lib/Packages/I18N/Languages.pm

   1 # Taken from the webwml CVS tree (english/templates/languages.wml)
   2
   3 package Packages::I18N::Languages;
   4
   5 use strict;
   6 use warnings;
   7
   8 use Exporter;
   9
  10 our @ISA = qw( Exporter );
  11 our @EXPORT = qw( langcmp get_transliteration get_selfname );
  12
  13 # language directory name => ISO 639 two-letter code for the language name
  14 my %langs = (
  15              english    => "en",
  16              arabic     => "ar",
  17              bulgarian  => "bg",
  18              catalan    => "ca",
  19              czech      => "cs",
  20              danish     => "da",
  21              german     => "de",
  22              greek      => "el",
  23              esperanto  => "eo",
  24              spanish    => "es",
  25              finnish    => "fi",
  26              french     => "fr",
  27              croatian   => "hr",
  28              hungarian  => "hu",
  29              indonesian => "id",
  30              italian    => "it",
  31              japanese   => "ja",
  32              korean     => "ko",
  33              lithuanian => "lt",
  34              dutch      => "nl",
  35              norwegian  => "no",
  36              polish     => "pl",
  37              portuguese => "pt",
  38              romanian   => "ro",
  39              russian    => "ru",
  40              swedish    => "sv",
  41              slovene    => "sl",
  42              slovak     => "sk",
  43              turkish    => "tr",
  44              chinese    => "zh",
  45              );
  46
  47 # language directory name => native name of the language
  48 # non-ASCII letters must be escaped (using entities)!
  49 my %selflang = (
  50                 ar     => '&#1593;&#1585;&#1576;&#1610;&#1577;',
  51                 bg     => '&#1041;&#1098;&#1083;&#1075;&#1072;&#1088;&#1089;&#1082;&#1080;',
  52                 ca     => 'catal&agrave;',
  53                 zh     => '&#20013;&#25991;',
  54                 hr     => 'hrvatski',
  55                 cs     => '&#269;esky',
  56                 da     => 'dansk',
  57                 nl     => 'Nederlands',
  58                 en     => 'English',
  59                 eo     => 'Esperanto',
  60                 fi     => 'suomi',
  61                 fr     => 'fran&ccedil;ais',
  62                 de     => 'Deutsch',
  63                 el     => '&#917;&#955;&#955;&#951;&#957;&#953;&#954;&#940;',
  64                 hu     => 'magyar',
  65                 id     => 'Indonesia',
  66                 it     => 'Italiano',
  67                 ja     => '&#26085;&#26412;&#35486;',
  68                 ko     => '&#54620;&#44397;&#50612;',
  69                 lt     => 'Lietuvi&#371;',
  70                 "no"   => 'norsk&nbsp;(bokm&aring;l)',
  71                 pl     => 'polski',
  72                 pt_PT  => 'Portugu&ecirc;s (pt)',
  73                 pt_BR  => 'Portugu&ecirc;s (br)',
  74                 ro     => 'rom&acirc;n&#259;',
  75                 ru     => '&#1056;&#1091;&#1089;&#1089;&#1082;&#1080;&#1081;',
  76                 es     => 'espa&ntilde;ol',
  77                 sk     => 'slovak', #FIXME
  78                 sv_SE  => 'svenska',
  79                 sl     => 'sloven&#353;&#269;ina',
  80                 tr     => 'T&uuml;rk&ccedil;e',
  81                 uk     => 'ukrainian', #FIXME
  82                 );
  83
  84 # language directory name => Latin transliteration of the language name
  85 # This is used for language names which consist entirely of non-Latin
  86 # characters, to aid those that have browsers which cannot show different
  87 # character sets at once.
  88 my %translit = (
  89                 ar => "Arabiya",
  90                 bg => "B&#601;lgarski",
  91                 zh => "Zhongzu", # Not printed due to Chinese-specific code; kept for sort order
  92                 el => "Ellinika",
  93                 ja => "Nihongo",
  94                 ko => "Hangul", # Not sure. "Hanguk-Mal" (=Spoken Korean)?
  95                 ru => "Russkij",
  96                 );
  97
  98 # second transliteration table, used for languages starting with a latin
  99 # diacritic letter
 100 my %translit2 = (
 101                  cs    => "cesky",
 102 );
 103
 104 sub langcmp ($$) {
 105   my ($first, $second) = @_;
 106
 107   # Handle sorting of non-latin characters
 108   # If there is a transliteration for this language available, use it
 109   $first = $translit{$first} if defined $translit{$first};
 110   $second = $translit{$second} if defined $translit{$second};
 111
 112   # Then handle special cases (initial latin letters with diacritics)
 113   $first = $translit2{$first} if defined $translit2{$first};
 114   $second = $translit2{$second} if defined $translit2{$second};
 115
 116   # Put remaining entity-only names last in the list
 117   if (substr($first,0,1) eq '&')
 118   {
 119     $first =~ s/^&/ZZZ&/;
 120   }
 121   if (substr($second,0,1) eq '&')
 122   {
 123     $second =~ s/^&/ZZZ&/;
 124   }
 125   #    There seems to be a bug with localization in
 126   #    Perl 5.005 so we need those extra variables.
 127   my ($ufirst, $usecond) = (uc($first), uc($second));
 128   return $ufirst cmp $usecond;
 129 }
 130
 131 sub get_selfname {
 132     return $selflang{$_[0]} if exists $selflang{$_[0]};
 133     return undef;
 134 }
 135
 136 sub get_transliteration {
 137     return $translit{$_[0]} if exists $translit{$_[0]};
 138     return undef;
 139 }
 140
 141 1;