This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
PATCH: [perl #38193] embedded perl always calls setlocale(LC_ALL,"")
[perl5.git] / pod / perllocale.pod
CommitLineData
66cbab2c
KW
1=encoding utf8
2
5f05dabc
PP
3=head1 NAME
4
b0c42ed9 5perllocale - Perl locale handling (internationalization and localization)
5f05dabc
PP
6
7=head1 DESCRIPTION
8
66cbab2c
KW
9In the beginning there was ASCII, the "American Standard Code for
10Information Interchange", which works quite well for Americans with
11their English alphabet and dollar-denominated currency. But it doesn't
12work so well even for other English speakers, who may use different
13currencies, such as the pound sterling (as the symbol for that currency
14is not in ASCII); and it's hopelessly inadequate for many of the
15thousands of the world's other languages.
16
17To address these deficiencies, the concept of locales was invented
18(formally the ISO C, XPG4, POSIX 1.c "locale system"). And applications
19were and are being written that use the locale mechanism. The process of
20making such an application take account of its users' preferences in
21these kinds of matters is called B<internationalization> (often
22abbreviated as B<i18n>); telling such an application about a particular
23set of preferences is known as B<localization> (B<l10n>).
24
39332f68 25Perl has been extended to support the locale system. This
66cbab2c
KW
26is controlled per application by using one pragma, one function call,
27and several environment variables.
28
29Unfortunately, there are quite a few deficiencies with the design (and
30often, the implementations) of locales, and their use for character sets
31has mostly been supplanted by Unicode (see L<perlunitut> for an
32introduction to that, and keep on reading here for how Unicode interacts
33with locales in Perl).
34
7ee2ae1e 35Perl continues to support the old locale system, and starting in v5.16,
66cbab2c
KW
36provides a hybrid way to use the Unicode character set, along with the
37other portions of locales that may not be so problematic.
38(Unicode is also creating C<CLDR>, the "Common Locale Data Repository",
39L<http://cldr.unicode.org/> which includes more types of information than
40are available in the POSIX locale system. At the time of this writing,
41there was no CPAN module that provides access to this XML-encoded data.
42However, many of its locales have the POSIX-only data extracted, and are
43available at L<http://unicode.org/Public/cldr/latest/>.)
44
45=head1 WHAT IS A LOCALE
46
47A locale is a set of data that describes various aspects of how various
48communities in the world categorize their world. These categories are
49broken down into the following types (some of which include a brief
50note here):
51
52=over
53
54=item Category LC_NUMERIC: Numeric formatting
55
56This indicates how numbers should be formatted for human readability,
57for example the character used as the decimal point.
58
59=item Category LC_MONETARY: Formatting of monetary amounts
60
61=for comment
62The nbsp below makes this look better
63
64E<160>
65
66=item Category LC_TIME: Date/Time formatting
67
68=for comment
69The nbsp below makes this look better
70
71E<160>
72
73=item Category LC_MESSAGES: Error and other messages
74
2619d284
KW
75This is used by Perl itself only for accessing operating system error
76messages via L<$!|perlvar/$ERRNO>.
66cbab2c
KW
77
78=item Category LC_COLLATE: Collation
79
76073c88 80This indicates the ordering of letters for comparison and sorting.
66cbab2c
KW
81In Latin alphabets, for example, "b", generally follows "a".
82
83=item Category LC_CTYPE: Character Types
84
85This indicates, for example if a character is an uppercase letter.
86
2619d284
KW
87=item Other categories
88
89Some platforms have other categories, dealing with such things as
90measurement units and paper sizes. None of these are used directly by
91Perl, but outside operations that Perl interacts with may use
92these. See L</Not within the scope of any use locale variant> below.
93
66cbab2c
KW
94=back
95
2619d284
KW
96More details on the categories used by Perl are given below in L</LOCALE
97CATEGORIES>.
66cbab2c
KW
98
99Together, these categories go a long way towards being able to customize
100a single program to run in many different locations. But there are
101deficiencies, so keep reading.
5f05dabc
PP
102
103=head1 PREPARING TO USE LOCALES
104
2619d284
KW
105Perl itself will not use locales unless specifically requested to (but
106again note that Perl may interact with code that does use them). Even
107if there is such a request, B<all> of the following must be true
b960a36e 108for it to work properly:
5f05dabc
PP
109
110=over 4
111
112=item *
113
114B<Your operating system must support the locale system>. If it does,
39332f68 115you should find that the C<setlocale()> function is a documented part of
5f05dabc
PP
116its C library.
117
118=item *
119
5a964f20 120B<Definitions for locales that you use must be installed>. You, or
14280422
DD
121your system administrator, must make sure that this is the case. The
122available locales, the location in which they are kept, and the manner
5a964f20
TC
123in which they are installed all vary from system to system. Some systems
124provide only a few, hard-wired locales and do not allow more to be
125added. Others allow you to add "canned" locales provided by the system
126supplier. Still others allow you or the system administrator to define
14280422 127and add arbitrary locales. (You may have to ask your supplier to
5a964f20 128provide canned locales that are not delivered with your operating
14280422 129system.) Read your system documentation for further illumination.
5f05dabc
PP
130
131=item *
132
133B<Perl must believe that the locale system is supported>. If it does,
134C<perl -V:d_setlocale> will say that the value for C<d_setlocale> is
135C<define>.
136
137=back
138
139If you want a Perl application to process and present your data
140according to a particular locale, the application code should include
2ae324a7 141the S<C<use locale>> pragma (see L<The use locale pragma>) where
5f05dabc
PP
142appropriate, and B<at least one> of the following must be true:
143
144=over 4
145
c052850d 146=item 1
5f05dabc 147
66cbab2c 148B<The locale-determining environment variables (see L</"ENVIRONMENT">)
5a964f20 149must be correctly set up> at the time the application is started, either
ef3087ec 150by yourself or by whomever set up your system account; or
5f05dabc 151
c052850d 152=item 2
5f05dabc 153
14280422
DD
154B<The application must set its own locale> using the method described in
155L<The setlocale function>.
5f05dabc
PP
156
157=back
158
159=head1 USING LOCALES
160
161=head2 The use locale pragma
162
2619d284 163By default, Perl itself ignores the current locale. The S<C<use locale>>
66cbab2c 164pragma tells Perl to use the current locale for some operations.
7ee2ae1e 165Starting in v5.16, there is an optional parameter to this pragma:
66cbab2c
KW
166
167 use locale ':not_characters';
168
169This parameter allows better mixing of locales and Unicode, and is
170described fully in L</Unicode and UTF-8>, but briefly, it tells Perl to
171not use the character portions of the locale definition, that is
172the C<LC_CTYPE> and C<LC_COLLATE> categories. Instead it will use the
2619d284 173native character set (extended by Unicode). When using this parameter,
66cbab2c
KW
174you are responsible for getting the external character set translated
175into the native/Unicode one (which it already will be if it is one of
176the increasingly popular UTF-8 locales). There are convenient ways of
177doing this, as described in L</Unicode and UTF-8>.
c052850d
KW
178
179The current locale is set at execution time by
180L<setlocale()|/The setlocale function> described below. If that function
181hasn't yet been called in the course of the program's execution, the
66cbab2c 182current locale is that which was determined by the L</"ENVIRONMENT"> in
c052850d
KW
183effect at the start of the program, except that
184C<L<LC_NUMERIC|/Category LC_NUMERIC: Numeric Formatting>> is always
185initialized to the C locale (mentioned under L<Finding locales>).
70709c68
KW
186If there is no valid environment, the current locale is undefined. It
187is likely, but not necessarily, the "C" locale.
c052850d
KW
188
189The operations that are affected by locale are:
5f05dabc
PP
190
191=over 4
192
b960a36e
KW
193=item B<Not within the scope of any C<use locale> variant>
194
2619d284
KW
195Only operations originating outside Perl should be affected.
196
197The variable L<$!|perlvar/$ERRNO> (and its synonyms C<$ERRNO> and
198C<$OS_ERROR>) when used as strings always are in terms of the current
199locale.
200
201The current locale is also used when going outside of Perl with
202operations like L<system()|perlfunc/system LIST> or
203L<qxE<sol>E<sol>|perlop/qxE<sol>STRINGE<sol>>, if those operations are
204locale-sensitive.
