doesn't save anything: Perl
640 magic (see L<perlguts/Magic Variables>) creates the transformed version of a
641 string the first time it's needed in a comparison, then keeps this version around
642 in case it's needed again. An example rewritten the easy way with
643 C<cmp> runs just about as fast. It also copes with null characters
644 embedded in strings; if you call strxfrm() directly, it treats the first
645 null it finds as a terminator. don't expect the transformed strings
646 it produces to be portable across systems--or even from one revision
647 of your operating system to the next. In short, don't call strxfrm()
648 directly: let Perl do it for you.
650 Note: C<use locale> isn't shown in some of these examples because it isn't
651 needed: strcoll() and strxfrm() exist only to generate locale-dependent
652 results, and so always obey the current C<LC_COLLATE> locale.
654 =head2 Category LC_CTYPE: Character Types
656 In the scope of S<C<use locale>> (but not a
657 C<use locale ':not_characters'>), Perl obeys the C<LC_CTYPE> locale
658 setting. This controls the application's notion of which characters are
659 alphabetic. This affects Perl's C<\w> regular expression metanotation,
660 which stands for alphanumeric characters--that is, alphabetic,
661 numeric, and including other special characters such as the underscore or
662 hyphen. (Consult L<perlre> for more information about
663 regular expressions.) Thanks to C<LC_CTYPE>, depending on your locale
664 setting, characters like "E<aelig>", "E<eth>", "E<szlig>", and
665 "E<oslash>" may be understood as C<\w> characters.
667 The C<LC_CTYPE> locale also provides the map used in transliterating
668 characters between lower and uppercase. This affects the case-mapping
669 functions--lc(), lcfirst, uc(), and ucfirst(); case-mapping
670 interpolation with C<\l>, C<\L>, C<\u>, or C<\U> in double-quoted strings
671 and C<s///> substitutions; and case-independent regular expression
672 pattern matching using the C<i> modifier.
674 Finally, C<LC_CTYPE> affects the POSIX character-class test
675 functions--isalpha(), islower(), and so on. For example, if you move
676 from the "C" locale to a 7-bit Scandinavian one, you may find--possibly
677 to your surprise--that "|" moves from the ispunct() class to isalpha().
678 Unfortunately, this creates big problems for regular expressions. "|" still
679 means alternation even though it matches C<\w>.
681 Note that there are quite a few things that are unaffected by the
682 current locale. All the escape sequences for particular characters,
683 C<\n> for example, always mean the platform's native one. This means,
684 for example, that C<\N> in regular expressions (every character
685 but new-line) work on the platform character set.
687 B<Note:> A broken or malicious C<LC_CTYPE> locale definition may result
688 in clearly ineligible characters being considered to be alphanumeric by
689 your application. For strict matching of (mundane) ASCII letters and
690 digits--for example, in command strings--locale-aware applications
691 should use C<\w> with the C</a> regular expression modifier. See L<"SECURITY">.
693 =head2 Category LC_NUMERIC: Numeric Formatting
695 After a proper POSIX::setlocale() call, Perl obeys the C<LC_NUMERIC>
696 locale information, which controls an application's idea of how numbers
697 should be formatted for human readability by the printf(), sprintf(), and
698 write() functions. String-to-numeric conversion by the POSIX::strtod()
699 function is also affected. In most implementations the only effect is to
700 change the character used for the decimal point--perhaps from "." to ",".
701 These functions aren't aware of such niceties as thousands separation and
702 so on. (See L<The localeconv function> if you care about these things.)
