Commit | Line | Data |
---|---|---|
a0d0e21e LW |
1 | package I18N::Collate; |
2 | ||
f06db76b AD |
3 | =head1 NAME |
4 | ||
69b19ea2 | 5 | I18N::Collate - compare 8-bit scalar data according to the current locale |
f06db76b | 6 | |
4413da2b JH |
7 | *** |
8 | ||
9 | WARNING: starting from the Perl version 5.003_06 | |
10 | the I18N::Collate interface for comparing 8-bit scalar data | |
11 | according to the current locale | |
12 | ||
13 | HAS BEEN DEPRECATED | |
14 | ||
15 | That is, please do not use it anymore for any new applications | |
16 | and please migrate the old applications away from it because its | |
17 | functionality was integrated into the Perl core language in the | |
18 | release 5.003_06. | |
19 | ||
20 | See the perllocale manual page for further information. | |
21 | ||
22 | *** | |
23 | ||
f06db76b AD |
24 | =head1 SYNOPSIS |
25 | ||
69b19ea2 | 26 | use I18N::Collate; |
f06db76b | 27 | setlocale(LC_COLLATE, 'locale-of-your-choice'); |
69b19ea2 | 28 | $s1 = new I18N::Collate "scalar_data_1"; |
29 | $s2 = new I18N::Collate "scalar_data_2"; | |
f06db76b AD |
30 | |
31 | =head1 DESCRIPTION | |
32 | ||
33 | This module provides you with objects that will collate | |
69b19ea2 | 34 | according to your national character set, provided that the |
35 | POSIX setlocale() function is supported on your system. | |
f06db76b AD |
36 | |
37 | You can compare $s1 and $s2 above with | |
38 | ||
39 | $s1 le $s2 | |
40 | ||
41 | to extract the data itself, you'll need a dereference: $$s1 | |
42 | ||
6158a1ac CS |
43 | This module uses POSIX::setlocale(). The basic collation conversion is |
44 | done by strxfrm() which terminates at NUL characters being a decent C | |
45 | routine. collate_xfrm() handles embedded NUL characters gracefully. | |
c2960299 | 46 | |
6158a1ac CS |
47 | The available locales depend on your operating system; try whether |
48 | C<locale -a> shows them or man pages for "locale" or "nlsinfo" or the | |
49 | direct approach C<ls /usr/lib/nls/loc> or C<ls /usr/lib/nls> or | |
50 | C<ls /usr/lib/locale>. Not all the locales that your vendor supports | |
51 | are necessarily installed: please consult your operating system's | |
52 | documentation and possibly your local system administration. The | |
53 | locale names are probably something like C<xx_XX.(ISO)?8859-N> or | |
54 | C<xx_XX.(ISO)?8859N>, for example C<fr_CH.ISO8859-1> is the Swiss (CH) | |
55 | variant of French (fr), ISO Latin (8859) 1 (-1) which is the Western | |
56 | European character set. | |
f06db76b AD |
57 | |
58 | =cut | |
59 | ||
69b19ea2 | 60 | # I18N::Collate.pm |
a0d0e21e | 61 | # |
5aabfad6 | 62 | # Author: Jarkko Hietaniemi <F<jhi@iki.fi>> |
a0d0e21e LW |
63 | # Helsinki University of Technology, Finland |
64 | # | |
5aabfad6 | 65 | # Acks: Guy Decoux <F<decoux@moulon.inra.fr>> understood |
a0d0e21e LW |
66 | # overloading magic much deeper than I and told |
67 | # how to cut the size of this code by more than half. | |
68 | # (my first version did overload all of lt gt eq le ge cmp) | |
69 | # | |
70 | # Purpose: compare 8-bit scalar data according to the current locale | |
71 | # | |
72 | # Requirements: Perl5 POSIX::setlocale() and POSIX::strxfrm() | |
73 | # | |
74 | # Exports: setlocale 1) | |
75 | # collate_xfrm 2) | |
76 | # | |
77 | # Overloads: cmp # 3) | |
78 | # | |
69b19ea2 | 79 | # Usage: use I18N::Collate; |
c2960299 | 80 | # setlocale(LC_COLLATE, 'locale-of-your-choice'); # 4) |
69b19ea2 | 81 | # $s1 = new I18N::Collate "scalar_data_1"; |
82 | # $s2 = new I18N::Collate "scalar_data_2"; | |
a0d0e21e LW |
83 | # |