205
206Also Perl gives access to various C library functions through the
207L<POSIX> module. Some of those functions are always affected by the
208current locale. For example, C<POSIX::strftime()> uses C<LC_TIME>;
209C<POSIX::strtod()> uses C<LC_NUMERIC>; C<POSIX::strcoll()> and
210C<POSIX::strxfrm()> use C<LC_COLLATE>; and character classification
211functions like C<POSIX::isalnum()> use C<LC_CTYPE>. All such functions
212will behave according to the current underlying locale, even if that
213isn't exposed to Perl space.
214
215And, certain Perl operations that are set-up within the scope of a
b960a36e
KW
216C<use locale> variant retain that effect even outside the scope.
217These include:
218
219=over 4
220
221=item *
222
223The output format of a L<write()|perlfunc/write> is determined by an
224earlier format declaration (L<perlfunc/format>), so whether or not the
225output is affected by locale is determined by if the C<format()> is
226within the scope of a C<use locale> variant, not whether the C<write()>
227is.
228
229=item *
230
231Regular expression patterns can be compiled using
232L<qrE<sol>E<sol>|perlop/qrE<sol>STRINGE<sol>msixpodual> with actual
233matching deferred to later. Again, it is whether or not the compilation
234was done within the scope of C<use locale> that determines the match
235behavior, not if the matches are done within such a scope or not.
236
237=back
238
66cbab2c
KW
239=item B<Under C<use locale ':not_characters';>>
240
241=over 4
242
243=item *
244
b960a36e
KW
245All the non-Perl operations.
246
247=item *
248
249B<Format declarations> (L<perlfunc/format>) and hence any subsequent
250C<write()>s use C<LC_NUMERIC>.
66cbab2c
KW
251
252=item *
253
b960a36e
KW
254B<stringification and output> use C<LC_NUMERIC>.
255These include the results of
256C<print()>,
257C<printf()>,
258C<say()>,
259and
260C<sprintf()>.
66cbab2c
KW
261
262=back
263
264=for comment
265The nbsp below makes this look better
266
267E<160>
268
269=item B<Under just plain C<use locale;>>
270
66cbab2c
KW
271=over 4
272
5f05dabc
PP
273=item *
274
b960a36e
KW
275All the above operations
276
277=item *
278
279B<The comparison operators> (C<lt>, C<le>, C<cmp>, C<ge>, and C<gt>) use
39332f68 280C<LC_COLLATE>. C<sort()> is also affected if used without an
5a964f20 281explicit comparison function, because it uses C<cmp> by default.
14280422 282
5a964f20 283B<Note:> C<eq> and C<ne> are unaffected by locale: they always
de108802 284perform a char-by-char comparison of their scalar operands. What's
14280422
DD
285more, if C<cmp> finds that its operands are equal according to the
286collation sequence specified by the current locale, it goes on to
de108802
RGS
287perform a char-by-char comparison, and only returns I<0> (equal) if the
288operands are char-for-char identical. If you really want to know whether
5a964f20 289two strings--which C<eq> and C<cmp> may consider different--are equal
14280422
DD
290as far as collation in the locale is concerned, see the discussion in
291L<Category LC_COLLATE: Collation>.
5f05dabc
PP
292
293=item *
294
39332f68
KW
295B<Regular expressions and case-modification functions> (C<uc()>, C<lc()>,
296C<ucfirst()>, and C<lcfirst()>) use C<LC_CTYPE>
5f05dabc 297
5f05dabc
PP
298=back
299
66cbab2c 300=back
5f05dabc 301
5a964f20 302The default behavior is restored with the S<C<no locale>> pragma, or
ef3087ec 303upon reaching the end of the block enclosing C<use locale>.
66cbab2c
KW
304Note that C<use locale> and C<use locale ':not_characters'> may be
305nested, and that what is in effect within an inner scope will revert to
306the outer scope's rules at the end of the inner scope.
5f05dabc 307
5a964f20 308The string result of any operation that uses locale
14280422
DD
309information is tainted, as it is possible for a locale to be
310untrustworthy. See L<"SECURITY">.
5f05dabc
PP
311
312=head2 The setlocale function
313
14280422 314You can switch locales as often as you wish at run time with the
39332f68 315C<POSIX::setlocale()> function:
5f05dabc 316
5f05dabc
PP
317 # Import locale-handling tool set from POSIX module.
318 # This example uses: setlocale -- the function call
319 # LC_CTYPE -- explained below
2619d284
KW
320 # (Showing the testing for success/failure of operations is
321 # omitted in these examples to avoid distracting from the main
6ea81ccf
KW
322 # point; also, any real work would almost certainly have to be
323 # done within the scope of "use locale".)
324
5f05dabc
PP
325 use POSIX qw(locale_h);
326
14280422 327 # query and save the old locale
5f05dabc
PP
328 $old_locale = setlocale(LC_CTYPE);
329
330 setlocale(LC_CTYPE, "fr_CA.ISO8859-1");
331 # LC_CTYPE now in locale "French, Canada, codeset ISO 8859-1"
332
333 setlocale(LC_CTYPE, "");
334 # LC_CTYPE now reset to default defined by LC_ALL/LC_CTYPE/LANG
335 # environment variables. See below for documentation.
336
337 # restore the old locale
338 setlocale(LC_CTYPE, $old_locale);
339
39332f68 340The first argument of C<setlocale()> gives the B<category>, the second the
14280422
DD
341B<locale>. The category tells in what aspect of data processing you
342want to apply locale-specific rules. Category names are discussed in
66cbab2c 343L</LOCALE CATEGORIES> and L</"ENVIRONMENT">. The locale is the name of a
14280422
DD
344collection of customization information corresponding to a particular
345combination of language, country or territory, and codeset. Read on for
346hints on the naming of locales: not all systems name locales as in the
347example.
348
39332f68 349If no second argument is provided and the category is something other
502a173a
JH
350than LC_ALL, the function returns a string naming the current locale
351for the category. You can use this value as the second argument in a
39332f68 352subsequent call to C<setlocale()>.
502a173a
JH
353
354If no second argument is provided and the category is LC_ALL, the
355result is implementation-dependent. It may be a string of
c052850d 356concatenated locale names (separator also implementation-dependent)
39332f68 357or a single locale name. Please consult your L<setlocale(3)> man page for
502a173a
JH
358details.
359
360If a second argument is given and it corresponds to a valid locale,
361the locale for the category is set to that value, and the function
362returns the now-current locale value. You can then use this in yet
39332f68 363another call to C<setlocale()>. (In some implementations, the return
502a173a
JH
364value may sometimes differ from the value you gave as the second
365argument--think of it as an alias for the value you gave.)
5f05dabc
PP
366
367As the example shows, if the second argument is an empty string, the
368category's locale is returned to the default specified by the
369corresponding environment variables. Generally, this results in a
5a964f20 370return to the default that was in force when Perl started up: changes
54310121 371to the environment made by the application after startup may or may not
5a964f20 372be noticed, depending on your system's C library.
5f05dabc 373
14280422 374If the second argument does not correspond to a valid locale, the locale
39332f68 375for the category is not changed, and the function returns C<undef>.
5f05dabc 376
66cbab2c
KW
377Note that Perl ignores the current C<LC_CTYPE> and C<LC_COLLATE> locales
378within the scope of a C<use locale ':not_characters'>.
379
2619d284
KW
380If C<set_locale()> fails for some reason (for example an attempt to set
381to a locale unknown to the system), C<undef> is returned.
382
39332f68 383For further information about the categories, consult L<setlocale(3)>.