704 Output produced by print() is also affected by the current locale: it
705 corresponds to what you'd get from printf() in the "C" locale. The
706 same is true for Perl's internal conversions between numeric and
709 use POSIX qw(strtod setlocale LC_NUMERIC);
711 setlocale LC_NUMERIC, "";
713 $n = 5/2; # Assign numeric 2.5 to $n
715 $a = " $n"; # Locale-dependent conversion to string
717 print "half five is $n\n"; # Locale-dependent output
719 printf "half five is %g\n", $n; # Locale-dependent output
721 print "DECIMAL POINT IS COMMA\n"
722 if $n == (strtod("2,5"))[0]; # Locale-dependent conversion
724 See also L<I18N::Langinfo> and C<RADIXCHAR>.
726 =head2 Category LC_MONETARY: Formatting of monetary amounts
728 The C standard defines the C<LC_MONETARY> category, but not a function
729 that is affected by its contents. (Those with experience of standards
730 committees will recognize that the working group decided to punt on the
731 issue.) Consequently, Perl takes no notice of it. If you really want
732 to use C<LC_MONETARY>, you can query its contents--see
733 L<The localeconv function>--and use the information that it returns in your
734 application's own formatting of currency amounts. However, you may well
735 find that the information, voluminous and complex though it may be, still
736 does not quite meet your requirements: currency formatting is a hard nut
739 See also L<I18N::Langinfo> and C<CRNCYSTR>.
743 Output produced by POSIX::strftime(), which builds a formatted
744 human-readable date/time string, is affected by the current C<LC_TIME>
745 locale. Thus, in a French locale, the output produced by the C<%B>
746 format element (full month name) for the first month of the year would
747 be "janvier". Here's how to get a list of long month names in the
750 use POSIX qw(strftime);
752 $long_month_name[$_] =
753 strftime("%B", 0, 0, 0, 1, $_, 96);
756 Note: C<use locale> isn't needed in this example: as a function that
757 exists only to generate locale-dependent results, strftime() always
758 obeys the current C<LC_TIME> locale.
760 See also L<I18N::Langinfo> and C<ABDAY_1>..C<ABDAY_7>, C<DAY_1>..C<DAY_7>,
761 C<ABMON_1>..C<ABMON_12>, and C<ABMON_1>..C<ABMON_12>.
763 =head2 Other categories
765 The remaining locale category, C<LC_MESSAGES> (possibly supplemented
766 by others in particular implementations) is not currently used by
767 Perl--except possibly to affect the behavior of library functions
768 called by extensions outside the standard Perl distribution and by the
769 operating system and its utilities. Note especially that the string
770 value of C<$!> and the error messages given by external utilities may
771 be changed by C<LC_MESSAGES>. If you want to have portable error
772 codes, use C<%!>. See L<Errno>.
776 Although the main discussion of Perl security issues can be found in
777 L<perlsec>, a discussion of Perl's locale handling would be incomplete
778 if it did not draw your attention to locale-dependent security issues.
779 Locales--particularly on systems that allow unprivileged users to
780 build their own locales--are untrustworthy. A malicious (or just plain
781 broken) locale can make a locale-aware application give unexpected
782 results. Here are a few possibilities:
788 Regular expression checks for safe file names or mail addresses using
789 C<\w> may be spoofed by an C<LC_CTYPE> locale that claims that
790 characters such as "E<gt>" and "|" are alphanumeric.
794 String interpolation with case-mapping, as in, say, C<$dest =
795 "C:\U$name.$ext">, may produce dangerous results if a bogus LC_CTYPE
796 case-mapping table is in effect.
800 A sneaky C<LC_COLLATE> locale could result in the names of students with
801 "D" grades appearing ahead of those with "A"s.
805 An application that takes the trouble to use information in
806 C<LC_MONETARY> may format debits as if they were credits and vice versa
807 if that locale has been subverted. Or it might make payments in US
808 dollars instead of Hong Kong dollars.
812 The date and day names in dates formatted by strftime() could be
813 manipulated to advantage by a malicious user able to subvert the
814 C<LC_DATE> locale. ("Look--it says I wasn't in the building on
819 Such dangers are not peculiar to the locale system: any aspect of an
820 application's environment which may be modified maliciously presents
821 similar challenges. Similarly, they are not specific to Perl: any
822 programming language that allows you to write programs that take
823 account of their environment exposes you to these issues.