84 | # now you can compare $s1 and $s2: $s1 le $s2 | |
85 | # to extract the data itself, you need to deref: $$s1 | |
86 | # | |
87 | # Notes: | |
88 | # 1) this uses POSIX::setlocale | |
89 | # 2) the basic collation conversion is done by strxfrm() which | |
90 | # terminates at NUL characters being a decent C routine. | |
91 | # collate_xfrm handles embedded NUL characters gracefully. | |
92 | # 3) due to cmp and overload magic, lt le eq ge gt work also | |
93 | # 4) the available locales depend on your operating system; | |
c2960299 AD |
94 | # try whether "locale -a" shows them or man pages for |
95 | # "locale" or "nlsinfo" work or the more direct | |
a0d0e21e | 96 | # approach "ls /usr/lib/nls/loc" or "ls /usr/lib/nls". |
c2960299 AD |
97 | # Not all the locales that your vendor supports |
98 | # are necessarily installed: please consult your | |
99 | # operating system's documentation. | |
a0d0e21e | 100 | # The locale names are probably something like |
c2960299 AD |
101 | # 'xx_XX.(ISO)?8859-N' or 'xx_XX.(ISO)?8859N', |
102 | # for example 'fr_CH.ISO8859-1' is the Swiss (CH) | |
103 | # variant of French (fr), ISO Latin (8859) 1 (-1) | |
104 | # which is the Western European character set. | |
a0d0e21e | 105 | # |
6b48aaa4 | 106 | # Updated: 19961005 |
a0d0e21e LW |
107 | # |
108 | # --- | |
109 | ||
110 | use POSIX qw(strxfrm LC_COLLATE); | |
111 | ||
112 | require Exporter; | |
113 | ||
114 | @ISA = qw(Exporter); | |
115 | @EXPORT = qw(collate_xfrm setlocale LC_COLLATE); | |
116 | @EXPORT_OK = qw(); | |
117 | ||
a5f75d66 | 118 | use overload qw( |
a0d0e21e LW |
119 | fallback 1 |
120 | cmp collate_cmp | |
121 | ); | |
122 | ||
6b48aaa4 JH |
123 | sub new { |
124 | my $new = $_[1]; | |
125 | ||
126 | if ($^W && $] >= 5.003_06) { | |
127 | unless ($please_use_I18N_Collate_even_if_deprecated) { | |
128 | warn <<___EOD___; | |
129 | *** | |
130 | ||
4413da2b JH |
131 | WARNING: starting from the Perl version 5.003_06 |
132 | the I18N::Collate interface for comparing 8-bit scalar data | |
133 | according to the current locale | |
6b48aaa4 JH |
134 | |
135 | HAS BEEN DEPRECATED | |
136 | ||
4413da2b JH |
137 | That is, please do not use it anymore for any new applications |
138 | and please migrate the old applications away from it because its | |
139 | functionality was integrated into the Perl core language in the | |
140 | release 5.003_06. | |
6b48aaa4 | 141 | |
4413da2b | 142 | See the perllocale manual page for further information. |
6b48aaa4 JH |
143 | |
144 | *** | |
145 | ___EOD___ | |
146 | $please_use_I18N_Collate_even_if_deprecated++; | |
147 | } | |
148 | } | |
149 | ||
150 | bless \$new; | |
151 | } | |
a0d0e21e LW |
152 | |
153 | sub setlocale { | |
154 | my ($category, $locale) = @_[0,1]; | |
155 | ||
156 | POSIX::setlocale($category, $locale) if (defined $category); | |
157 | # the current $LOCALE | |
158 | $LOCALE = $locale || $ENV{'LC_COLLATE'} || $ENV{'LC_ALL'} || ''; | |
159 | } | |
160 | ||
161 | sub C { | |
162 | my $s = ${$_[0]}; | |
163 | ||
164 | $C->{$LOCALE}->{$s} = collate_xfrm($s) | |
165 | unless (defined $C->{$LOCALE}->{$s}); # cache when met | |
166 | ||
167 | $C->{$LOCALE}->{$s}; | |
168 | } | |
169 | ||
170 | sub collate_xfrm { | |
171 | my $s = $_[0]; | |
172 | my $x = ''; | |
173 | ||
174 | for (split(/(\000+)/, $s)) { | |
175 | $x .= (/^\000/) ? $_ : strxfrm("$_\000"); | |
176 | } | |
177 | ||
178 | $x; | |
179 | } | |
180 | ||
181 | sub collate_cmp { | |
182 | &C($_[0]) cmp &C($_[1]); | |
183 | } | |
184 | ||
185 | # init $LOCALE | |
186 | ||
187 | &I18N::Collate::setlocale(); | |
188 | ||
189 | 1; # keep require happy |