3e6e419a
JH
384
385=head2 Finding locales
386
39332f68 387For locales available in your system, consult also L<setlocale(3)> to
5a964f20
TC
388see whether it leads to the list of available locales (search for the
389I<SEE ALSO> section). If that fails, try the following command lines:
5f05dabc
PP
390
391 locale -a
392
393 nlsinfo
394
395 ls /usr/lib/nls/loc
396
397 ls /usr/lib/locale
398
399 ls /usr/lib/nls
400
b478f28d
JH
401 ls /usr/share/locale
402
5f05dabc
PP
403and see whether they list something resembling these
404
2bdf8add 405 en_US.ISO8859-1 de_DE.ISO8859-1 ru_RU.ISO8859-5
502a173a 406 en_US.iso88591 de_DE.iso88591 ru_RU.iso88595
2bdf8add 407 en_US de_DE ru_RU
14280422 408 en de ru
2bdf8add
JH
409 english german russian
410 english.iso88591 german.iso88591 russian.iso88595
502a173a 411 english.roman8 russian.koi8r
5f05dabc 412
39332f68 413Sadly, even though the calling interface for C<setlocale()> has been
528d65ad 414standardized, names of locales and the directories where the
5a964f20 415configuration resides have not been. The basic form of the name is
528d65ad
JH
416I<language_territory>B<.>I<codeset>, but the latter parts after
417I<language> are not always present. The I<language> and I<country>
418are usually from the standards B<ISO 3166> and B<ISO 639>, the
419two-letter abbreviations for the countries and the languages of the
420world, respectively. The I<codeset> part often mentions some B<ISO
4218859> character set, the Latin codesets. For example, C<ISO 8859-1>
422is the so-called "Western European codeset" that can be used to encode
423most Western European languages adequately. Again, there are several
424ways to write even the name of that one standard. Lamentably.
5f05dabc 425
14280422
DD
426Two special locales are worth particular mention: "C" and "POSIX".
427Currently these are effectively the same locale: the difference is
5a964f20
TC
428mainly that the first one is defined by the C standard, the second by
429the POSIX standard. They define the B<default locale> in which
14280422 430every program starts in the absence of locale information in its
5a964f20 431environment. (The I<default> default locale, if you will.) Its language
39332f68
KW
432is (American) English and its character codeset ASCII or, rarely, a
433superset thereof (such as the "DEC Multinational Character Set
434(DEC-MCS)"). B<Warning>. The C locale delivered by some vendors
435may not actually exactly match what the C standard calls for. So
436beware.
5f05dabc 437
14280422
DD
438B<NOTE>: Not all systems have the "POSIX" locale (not all systems are
439POSIX-conformant), so use "C" when you need explicitly to specify this
440default locale.
5f05dabc 441
3e6e419a
JH
442=head2 LOCALE PROBLEMS
443
5a964f20 444You may encounter the following warning message at Perl startup:
3e6e419a
JH
445
446 perl: warning: Setting locale failed.
447 perl: warning: Please check that your locale settings:
448 LC_ALL = "En_US",
449 LANG = (unset)
450 are supported and installed on your system.
451 perl: warning: Falling back to the standard locale ("C").
452
5a964f20
TC
453This means that your locale settings had LC_ALL set to "En_US" and
454LANG exists but has no value. Perl tried to believe you but could not.
455Instead, Perl gave up and fell back to the "C" locale, the default locale
456that is supposed to work no matter what. This usually means your locale
457settings were wrong, they mention locales your system has never heard
458of, or the locale installation in your system has problems (for example,
459some system files are broken or missing). There are quick and temporary
460fixes to these problems, as well as more thorough and lasting fixes.
3e6e419a
JH
461
462=head2 Temporarily fixing locale problems
463
5a964f20 464The two quickest fixes are either to render Perl silent about any
3e6e419a
JH
465locale inconsistencies or to run Perl under the default locale "C".
466
467Perl's moaning about locale problems can be silenced by setting the
900bd440
JH
468environment variable PERL_BADLANG to a zero value, for example "0".
469This method really just sweeps the problem under the carpet: you tell
470Perl to shut up even when Perl sees that something is wrong. Do not
471be surprised if later something locale-dependent misbehaves.
3e6e419a
JH
472
473Perl can be run under the "C" locale by setting the environment
5a964f20
TC
474variable LC_ALL to "C". This method is perhaps a bit more civilized
475than the PERL_BADLANG approach, but setting LC_ALL (or
476other locale variables) may affect other programs as well, not just
477Perl. In particular, external programs run from within Perl will see
3e6e419a 478these changes. If you make the new settings permanent (read on), all
f979aebc 479programs you run see the changes. See L<"ENVIRONMENT"> for
5a964f20 480the full list of relevant environment variables and L<USING LOCALES>
e05ffc7d 481for their effects in Perl. Effects in other programs are
5a964f20 482easily deducible. For example, the variable LC_COLLATE may well affect
b432a672 483your B<sort> program (or whatever the program that arranges "records"
3e6e419a
JH
484alphabetically in your system is called).
485
5a964f20
TC
486You can test out changing these variables temporarily, and if the
487new settings seem to help, put those settings into your shell startup
488files. Consult your local documentation for the exact details. For in
489Bourne-like shells (B<sh>, B<ksh>, B<bash>, B<zsh>):
3e6e419a
JH
490
491 LC_ALL=en_US.ISO8859-1
492 export LC_ALL
493
5a964f20
TC
494This assumes that we saw the locale "en_US.ISO8859-1" using the commands
495discussed above. We decided to try that instead of the above faulty
496locale "En_US"--and in Cshish shells (B<csh>, B<tcsh>)
3e6e419a
JH
497
498 setenv LC_ALL en_US.ISO8859-1
c47ff5f1 499
c406981e
JH
500or if you have the "env" application you can do in any shell
501
502 env LC_ALL=en_US.ISO8859-1 perl ...
503
5a964f20 504If you do not know what shell you have, consult your local
3e6e419a
JH
505helpdesk or the equivalent.
506
507=head2 Permanently fixing locale problems
508
5a964f20
TC
509The slower but superior fixes are when you may be able to yourself
510fix the misconfiguration of your own environment variables. The
3e6e419a
JH
511mis(sing)configuration of the whole system's locales usually requires
512the help of your friendly system administrator.
513
5a964f20
TC
514First, see earlier in this document about L<Finding locales>. That tells
515how to find which locales are really supported--and more importantly,
516installed--on your system. In our example error message, environment
517variables affecting the locale are listed in the order of decreasing
518importance (and unset variables do not matter). Therefore, having
519LC_ALL set to "En_US" must have been the bad choice, as shown by the
520error message. First try fixing locale settings listed first.
3e6e419a 521
5a964f20
TC
522Second, if using the listed commands you see something B<exactly>
523(prefix matches do not count and case usually counts) like "En_US"
524without the quotes, then you should be okay because you are using a
525locale name that should be installed and available in your system.
4a4eefd0 526In this case, see L<Permanently fixing your system's locale configuration>.
3e6e419a 527
4a4eefd0 528=head2 Permanently fixing your system's locale configuration
3e6e419a 529
5a964f20 530This is when you see something like:
3e6e419a
JH
531
532 perl: warning: Please check that your locale settings:
533 LC_ALL = "En_US",
534 LANG = (unset)
535 are supported and installed on your system.
536
537but then cannot see that "En_US" listed by the above-mentioned
5a964f20
TC
538commands. You may see things like "en_US.ISO8859-1", but that isn't
539the same. In this case, try running under a locale
540that you can list and which somehow matches what you tried. The
3e6e419a 541rules for matching locale names are a bit vague because
e05ffc7d 542standardization is weak in this area. See again the
13a2d996 543L<Finding locales> about general rules.
3e6e419a 544
b687b08b 545=head2 Fixing system locale configuration
3e6e419a 546
5a964f20
TC
547Contact a system administrator (preferably your own) and report the exact
548error message you get, and ask them to read this same documentation you
549are now reading. They should be able to check whether there is something
550wrong with the locale configuration of the system. The L<Finding locales>
551section is unfortunately a bit vague about the exact commands and places
552because these things are not that standardized.
3e6e419a 553
5f05dabc
PP
554=head2 The localeconv function
555
39332f68 556The C<POSIX::localeconv()> function allows you to get particulars of the
14280422
DD
557locale-dependent numeric formatting information specified by the current
558C<LC_NUMERIC> and C<LC_MONETARY> locales. (If you just want the name of
39332f68 559the current locale for a particular category, use C<POSIX::setlocale()>
5a964f20 560with a single parameter--see L<The setlocale function>.)
5f05dabc
PP
561
562 use POSIX qw(locale_h);
5f05dabc
PP
563
564 # Get a reference to a hash of locale-dependent info
565 $locale_values = localeconv();
566
567 # Output sorted list of the values
568 for (sort keys %$locale_values) {
14280422 569 printf "%-20s = %s\n", $_, $locale_values->{$_}
5f05dabc
PP
570 }
571
39332f68 572C<localeconv()> takes no arguments, and returns B<a reference to> a hash.