825 Perl cannot protect you from all possibilities shown in the
826 examples--there is no substitute for your own vigilance--but, when
827 C<use locale> is in effect, Perl uses the tainting mechanism (see
828 L<perlsec>) to mark string results that become locale-dependent, and
829 which may be untrustworthy in consequence. Here is a summary of the
830 tainting behavior of operators and functions that may be affected by
837 B<Comparison operators> (C<lt>, C<le>, C<ge>, C<gt> and C<cmp>):
839 Scalar true/false (or less/equal/greater) result is never tainted.
843 B<Case-mapping interpolation> (with C<\l>, C<\L>, C<\u> or C<\U>)
845 Result string containing interpolated material is tainted if
846 C<use locale> (but not S<C<use locale ':not_characters'>>) is in effect.
850 B<Matching operator> (C<m//>):
852 Scalar true/false result never tainted.
854 Subpatterns, either delivered as a list-context result or as $1 etc.
855 are tainted if C<use locale> (but not S<C<use locale ':not_characters'>>)
856 is in effect, and the subpattern regular
857 expression contains C<\w> (to match an alphanumeric character), C<\W>
858 (non-alphanumeric character), C<\s> (whitespace character), or C<\S>
859 (non whitespace character). The matched-pattern variable, $&, $`
860 (pre-match), $' (post-match), and $+ (last match) are also tainted if
861 C<use locale> is in effect and the regular expression contains C<\w>,
862 C<\W>, C<\s>, or C<\S>.
866 B<Substitution operator> (C<s///>):
868 Has the same behavior as the match operator. Also, the left
869 operand of C<=~> becomes tainted when C<use locale>
870 (but not S<C<use locale ':not_characters'>>) is in effect if modified as
871 a result of a substitution based on a regular
872 expression match involving C<\w>, C<\W>, C<\s>, or C<\S>; or of
873 case-mapping with C<\l>, C<\L>,C<\u> or C<\U>.
877 B<Output formatting functions> (printf() and write()):
879 Results are never tainted because otherwise even output from print,
880 for example C<print(1/7)>, should be tainted if C<use locale> is in
885 B<Case-mapping functions> (lc(), lcfirst(), uc(), ucfirst()):
887 Results are tainted if C<use locale> (but not
888 S<C<use locale ':not_characters'>>) is in effect.
892 B<POSIX locale-dependent functions> (localeconv(), strcoll(),
893 strftime(), strxfrm()):
895 Results are never tainted.
899 B<POSIX character class tests> (isalnum(), isalpha(), isdigit(),
900 isgraph(), islower(), isprint(), ispunct(), isspace(), isupper(),
903 True/false results are never tainted.
907 Three examples illustrate locale-dependent tainting.
908 The first program, which ignores its locale, won't run: a value taken
909 directly from the command line may not be used to name an output file
910 when taint checks are enabled.
912 #/usr/local/bin/perl -T
913 # Run with taint checking
915 # Command line sanity check omitted...
916 $tainted_output_file = shift;
918 open(F, ">$tainted_output_file")
919 or warn "Open of $tainted_output_file failed: $!\n";
921 The program can be made to run by "laundering" the tainted value through
922 a regular expression: the second example--which still ignores locale
923 information--runs, creating the file named on its command line
926 #/usr/local/bin/perl -T
928 $tainted_output_file = shift;
929 $tainted_output_file =~ m%[\w/]+%;
930 $untainted_output_file = $&;
932 open(F, ">$untainted_output_file")
933 or warn "Open of $untainted_output_file failed: $!\n";
935 Compare this with a similar but locale-aware program:
937 #/usr/local/bin/perl -T
939 $tainted_output_file = shift;
941 $tainted_output_file =~ m%[\w/]+%;
942 $localized_output_file = $&;
944 open(F, ">$localized_output_file")
945 or warn "Open of $localized_output_file failed: $!\n";
947 This third program fails to run because $& is tainted: it is the result
948 of a match involving C<\w> while C<use locale> is in effect.