5a964f20 573The keys of this hash are variable names for formatting, such as
502a173a 574C<decimal_point> and C<thousands_sep>. The values are the
cea6626f 575corresponding, er, values. See L<POSIX/localeconv> for a longer
502a173a
JH
576example listing the categories an implementation might be expected to
577provide; some provide more and others fewer. You don't need an
39332f68 578explicit C<use locale>, because C<localeconv()> always observes the
502a173a 579current locale.
5f05dabc 580
5a964f20
TC
581Here's a simple-minded example program that rewrites its command-line
582parameters as integers correctly formatted in the current locale:
5f05dabc 583
ef3087ec
KW
584 use POSIX qw(locale_h);
585
586 # Get some of locale's numeric formatting parameters
587 my ($thousands_sep, $grouping) =
588 @{localeconv()}{'thousands_sep', 'grouping'};
589
590 # Apply defaults if values are missing
591 $thousands_sep = ',' unless $thousands_sep;
592
593 # grouping and mon_grouping are packed lists
594 # of small integers (characters) telling the
595 # grouping (thousand_seps and mon_thousand_seps
596 # being the group dividers) of numbers and
597 # monetary quantities. The integers' meanings:
598 # 255 means no more grouping, 0 means repeat
599 # the previous grouping, 1-254 means use that
600 # as the current grouping. Grouping goes from
601 # right to left (low to high digits). In the
602 # below we cheat slightly by never using anything
603 # else than the first grouping (whatever that is).
604 if ($grouping) {
605 @grouping = unpack("C*", $grouping);
606 } else {
607 @grouping = (3);
608 }
609
610 # Format command line params for current locale
611 for (@ARGV) {
612 $_ = int; # Chop non-integer part
613 1 while
614 s/(\d)(\d{$grouping[0]}($|$thousands_sep))/$1$thousands_sep$2/;
615 print "$_";
616 }
617 print "\n";
5f05dabc 618
74c76037 619=head2 I18N::Langinfo
4bbcc6e8
JH
620
621Another interface for querying locale-dependent information is the
39332f68 622C<I18N::Langinfo::langinfo()> function, available at least in Unix-like
4bbcc6e8
JH
623systems and VMS.
624
39332f68
KW
625The following example will import the C<langinfo()> function itself and
626three constants to be used as arguments to C<langinfo()>: a constant for
74c76037
JH
627the abbreviated first day of the week (the numbering starts from
628Sunday = 1) and two more constants for the affirmative and negative
629answers for a yes/no question in the current locale.
4bbcc6e8 630
74c76037 631 use I18N::Langinfo qw(langinfo ABDAY_1 YESSTR NOSTR);
4bbcc6e8 632
ef3087ec
KW
633 my ($abday_1, $yesstr, $nostr)
634 = map { langinfo } qw(ABDAY_1 YESSTR NOSTR);
4bbcc6e8 635
74c76037 636 print "$abday_1? [$yesstr/$nostr] ";
4bbcc6e8 637
74c76037
JH
638In other words, in the "C" (or English) locale the above will probably
639print something like:
640
e05ffc7d 641 Sun? [yes/no]
4bbcc6e8
JH
642
643See L<I18N::Langinfo> for more information.
644
5f05dabc
PP
645=head1 LOCALE CATEGORIES
646
5a964f20
TC
647The following subsections describe basic locale categories. Beyond these,
648some combination categories allow manipulation of more than one
649basic category at a time. See L<"ENVIRONMENT"> for a discussion of these.
5f05dabc
PP
650
651=head2 Category LC_COLLATE: Collation
652
66cbab2c
KW
653In the scope of S<C<use locale>> (but not a
654C<use locale ':not_characters'>), Perl looks to the C<LC_COLLATE>
5a964f20 655environment variable to determine the application's notions on collation
b4ffc3db
TC
656(ordering) of characters. For example, "b" follows "a" in Latin
657alphabets, but where do "E<aacute>" and "E<aring>" belong? And while
f87fa335 658"color" follows "chocolate" in English, what about in traditional Spanish?
5f05dabc 659
60f0fa02
JH
660The following collations all make sense and you may meet any of them
661if you "use locale".
662
663 A B C D E a b c d e
35316ca3 664 A a B b C c D d E e
60f0fa02
JH
665 a A b B c C d D e E
666 a b c d e A B C D E
667
f1cbbd6e 668Here is a code snippet to tell what "word"
5a964f20 669characters are in the current locale, in that locale's order:
5f05dabc
PP
670
671 use locale;
35316ca3 672 print +(sort grep /\w/, map { chr } 0..255), "\n";
5f05dabc 673
14280422
DD
674Compare this with the characters that you see and their order if you
675state explicitly that the locale should be ignored:
5f05dabc
PP
676
677 no locale;
35316ca3 678 print +(sort grep /\w/, map { chr } 0..255), "\n";
5f05dabc
PP
679
680This machine-native collation (which is what you get unless S<C<use
681locale>> has appeared earlier in the same block) must be used for
682sorting raw binary data, whereas the locale-dependent collation of the
b0c42ed9 683first example is useful for natural text.
5f05dabc 684
14280422
DD
685As noted in L<USING LOCALES>, C<cmp> compares according to the current
686collation locale when C<use locale> is in effect, but falls back to a
de108802 687char-by-char comparison for strings that the locale says are equal. You
39332f68 688can use C<POSIX::strcoll()> if you don't want this fall-back:
14280422
DD
689
690 use POSIX qw(strcoll);
691 $equal_in_locale =
692 !strcoll("space and case ignored", "SpaceAndCaseIgnored");
693
39332f68 694C<$equal_in_locale> will be true if the collation locale specifies a
5a964f20 695dictionary-like ordering that ignores space characters completely and
9e3a2af8 696which folds case.
14280422 697
5a964f20 698If you have a single string that you want to check for "equality in
14280422 699locale" against several others, you might think you could gain a little
39332f68 700efficiency by using C<POSIX::strxfrm()> in conjunction with C<eq>:
14280422
DD
701
702 use POSIX qw(strxfrm);
703 $xfrm_string = strxfrm("Mixed-case string");
704 print "locale collation ignores spaces\n"
705 if $xfrm_string eq strxfrm("Mixed-casestring");
706 print "locale collation ignores hyphens\n"
707 if $xfrm_string eq strxfrm("Mixedcase string");
708 print "locale collation ignores case\n"
709 if $xfrm_string eq strxfrm("mixed-case string");
710
39332f68 711C<strxfrm()> takes a string and maps it into a transformed string for use
de108802 712in char-by-char comparisons against other transformed strings during
14280422 713collation. "Under the hood", locale-affected Perl comparison operators
39332f68
KW
714call C<strxfrm()> for both operands, then do a char-by-char
715comparison of the transformed strings. By calling C<strxfrm()> explicitly
14280422 716and using a non locale-affected comparison, the example attempts to save
5a964f20 717a couple of transformations. But in fact, it doesn't save anything: Perl
2ae324a7 718magic (see L<perlguts/Magic Variables>) creates the transformed version of a
5a964f20 719string the first time it's needed in a comparison, then keeps this version around
14280422 720in case it's needed again. An example rewritten the easy way with
e38874e2 721C<cmp> runs just about as fast. It also copes with null characters
39332f68 722embedded in strings; if you call C<strxfrm()> directly, it treats the first
5a964f20
TC
723null it finds as a terminator. don't expect the transformed strings
724it produces to be portable across systems--or even from one revision
39332f68 725of your operating system to the next. In short, don't call C<strxfrm()>
e38874e2 726directly: let Perl do it for you.
14280422 727
5a964f20 728Note: C<use locale> isn't shown in some of these examples because it isn't
39332f68 729needed: C<strcoll()> and C<strxfrm()> exist only to generate locale-dependent
14280422 730results, and so always obey the current C<LC_COLLATE> locale.