956 A string that can suppress Perl's warning about failed locale settings
957 at startup. Failure can occur if the locale support in the operating
958 system is lacking (broken) in some way--or if you mistyped the name of
959 a locale when you set up your environment. If this environment
960 variable is absent, or has a value that does not evaluate to integer
961 zero--that is, "0" or ""-- Perl will complain about locale setting
964 B<NOTE>: PERL_BADLANG only gives you a way to hide the warning message.
965 The message tells about some problem in your system's locale support,
966 and you should investigate what the problem is.
970 The following environment variables are not specific to Perl: They are
971 part of the standardized (ISO C, XPG4, POSIX 1.c) setlocale() method
972 for controlling an application's opinion on data.
978 C<LC_ALL> is the "override-all" locale environment variable. If
979 set, it overrides all the rest of the locale environment variables.
983 B<NOTE>: C<LANGUAGE> is a GNU extension, it affects you only if you
984 are using the GNU libc. This is the case if you are using e.g. Linux.
985 If you are using "commercial" Unixes you are most probably I<not>
986 using GNU libc and you can ignore C<LANGUAGE>.
988 However, in the case you are using C<LANGUAGE>: it affects the
989 language of informational, warning, and error messages output by
990 commands (in other words, it's like C<LC_MESSAGES>) but it has higher
991 priority than C<LC_ALL>. Moreover, it's not a single value but
992 instead a "path" (":"-separated list) of I<languages> (not locales).
993 See the GNU C<gettext> library documentation for more information.
997 In the absence of C<LC_ALL>, C<LC_CTYPE> chooses the character type
998 locale. In the absence of both C<LC_ALL> and C<LC_CTYPE>, C<LANG>
999 chooses the character type locale.
1003 In the absence of C<LC_ALL>, C<LC_COLLATE> chooses the collation
1004 (sorting) locale. In the absence of both C<LC_ALL> and C<LC_COLLATE>,
1005 C<LANG> chooses the collation locale.
1009 In the absence of C<LC_ALL>, C<LC_MONETARY> chooses the monetary
1010 formatting locale. In the absence of both C<LC_ALL> and C<LC_MONETARY>,
1011 C<LANG> chooses the monetary formatting locale.
1015 In the absence of C<LC_ALL>, C<LC_NUMERIC> chooses the numeric format
1016 locale. In the absence of both C<LC_ALL> and C<LC_NUMERIC>, C<LANG>
1017 chooses the numeric format.
1021 In the absence of C<LC_ALL>, C<LC_TIME> chooses the date and time
1022 formatting locale. In the absence of both C<LC_ALL> and C<LC_TIME>,
1023 C<LANG> chooses the date and time formatting locale.
1027 C<LANG> is the "catch-all" locale environment variable. If it is set, it
1028 is used as the last resort after the overall C<LC_ALL> and the
1029 category-specific C<LC_...>.
1035 The LC_NUMERIC controls the numeric output:
1038 use POSIX qw(locale_h); # Imports setlocale() and the LC_ constants.
1039 setlocale(LC_NUMERIC, "fr_FR") or die "Pardon";
1040 printf "%g\n", 1.23; # If the "fr_FR" succeeded, probably shows 1,23.
1042 and also how strings are parsed by POSIX::strtod() as numbers:
1045 use POSIX qw(locale_h strtod);
1046 setlocale(LC_NUMERIC, "de_DE") or die "Entschuldigung";
1047 my $x = strtod("2,34") + 5;
1048 print $x, "\n"; # Probably shows 7,34.
1052 =head2 Backward compatibility
1054 Versions of Perl prior to 5.004 B<mostly> ignored locale information,
1055 generally behaving as if something similar to the C<"C"> locale were
1056 always in force, even if the program environment suggested otherwise
1057 (see L<The setlocale function>). By default, Perl still behaves this
1058 way for backward compatibility. If you want a Perl application to pay
1059 attention to locale information, you B<must> use the S<C<use locale>>
1060 pragma (see L<The use locale pragma>) or, in the unlikely event
1061 that you want to do so for just pattern matching, the
1062 C</l> regular expression modifier (see L<perlre/Character set
1063 modifiers>) to instruct it to do so.
1065 Versions of Perl from 5.002 to 5.003 did use the C<LC_CTYPE>
1066 information if available; that is, C<\w> did understand what
1067 were the letters according to the locale environment variables.
1068 The problem was that the user had no control over the feature:
1069 if the C library supported locales, Perl used them.
1071 =head2 I18N:Collate obsolete
1073 In versions of Perl prior to 5.004, per-locale collation was possible
1074 using the C<I18N::Collate> library module. This module is now mildly
1075 obsolete and should be avoided in new applications. The C<LC_COLLATE>
1076 functionality is now integrated into the Perl core language: One can
1077 use locale-specific scalar data completely normally with C<use locale>,
1078 so there is no longer any need to juggle with the scalar references of
1081 =head2 Sort speed and memory use impacts
1083 Comparing and sorting by locale is usually slower than the default
1084 sorting; slow-downs of two to four times have been observed. It will
1085 also consume more memory: once a Perl scalar variable has participated
1086 in any string comparison or sorting operation obeying the locale
1087 collation rules, it will take 3-15 times more memory than before. (The
1088 exact multiplier depends on the string's contents, the operating system
1089 and the locale.) These downsides are dictated more by the operating
1090 system's implementation of the locale system than by Perl.
1092 =head2 write() and LC_NUMERIC
1094 If a program's environment specifies an LC_NUMERIC locale and C<use
1095 locale> is in effect when the format is declared, the locale is used
1096 to specify the decimal point character in formatted output. Formatted
1097 output cannot be controlled by C<use locale> at the time when write()
1100 =head2 Freely available locale definitions
1102 The Unicode CLDR project extracts the POSIX portion of many of its
1103 locales, available at
1105 http://unicode.org/Public/cldr/latest/
1107 There is a large collection of locale definitions at:
1109 http://std.dkuug.dk/i18n/WG15-collection/locales/
1111 You should be aware that it is
1112 unsupported, and is not claimed to be fit for any purpose. If your
1113 system allows installation of arbitrary locales, you may find the
1114 definitions useful as they are, or as a basis for the development of
1117 =head2 I18n and l10n
1119 "Internationalization" is often abbreviated as B<i18n> because its first
1120 and last letters are separated by eighteen others. (You may guess why
1121 the internalin ... internaliti ... i18n tends to get abbreviated.) In
1122 the same way, "localization" is often abbreviated to B<l10n>.
1124 =head2 An imperfect standard
1126 Internationalization, as defined in the C and POSIX standards, can be
1127 criticized as incomplete, ungainly, and having too large a granularity.
1128 (Locales apply to a whole process, when it would arguably be more useful
1129 to have them apply to a single thread, window group, or whatever.) They
1130 also have a tendency, like standards groups, to divide the world into
1131 nations, when we all know that the world can equally well be divided
1132 into bankers, bikers, gamers, and so on.