5f05dabc
PP
731
732=head2 Category LC_CTYPE: Character Types
733
66cbab2c
KW
734In the scope of S<C<use locale>> (but not a
735C<use locale ':not_characters'>), Perl obeys the C<LC_CTYPE> locale
14280422
DD
736setting. This controls the application's notion of which characters are
737alphabetic. This affects Perl's C<\w> regular expression metanotation,
f1cbbd6e
GS
738which stands for alphanumeric characters--that is, alphabetic,
739numeric, and including other special characters such as the underscore or
740hyphen. (Consult L<perlre> for more information about
14280422 741regular expressions.) Thanks to C<LC_CTYPE>, depending on your locale
b4ffc3db
TC
742setting, characters like "E<aelig>", "E<eth>", "E<szlig>", and
743"E<oslash>" may be understood as C<\w> characters.
5f05dabc 744
2c268ad5 745The C<LC_CTYPE> locale also provides the map used in transliterating
68dc0745 746characters between lower and uppercase. This affects the case-mapping
39332f68 747functions--C<fc()>, C<lc()>, C<lcfirst()>, C<uc()>, and C<ucfirst()>; case-mapping
b9cc4f69
KW
748interpolation with C<\F>, C<\l>, C<\L>, C<\u>, or C<\U> in double-quoted
749strings and C<s///> substitutions; and case-independent regular expression
e38874e2
DD
750pattern matching using the C<i> modifier.
751
5a964f20 752Finally, C<LC_CTYPE> affects the POSIX character-class test
39332f68 753functions--C<isalpha()>, C<islower()>, and so on. For example, if you move
5a964f20 754from the "C" locale to a 7-bit Scandinavian one, you may find--possibly
39332f68 755to your surprise--that "|" moves from the C<ispunct()> class to C<isalpha()>.
ef3087ec
KW
756Unfortunately, this creates big problems for regular expressions. "|" still
757means alternation even though it matches C<\w>.
5f05dabc 758
5d63e270
KW
759Note that there are quite a few things that are unaffected by the
760current locale. All the escape sequences for particular characters,
761C<\n> for example, always mean the platform's native one. This means,
762for example, that C<\N> in regular expressions (every character
763but new-line) work on the platform character set.
764
14280422
DD
765B<Note:> A broken or malicious C<LC_CTYPE> locale definition may result
766in clearly ineligible characters being considered to be alphanumeric by
e199995e 767your application. For strict matching of (mundane) ASCII letters and
5a964f20 768digits--for example, in command strings--locale-aware applications
e199995e 769should use C<\w> with the C</a> regular expression modifier. See L<"SECURITY">.
5f05dabc
PP
770
771=head2 Category LC_NUMERIC: Numeric Formatting
772
b960a36e
KW
773After a proper C<POSIX::setlocale()> call, and within the scope of one
774of the C<use locale> variants, Perl obeys the C<LC_NUMERIC>
2095dafa 775locale information, which controls an application's idea of how numbers
b960a36e
KW
776should be formatted for human readability.
777In most implementations the only effect is to
b4ffc3db 778change the character used for the decimal point--perhaps from "." to ",".
b960a36e 779The functions aren't aware of such niceties as thousands separation and
2095dafa 780so on. (See L<The localeconv function> if you care about these things.)
5a964f20 781
b960a36e
KW
782 use POSIX qw(strtod setlocale LC_NUMERIC);
783 use locale;
5f05dabc 784
b960a36e 785 setlocale LC_NUMERIC, "";
14280422 786
b960a36e 787 $n = 5/2; # Assign numeric 2.5 to $n
5f05dabc 788
b960a36e 789 $a = " $n"; # Locale-dependent conversion to string
5f05dabc 790
b960a36e 791 print "half five is $n\n"; # Locale-dependent output
5f05dabc 792
b960a36e 793 printf "half five is %g\n", $n; # Locale-dependent output
5f05dabc 794
b960a36e
KW
795 print "DECIMAL POINT IS COMMA\n"
796 if $n == (strtod("2,5"))[0]; # Locale-dependent conversion
5f05dabc 797
4bbcc6e8
JH
798See also L<I18N::Langinfo> and C<RADIXCHAR>.
799
5f05dabc
PP
800=head2 Category LC_MONETARY: Formatting of monetary amounts
801
e199995e 802The C standard defines the C<LC_MONETARY> category, but not a function
5a964f20 803that is affected by its contents. (Those with experience of standards
b0c42ed9 804committees will recognize that the working group decided to punt on the
fa9b773e
KW
805issue.) Consequently, Perl essentially takes no notice of it. If you
806really want to use C<LC_MONETARY>, you can query its contents--see
e05ffc7d
KW
807L<The localeconv function>--and use the information that it returns in your
808application's own formatting of currency amounts. However, you may well
809find that the information, voluminous and complex though it may be, still
810does not quite meet your requirements: currency formatting is a hard nut
13a2d996 811to crack.
5f05dabc 812
4bbcc6e8
JH
813See also L<I18N::Langinfo> and C<CRNCYSTR>.
814
5f05dabc
PP
815=head2 LC_TIME
816
39332f68 817Output produced by C<POSIX::strftime()>, which builds a formatted
5f05dabc
PP
818human-readable date/time string, is affected by the current C<LC_TIME>
819locale. Thus, in a French locale, the output produced by the C<%B>
820format element (full month name) for the first month of the year would
5a964f20 821be "janvier". Here's how to get a list of long month names in the
5f05dabc
PP
822current locale:
823
824 use POSIX qw(strftime);
14280422
DD
825 for (0..11) {
826 $long_month_name[$_] =
827 strftime("%B", 0, 0, 0, 1, $_, 96);
5f05dabc
PP
828 }
829
2619d284
KW
830Note: C<use locale> isn't needed in this example: C<strftime()> is a POSIX
831function which uses the standard system-supplied C<libc> function that
832always obeys the current C<LC_TIME> locale.
5f05dabc 833
4bbcc6e8 834See also L<I18N::Langinfo> and C<ABDAY_1>..C<ABDAY_7>, C<DAY_1>..C<DAY_7>,
2a2bf5f4 835C<ABMON_1>..C<ABMON_12>, and C<ABMON_1>..C<ABMON_12>.
4bbcc6e8 836
5f05dabc
PP
837=head2 Other categories
838
2619d284
KW
839The remaining locale categories are not currently used by Perl itself.
840But again note that things Perl interacts with may use these, including
841extensions outside the standard Perl distribution, and by the
98a6f11e 842operating system and its utilities. Note especially that the string
843value of C<$!> and the error messages given by external utilities may
844be changed by C<LC_MESSAGES>. If you want to have portable error
265f5c4a 845codes, use C<%!>. See L<Errno>.
14280422
DD
846
847=head1 SECURITY
848
5a964f20 849Although the main discussion of Perl security issues can be found in
14280422
DD
850L<perlsec>, a discussion of Perl's locale handling would be incomplete
851if it did not draw your attention to locale-dependent security issues.
5a964f20
TC
852Locales--particularly on systems that allow unprivileged users to
853build their own locales--are untrustworthy. A malicious (or just plain
14280422
DD
854broken) locale can make a locale-aware application give unexpected
855results. Here are a few possibilities:
856
857=over 4
858
859=item *
860
861Regular expression checks for safe file names or mail addresses using
5a964f20 862C<\w> may be spoofed by an C<LC_CTYPE> locale that claims that
14280422
DD
863characters such as "E<gt>" and "|" are alphanumeric.
864
865=item *
866
e38874e2
DD
867String interpolation with case-mapping, as in, say, C<$dest =
868"C:\U$name.$ext">, may produce dangerous results if a bogus LC_CTYPE
869case-mapping table is in effect.
870
871=item *
872
14280422
DD
873A sneaky C<LC_COLLATE> locale could result in the names of students with
874"D" grades appearing ahead of those with "A"s.
875
876=item *
877
5a964f20 878An application that takes the trouble to use information in
14280422 879C<LC_MONETARY> may format debits as if they were credits and vice versa
5a964f20 880if that locale has been subverted. Or it might make payments in US
14280422
DD
881dollars instead of Hong Kong dollars.
882
883=item *
884
39332f68 885The date and day names in dates formatted by C<strftime()> could be
14280422 886manipulated to advantage by a malicious user able to subvert the
5a964f20 887C<LC_DATE> locale. ("Look--it says I wasn't in the building on
14280422
DD
888Sunday.")