1134 =head1 Unicode and UTF-8
1136 The support of Unicode is new starting from Perl version v5.6, and more fully
1137 implemented in version v5.8 and later. See L<perluniintro>. It is
1138 strongly recommended that when combining Unicode and locale (starting in
1141 use locale ':not_characters';
1143 When this form of the pragma is used, only the non-character portions of
1144 locales are used by Perl, for example C<LC_NUMERIC>. Perl assumes that
1145 you have translated all the characters it is to operate on into Unicode
1146 (actually the platform's native character set (ASCII or EBCDIC) plus
1147 Unicode). For data in files, this can conveniently be done by also
1152 This pragma arranges for all inputs from files to be translated into
1153 Unicode from the current locale as specified in the environment (see
1154 L</ENVIRONMENT>), and all outputs to files to be translated back
1155 into the locale. (See L<open>). On a per-filehandle basis, you can
1156 instead use the L<PerlIO::locale> module, or the L<Encode::Locale>
1157 module, both available from CPAN. The latter module also has methods to
1158 ease the handling of C<ARGV> and environment variables, and can be used
1159 on individual strings. Also, if you know that all your locales will be
1160 UTF-8, as many are these days, you can use the L<B<-C>|perlrun/-C>
1161 command line switch.
1163 This form of the pragma allows essentially seamless handling of locales
1164 with Unicode. The collation order will be Unicode's. It is strongly
1165 recommended that when you need to order and sort strings that you use
1166 the standard module L<Unicode::Collate> which gives much better results
1167 in many instances than you can get with the old-style locale handling.
1169 For pre-v5.16 Perls, or if you use the locale pragma without the
1170 C<:not_characters> parameter, Perl tries to work with both Unicode and
1171 locales--but there are problems.
1173 Perl does not handle multi-byte locales in this case, such as have been
1175 Asian languages, such as Big5 or Shift JIS. However, the increasingly
1176 common multi-byte UTF-8 locales, if properly implemented, may work
1177 reasonably well (depending on your C library implementation) in this
1178 form of the locale pragma, simply because both
1179 they and Perl store characters that take up multiple bytes the same way.
1180 However, some, if not most, C library implementations may not process
1181 the characters in the upper half of the Latin-1 range (128 - 255)
1182 properly under LC_CTYPE. To see if a character is a particular type
1183 under a locale, Perl uses the functions like C<isalnum()>. Your C
1184 library may not work for UTF-8 locales with those functions, instead
1185 only working under the newer wide library functions like C<iswalnum()>.
1187 Perl generally takes the tack to use locale rules on code points that can fit
1188 in a single byte, and Unicode rules for those that can't (though this
1189 isn't uniformly applied, see the note at the end of this section). This
1190 prevents many problems in locales that aren't UTF-8. Suppose the locale
1191 is ISO8859-7, Greek. The character at 0xD7 there is a capital Chi. But
1192 in the ISO8859-1 locale, Latin1, it is a multiplication sign. The POSIX
1193 regular expression character class C<[[:alpha:]]> will magically match
1194 0xD7 in the Greek locale but not in the Latin one.
1196 However, there are places where this breaks down. Certain constructs are
1197 for Unicode only, such as C<\p{Alpha}>. They assume that 0xD7 always has its
1198 Unicode meaning (or the equivalent on EBCDIC platforms). Since Latin1 is a
1199 subset of Unicode and 0xD7 is the multiplication sign in both Latin1 and
1200 Unicode, C<\p{Alpha}> will never match it, regardless of locale. A similar
1201 issue occurs with C<\N{...}>. It is therefore a bad idea to use C<\p{}> or
1202 C<\N{}> under plain C<use locale>--I<unless> you can guarantee that the
1203 locale will be a ISO8859-1. Use POSIX character classes instead.
1205 Another problem with this approach is that operations that cross the
1206 single byte/multiple byte boundary are not well-defined, and so are
1207 disallowed. (This boundary is between the codepoints at 255/256.).
1208 For example, lower casing LATIN CAPITAL LETTER Y WITH DIAERESIS (U+0178)
1209 should return LATIN SMALL LETTER Y WITH DIAERESIS (U+00FF). But in the
1210 Greek locale, for example, there is no character at 0xFF, and Perl
1211 has no way of knowing what the character at 0xFF is really supposed to
1212 represent. Thus it disallows the operation. In this mode, the
1213 lowercase of U+0178 is itself.