889
890=back
891
892Such dangers are not peculiar to the locale system: any aspect of an
5a964f20 893application's environment which may be modified maliciously presents
14280422 894similar challenges. Similarly, they are not specific to Perl: any
5a964f20 895programming language that allows you to write programs that take
14280422
DD
896account of their environment exposes you to these issues.
897
5a964f20
TC
898Perl cannot protect you from all possibilities shown in the
899examples--there is no substitute for your own vigilance--but, when
14280422 900C<use locale> is in effect, Perl uses the tainting mechanism (see
5a964f20 901L<perlsec>) to mark string results that become locale-dependent, and
14280422 902which may be untrustworthy in consequence. Here is a summary of the
5a964f20 903tainting behavior of operators and functions that may be affected by
14280422
DD
904the locale:
905
906=over 4
907
551e1d92
RB
908=item *
909
910B<Comparison operators> (C<lt>, C<le>, C<ge>, C<gt> and C<cmp>):
14280422
DD
911
912Scalar true/false (or less/equal/greater) result is never tainted.
913
551e1d92
RB
914=item *
915
916B<Case-mapping interpolation> (with C<\l>, C<\L>, C<\u> or C<\U>)
e38874e2
DD
917
918Result string containing interpolated material is tainted if
66cbab2c 919C<use locale> (but not S<C<use locale ':not_characters'>>) is in effect.
e38874e2 920
551e1d92
RB
921=item *
922
923B<Matching operator> (C<m//>):
14280422
DD
924
925Scalar true/false result never tainted.
926
5a964f20 927Subpatterns, either delivered as a list-context result or as $1 etc.
66cbab2c
KW
928are tainted if C<use locale> (but not S<C<use locale ':not_characters'>>)
929is in effect, and the subpattern regular
e38874e2 930expression contains C<\w> (to match an alphanumeric character), C<\W>
6b0ac556
OK
931(non-alphanumeric character), C<\s> (whitespace character), or C<\S>
932(non whitespace character). The matched-pattern variable, $&, $`
e38874e2
DD
933(pre-match), $' (post-match), and $+ (last match) are also tainted if
934C<use locale> is in effect and the regular expression contains C<\w>,
935C<\W>, C<\s>, or C<\S>.
14280422 936
551e1d92
RB
937=item *
938
939B<Substitution operator> (C<s///>):
14280422 940
e38874e2 941Has the same behavior as the match operator. Also, the left
66cbab2c
KW
942operand of C<=~> becomes tainted when C<use locale>
943(but not S<C<use locale ':not_characters'>>) is in effect if modified as
944a result of a substitution based on a regular
e38874e2 945expression match involving C<\w>, C<\W>, C<\s>, or C<\S>; or of
7b8d334a 946case-mapping with C<\l>, C<\L>,C<\u> or C<\U>.
14280422 947
551e1d92
RB
948=item *
949
39332f68 950B<Output formatting functions> (C<printf()> and C<write()>):
14280422 951
3cf03d68
JH
952Results are never tainted because otherwise even output from print,
953for example C<print(1/7)>, should be tainted if C<use locale> is in
954effect.
14280422 955
551e1d92
RB
956=item *
957
39332f68 958B<Case-mapping functions> (C<lc()>, C<lcfirst()>, C<uc()>, C<ucfirst()>):
14280422 959
66cbab2c
KW
960Results are tainted if C<use locale> (but not
961S<C<use locale ':not_characters'>>) is in effect.
14280422 962
551e1d92
RB
963=item *
964
39332f68
KW
965B<POSIX locale-dependent functions> (C<localeconv()>, C<strcoll()>,
966C<strftime()>, C<strxfrm()>):
14280422
DD
967
968Results are never tainted.
969
551e1d92
RB
970=item *
971
39332f68
KW
972B<POSIX character class tests> (C<isalnum()>, C<isalpha()>, C<isdigit()>,
973C<isgraph()>, C<islower()>, C<isprint()>, C<ispunct()>, C<isspace()>, C<isupper()>,
974C<isxdigit()>):
14280422
DD
975
976True/false results are never tainted.
977
978=back
979
980Three examples illustrate locale-dependent tainting.
981The first program, which ignores its locale, won't run: a value taken
54310121 982directly from the command line may not be used to name an output file
14280422
DD
983when taint checks are enabled.
984
985 #/usr/local/bin/perl -T
986 # Run with taint checking
987
54310121 988 # Command line sanity check omitted...
14280422
DD
989 $tainted_output_file = shift;
990
991 open(F, ">$tainted_output_file")
3183d96c 992 or warn "Open of $tainted_output_file failed: $!\n";
14280422
DD
993
994The program can be made to run by "laundering" the tainted value through
5a964f20
TC
995a regular expression: the second example--which still ignores locale
996information--runs, creating the file named on its command line
14280422
DD
997if it can.
998
999 #/usr/local/bin/perl -T
1000
1001 $tainted_output_file = shift;
1002 $tainted_output_file =~ m%[\w/]+%;
1003 $untainted_output_file = $&;
1004
1005 open(F, ">$untainted_output_file")
1006 or warn "Open of $untainted_output_file failed: $!\n";
1007
5a964f20 1008Compare this with a similar but locale-aware program:
14280422
DD
1009
1010 #/usr/local/bin/perl -T
1011
1012 $tainted_output_file = shift;
1013 use locale;
1014 $tainted_output_file =~ m%[\w/]+%;
1015 $localized_output_file = $&;
1016
1017 open(F, ">$localized_output_file")
1018 or warn "Open of $localized_output_file failed: $!\n";
1019
1020This third program fails to run because $& is tainted: it is the result
5a964f20 1021of a match involving C<\w> while C<use locale> is in effect.
5f05dabc
PP
1022
1023=head1 ENVIRONMENT
1024
1025=over 12
1026
1027=item PERL_BADLANG
1028
14280422 1029A string that can suppress Perl's warning about failed locale settings
54310121 1030at startup. Failure can occur if the locale support in the operating
5a964f20 1031system is lacking (broken) in some way--or if you mistyped the name of
900bd440
JH
1032a locale when you set up your environment. If this environment
1033variable is absent, or has a value that does not evaluate to integer
1034zero--that is, "0" or ""-- Perl will complain about locale setting
1035failures.
5f05dabc 1036
14280422
DD
1037B<NOTE>: PERL_BADLANG only gives you a way to hide the warning message.
1038The message tells about some problem in your system's locale support,
1039and you should investigate what the problem is.
5f05dabc
PP
1040
1041=back
1042
1043The following environment variables are not specific to Perl: They are
39332f68 1044part of the standardized (ISO C, XPG4, POSIX 1.c) C<setlocale()> method
14280422 1045for controlling an application's opinion on data.
5f05dabc
PP
1046
1047=over 12
1048
1049=item LC_ALL
1050
5a964f20 1051C<LC_ALL> is the "override-all" locale environment variable. If
5f05dabc
PP
1052set, it overrides all the rest of the locale environment variables.
1053
528d65ad
JH
1054=item LANGUAGE
1055
1056B<NOTE>: C<LANGUAGE> is a GNU extension, it affects you only if you
1057are using the GNU libc. This is the case if you are using e.g. Linux.
e1020413 1058If you are using "commercial" Unixes you are most probably I<not>
22b6f60d
JH
1059using GNU libc and you can ignore C<LANGUAGE>.
1060
1061However, in the case you are using C<LANGUAGE>: it affects the
1062language of informational, warning, and error messages output by
1063commands (in other words, it's like C<LC_MESSAGES>) but it has higher
96090e4f 1064priority than C<LC_ALL>. Moreover, it's not a single value but
22b6f60d
JH
1065instead a "path" (":"-separated list) of I<languages> (not locales).
1066See the GNU C<gettext> library documentation for more information.
528d65ad 1067
5f05dabc
PP
1068=item LC_CTYPE
1069
1070In the absence of C<LC_ALL>, C<LC_CTYPE> chooses the character type
1071locale. In the absence of both C<LC_ALL> and C<LC_CTYPE>, C<LANG>
1072chooses the character type locale.
1073
1074=item LC_COLLATE
1075
14280422
DD
1076In the absence of C<LC_ALL>, C<LC_COLLATE> chooses the collation
1077(sorting) locale. In the absence of both C<LC_ALL> and C<LC_COLLATE>,
1078C<LANG> chooses the collation locale.
5f05dabc
PP
1079
1080=item LC_MONETARY
1081
14280422
DD
1082In the absence of C<LC_ALL>, C<LC_MONETARY> chooses the monetary
1083formatting locale. In the absence of both C<LC_ALL> and C<LC_MONETARY>,
1084C<LANG> chooses the monetary formatting locale.
5f05dabc
PP
1085
1086=item LC_NUMERIC
1087
1088In the absence of C<LC_ALL>, C<LC_NUMERIC> chooses the numeric format
1089locale. In the absence of both C<LC_ALL> and C<LC_NUMERIC>, C<LANG>
1090chooses the numeric format.
1091
1092=item LC_TIME
1093
14280422
DD
1094In the absence of C<LC_ALL>, C<LC_TIME> chooses the date and time
1095formatting locale. In the absence of both C<LC_ALL> and C<LC_TIME>,
1096C<LANG> chooses the date and time formatting locale.
5f05dabc
PP
1097
1098=item LANG
1099
14280422
DD
1100C<LANG> is the "catch-all" locale environment variable. If it is set, it
1101is used as the last resort after the overall C<LC_ALL> and the
5f05dabc
PP
1102category-specific C<LC_...>.
1103
1104=back
1105
7e4353e9
RGS
1106=head2 Examples
1107
1108The LC_NUMERIC controls the numeric output:
1109
ef3087ec
KW
1110 use locale;
1111 use POSIX qw(locale_h); # Imports setlocale() and the LC_ constants.
1112 setlocale(LC_NUMERIC, "fr_FR") or die "Pardon";
1113 printf "%g\n", 1.23; # If the "fr_FR" succeeded, probably shows 1,23.
7e4353e9 1114
39332f68 1115and also how strings are parsed by C<POSIX::strtod()> as numbers:
7e4353e9 1116
ef3087ec
KW
1117 use locale;
1118 use POSIX qw(locale_h strtod);
1119 setlocale(LC_NUMERIC, "de_DE") or die "Entschuldigung";
1120 my $x = strtod("2,34") + 5;
1121 print $x, "\n"; # Probably shows 7,34.
7e4353e9 1122
5f05dabc
PP
1123=head1 NOTES
1124
b960a36e
KW
1125=head2 String C<eval> and C<LC_NUMERIC>
1126
1127A string L<eval|perlfunc/eval EXPR> parses its expression as standard
1128Perl. It is therefore expecting the decimal point to be a dot. If
1129C<LC_NUMERIC> is set to have this be a comma instead, the parsing will
1130be confused, perhaps silently.
1131
1132 use locale;
1133 use POSIX qw(locale_h);
1134 setlocale(LC_NUMERIC, "fr_FR") or die "Pardon";
1135 my $a = 1.2;
1136 print eval "$a + 1.5";
1137 print "\n";
1138
1139prints C<13,5>. This is because in that locale, the comma is the
1140decimal point character. The C<eval> thus expands to:
1141
1142 eval "1,2 + 1.5"
1143
1144and the result is not what you likely expected. No warnings are
1145generated. If you do string C<eval>'s within the scope of
1146S<C<use locale>>, you should instead change the C<eval> line to do
1147something like:
1148
1149 print eval "no locale; $a + 1.5";
1150
1151This prints C<2.7>.
1152
5f05dabc
PP
1153=head2 Backward compatibility
1154
b0c42ed9 1155Versions of Perl prior to 5.004 B<mostly> ignored locale information,
5a964f20
TC
1156generally behaving as if something similar to the C<"C"> locale were
1157always in force, even if the program environment suggested otherwise
1158(see L<The setlocale function>). By default, Perl still behaves this
1159way for backward compatibility. If you want a Perl application to pay
1160attention to locale information, you B<must> use the S<C<use locale>>
062ca197
KW
1161pragma (see L<The use locale pragma>) or, in the unlikely event
1162that you want to do so for just pattern matching, the
70709c68
KW
1163C</l> regular expression modifier (see L<perlre/Character set
1164modifiers>) to instruct it to do so.
b0c42ed9
JH
1165
1166Versions of Perl from 5.002 to 5.003 did use the C<LC_CTYPE>
5a964f20
TC
1167information if available; that is, C<\w> did understand what
1168were the letters according to the locale environment variables.
b0c42ed9
JH
1169The problem was that the user had no control over the feature:
1170if the C library supported locales, Perl used them.
1171
1172=head2 I18N:Collate obsolete
1173
5a964f20 1174In versions of Perl prior to 5.004, per-locale collation was possible
b0c42ed9
JH
1175using the C<I18N::Collate> library module. This module is now mildly
1176obsolete and should be avoided in new applications. The C<LC_COLLATE>
1177functionality is now integrated into the Perl core language: One can
1178use locale-specific scalar data completely normally with C<use locale>,
1179so there is no longer any need to juggle with the scalar references of
1180C<I18N::Collate>.
5f05dabc 1181
14280422 1182=head2 Sort speed and memory use impacts
5f05dabc
PP
1183
1184Comparing and sorting by locale is usually slower than the default
14280422
DD
1185sorting; slow-downs of two to four times have been observed. It will
1186also consume more memory: once a Perl scalar variable has participated
1187in any string comparison or sorting operation obeying the locale
1188collation rules, it will take 3-15 times more memory than before. (The
1189exact multiplier depends on the string's contents, the operating system
1190and the locale.) These downsides are dictated more by the operating
1191system's implementation of the locale system than by Perl.
5f05dabc 1192
5f05dabc
PP
1193=head2 Freely available locale definitions
1194
66cbab2c
KW
1195The Unicode CLDR project extracts the POSIX portion of many of its
1196locales, available at
1197
1198 http://unicode.org/Public/cldr/latest/
1199
08d7a6b2
LB
1200There is a large collection of locale definitions at:
1201
1202 http://std.dkuug.dk/i18n/WG15-collection/locales/
1203
1204You should be aware that it is
14280422 1205unsupported, and is not claimed to be fit for any purpose. If your
5a964f20 1206system allows installation of arbitrary locales, you may find the
14280422
DD
1207definitions useful as they are, or as a basis for the development of
1208your own locales.
5f05dabc 1209
14280422 1210=head2 I18n and l10n
5f05dabc 1211
b0c42ed9
JH
1212"Internationalization" is often abbreviated as B<i18n> because its first
1213and last letters are separated by eighteen others. (You may guess why
1214the internalin ... internaliti ... i18n tends to get abbreviated.) In
1215the same way, "localization" is often abbreviated to B<l10n>.
14280422
DD
1216
1217=head2 An imperfect standard
1218
1219Internationalization, as defined in the C and POSIX standards, can be
1220criticized as incomplete, ungainly, and having too large a granularity.
1221(Locales apply to a whole process, when it would arguably be more useful
1222to have them apply to a single thread, window group, or whatever.) They
1223also have a tendency, like standards groups, to divide the world into
1224nations, when we all know that the world can equally well be divided
e199995e 1225into bankers, bikers, gamers, and so on.
5f05dabc 1226
b310b053
JH
1227=head1 Unicode and UTF-8
1228
7ee2ae1e
KW
1229The support of Unicode is new starting from Perl version v5.6, and more fully
1230implemented in version v5.8 and later. See L<perluniintro>. It is
66cbab2c 1231strongly recommended that when combining Unicode and locale (starting in
7ee2ae1e 1232v5.16), you use
66cbab2c
KW
1233
1234 use locale ':not_characters';
1235
1236When this form of the pragma is used, only the non-character portions of
1237locales are used by Perl, for example C<LC_NUMERIC>. Perl assumes that
1238you have translated all the characters it is to operate on into Unicode
1239(actually the platform's native character set (ASCII or EBCDIC) plus
1240Unicode). For data in files, this can conveniently be done by also
1241specifying
1242
1243 use open ':locale';
1244
1245This pragma arranges for all inputs from files to be translated into
1246Unicode from the current locale as specified in the environment (see
1247L</ENVIRONMENT>), and all outputs to files to be translated back
1248into the locale. (See L<open>). On a per-filehandle basis, you can
1249instead use the L<PerlIO::locale> module, or the L<Encode::Locale>
1250module, both available from CPAN. The latter module also has methods to
1251ease the handling of C<ARGV> and environment variables, and can be used
1252on individual strings. Also, if you know that all your locales will be
1253UTF-8, as many are these days, you can use the L<B<-C>|perlrun/-C>
1254command line switch.