1215 The same problems ensue if you enable automatic UTF-8-ification of your
1216 standard file handles, default C<open()> layer, and C<@ARGV> on non-ISO8859-1,
1217 non-UTF-8 locales (by using either the B<-C> command line switch or the
1218 C<PERL_UNICODE> environment variable; see L<perlrun>).
1219 Things are read in as UTF-8, which would normally imply a Unicode
1220 interpretation, but the presence of a locale causes them to be interpreted
1221 in that locale instead. For example, a 0xD7 code point in the Unicode
1222 input, which should mean the multiplication sign, won't be interpreted by
1223 Perl that way under the Greek locale. This is not a problem
1224 I<provided> you make certain that all locales will always and only be either
1225 an ISO8859-1, or, if you don't have a deficient C library, a UTF-8 locale.
1227 Vendor locales are notoriously buggy, and it is difficult for Perl to test
1228 its locale-handling code because this interacts with code that Perl has no
1229 control over; therefore the locale-handling code in Perl may be buggy as
1230 well. (However, the Unicode-supplied locales should be better, and
1231 there is a feed back mechanism to correct any problems. See
1232 L</Freely available locale definitions>.)
1234 If you have Perl v5.16, the problems mentioned above go away if you use
1235 the C<:not_characters> parameter to the locale pragma (except for vendor
1236 bugs in the non-character portions). If you don't have v5.16, and you
1237 I<do> have locales that work, using them may be worthwhile for certain
1238 specific purposes, as long as you keep in mind the gotchas already
1239 mentioned. For example, if the collation for your locales works, it
1240 runs faster under locales than under L<Unicode::Collate>; and you gain
1241 access to such things as the local currency symbol and the names of the
1242 months and days of the week. (But to hammer home the point, in v5.16,
1243 you get this access without the downsides of locales by using the
1244 C<:not_characters> form of the pragma.)
1246 Note: The policy of using locale rules for code points that can fit in a
1247 byte, and Unicode rules for those that can't is not uniformly applied.
1248 Pre-v5.12, it was somewhat haphazard; in v5.12 it was applied fairly
1249 consistently to regular expression matching except for bracketed
1250 character classes; in v5.14 it was extended to all regex matches; and in
1251 v5.16 to the casing operations such as C<"\L"> and C<uc()>. For
1252 collation, in all releases, the system's C<strxfrm()> function is called,
1253 and whatever it does is what you get.
1257 =head2 Broken systems
1259 In certain systems, the operating system's locale support
1260 is broken and cannot be fixed or used by Perl. Such deficiencies can
1261 and will result in mysterious hangs and/or Perl core dumps when
1262 C<use locale> is in effect. When confronted with such a system,
1263 please report in excruciating detail to <F<perlbug@perl.org>>, and
1264 also contact your vendor: bug fixes may exist for these problems
1265 in your operating system. Sometimes such bug fixes are called an
1266 operating system upgrade.
1270 L<I18N::Langinfo>, L<perluniintro>, L<perlunicode>, L<open>,
1271 L<POSIX/isalnum>, L<POSIX/isalpha>,
1272 L<POSIX/isdigit>, L<POSIX/isgraph>, L<POSIX/islower>,
1273 L<POSIX/isprint>, L<POSIX/ispunct>, L<POSIX/isspace>,
1274 L<POSIX/isupper>, L<POSIX/isxdigit>, L<POSIX/localeconv>,
1275 L<POSIX/setlocale>, L<POSIX/strcoll>, L<POSIX/strftime>,
1276 L<POSIX/strtod>, L<POSIX/strxfrm>.
1280 Jarkko Hietaniemi's original F<perli18n.pod> heavily hacked by Dominic
1281 Dunlop, assisted by the perl5-porters. Prose worked over a bit by
1282 Tom Christiansen, and updated by Perl 5 porters.