1255
1256This form of the pragma allows essentially seamless handling of locales
1257with Unicode. The collation order will be Unicode's. It is strongly
1258recommended that when you need to order and sort strings that you use
1259the standard module L<Unicode::Collate> which gives much better results
1260in many instances than you can get with the old-style locale handling.
1261
7ee2ae1e 1262For pre-v5.16 Perls, or if you use the locale pragma without the
66cbab2c
KW
1263C<:not_characters> parameter, Perl tries to work with both Unicode and
1264locales--but there are problems.
1265
1266Perl does not handle multi-byte locales in this case, such as have been
1267used for various
dc4bfc4b
KW
1268Asian languages, such as Big5 or Shift JIS. However, the increasingly
1269common multi-byte UTF-8 locales, if properly implemented, may work
1270reasonably well (depending on your C library implementation) in this
1271form of the locale pragma, simply because both
1272they and Perl store characters that take up multiple bytes the same way.
1273However, some, if not most, C library implementations may not process
1274the characters in the upper half of the Latin-1 range (128 - 255)
1275properly under LC_CTYPE. To see if a character is a particular type
1276under a locale, Perl uses the functions like C<isalnum()>. Your C
1277library may not work for UTF-8 locales with those functions, instead
1278only working under the newer wide library functions like C<iswalnum()>.
e199995e
KW
1279
1280Perl generally takes the tack to use locale rules on code points that can fit
66cbab2c
KW
1281in a single byte, and Unicode rules for those that can't (though this
1282isn't uniformly applied, see the note at the end of this section). This
1283prevents many problems in locales that aren't UTF-8. Suppose the locale
1284is ISO8859-7, Greek. The character at 0xD7 there is a capital Chi. But
1285in the ISO8859-1 locale, Latin1, it is a multiplication sign. The POSIX
1286regular expression character class C<[[:alpha:]]> will magically match
12870xD7 in the Greek locale but not in the Latin one.
e199995e
KW
1288
1289However, there are places where this breaks down. Certain constructs are
b4ffc3db
TC
1290for Unicode only, such as C<\p{Alpha}>. They assume that 0xD7 always has its
1291Unicode meaning (or the equivalent on EBCDIC platforms). Since Latin1 is a
1292subset of Unicode and 0xD7 is the multiplication sign in both Latin1 and
1293Unicode, C<\p{Alpha}> will never match it, regardless of locale. A similar
1294issue occurs with C<\N{...}>. It is therefore a bad idea to use C<\p{}> or
66cbab2c
KW
1295C<\N{}> under plain C<use locale>--I<unless> you can guarantee that the
1296locale will be a ISO8859-1. Use POSIX character classes instead.
1297
1298Another problem with this approach is that operations that cross the
1299single byte/multiple byte boundary are not well-defined, and so are
1300disallowed. (This boundary is between the codepoints at 255/256.).
1301For example, lower casing LATIN CAPITAL LETTER Y WITH DIAERESIS (U+0178)
1302should return LATIN SMALL LETTER Y WITH DIAERESIS (U+00FF). But in the
1303Greek locale, for example, there is no character at 0xFF, and Perl
1304has no way of knowing what the character at 0xFF is really supposed to
1305represent. Thus it disallows the operation. In this mode, the
1306lowercase of U+0178 is itself.
1307
1308The same problems ensue if you enable automatic UTF-8-ification of your
e199995e 1309standard file handles, default C<open()> layer, and C<@ARGV> on non-ISO8859-1,
b4ffc3db
TC
1310non-UTF-8 locales (by using either the B<-C> command line switch or the
1311C<PERL_UNICODE> environment variable; see L<perlrun>).
1312Things are read in as UTF-8, which would normally imply a Unicode
1313interpretation, but the presence of a locale causes them to be interpreted
1314in that locale instead. For example, a 0xD7 code point in the Unicode
1315input, which should mean the multiplication sign, won't be interpreted by
66cbab2c 1316Perl that way under the Greek locale. This is not a problem
b4ffc3db 1317I<provided> you make certain that all locales will always and only be either
66cbab2c 1318an ISO8859-1, or, if you don't have a deficient C library, a UTF-8 locale.
b4ffc3db
TC
1319
1320Vendor locales are notoriously buggy, and it is difficult for Perl to test
1321its locale-handling code because this interacts with code that Perl has no
1322control over; therefore the locale-handling code in Perl may be buggy as
66cbab2c
KW
1323well. (However, the Unicode-supplied locales should be better, and
1324there is a feed back mechanism to correct any problems. See
1325L</Freely available locale definitions>.)
1326
7ee2ae1e 1327If you have Perl v5.16, the problems mentioned above go away if you use
66cbab2c 1328the C<:not_characters> parameter to the locale pragma (except for vendor
7ee2ae1e 1329bugs in the non-character portions). If you don't have v5.16, and you
66cbab2c
KW
1330I<do> have locales that work, using them may be worthwhile for certain
1331specific purposes, as long as you keep in mind the gotchas already
1332mentioned. For example, if the collation for your locales works, it
1333runs faster under locales than under L<Unicode::Collate>; and you gain
1334access to such things as the local currency symbol and the names of the
7ee2ae1e 1335months and days of the week. (But to hammer home the point, in v5.16,
66cbab2c
KW
1336you get this access without the downsides of locales by using the
1337C<:not_characters> form of the pragma.)
1338
1339Note: The policy of using locale rules for code points that can fit in a
1340byte, and Unicode rules for those that can't is not uniformly applied.
7ee2ae1e 1341Pre-v5.12, it was somewhat haphazard; in v5.12 it was applied fairly
66cbab2c 1342consistently to regular expression matching except for bracketed
7ee2ae1e
KW
1343character classes; in v5.14 it was extended to all regex matches; and in
1344v5.16 to the casing operations such as C<"\L"> and C<uc()>. For
66cbab2c
KW
1345collation, in all releases, the system's C<strxfrm()> function is called,
1346and whatever it does is what you get.
b310b053 1347
5f05dabc
PP
1348=head1 BUGS
1349
1350=head2 Broken systems
1351
5a964f20 1352In certain systems, the operating system's locale support
2bdf8add 1353is broken and cannot be fixed or used by Perl. Such deficiencies can
b4ffc3db 1354and will result in mysterious hangs and/or Perl core dumps when
2bdf8add 1355C<use locale> is in effect. When confronted with such a system,
7f2de2d2 1356please report in excruciating detail to <F<perlbug@perl.org>>, and
b4ffc3db 1357also contact your vendor: bug fixes may exist for these problems
2bdf8add
JH
1358in your operating system. Sometimes such bug fixes are called an
1359operating system upgrade.
5f05dabc
PP
1360
1361=head1 SEE ALSO
1362
b310b053
JH
1363L<I18N::Langinfo>, L<perluniintro>, L<perlunicode>, L<open>,
1364L<POSIX/isalnum>, L<POSIX/isalpha>,
4bbcc6e8
JH
1365L<POSIX/isdigit>, L<POSIX/isgraph>, L<POSIX/islower>,
1366L<POSIX/isprint>, L<POSIX/ispunct>, L<POSIX/isspace>,
1367L<POSIX/isupper>, L<POSIX/isxdigit>, L<POSIX/localeconv>,
1368L<POSIX/setlocale>, L<POSIX/strcoll>, L<POSIX/strftime>,
1369L<POSIX/strtod>, L<POSIX/strxfrm>.
5f05dabc 1370
ccd65d51
KW
1371For special considerations when Perl is embedded in a C program,
1372see L<perlembed/Using embedded Perl with POSIX locales>.
1373
5f05dabc
PP
1374=head1 HISTORY
1375
b0c42ed9 1376Jarkko Hietaniemi's original F<perli18n.pod> heavily hacked by Dominic
5a964f20 1377Dunlop, assisted by the perl5-porters. Prose worked over a bit by
c052850d 1378Tom Christiansen, and updated by Perl 5 porters.