languages/classes/LanguageRu.php

   1 <?php
   2 /**
   3  * Russian (русский язык) specific code.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Language
  22  */
  23
  24 /**
  25  * Russian (русский язык)
  26  *
  27  * You can contact:
  28  * Alexander Sigachov (alexander.sigachov at Googgle Mail)
  29  * Amir E. Aharoni (amir.aharoni@mail.huji.ac.il)
  30  *
  31  * @ingroup Language
  32  */
  33 class LanguageRu extends Language {
  34
  35         /**
  36          * Convert from the nominative form of a noun to some other case
  37          * Invoked with {{grammar:case|word}}
  38          *
  39          * @param string $word
  40          * @param string $case
  41          * @return string
  42          */
  43         function convertGrammar( $word, $case ) {
  44                 global $wgGrammarForms;
  45                 if ( isset( $wgGrammarForms['ru'][$case][$word] ) ) {
  46                         return $wgGrammarForms['ru'][$case][$word];
  47                 }
  48
  49                 # These rules don't cover the whole language, and are intended only for
  50                 # names of languages and Wikimedia sites.
  51
  52                 # substr doesn't support Unicode and mb_substr has issues,
  53                 # so break it to characters using preg_match_all and then use array_slice and join
  54                 $chars = array();
  55                 preg_match_all( '/./us', $word, $chars );
  56                 if ( !preg_match( "/[a-zA-Z_]/us", $word ) ) {
  57                         switch ( $case ) {
  58                                 case 'genitive': # родительный падеж
  59                                         if ( join( '', array_slice( $chars[0], -1 ) ) === 'ь' ) {
  60                                                 $word = join( '', array_slice( $chars[0], 0, -1 ) ) . 'я';
  61                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ия' ) {
  62                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'ии';
  63                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ка' ) {
  64                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'ки';
  65                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ти' ) {
  66                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'тей';
  67                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ды' ) {
  68                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'дов';
  69                                         } elseif ( join( '', array_slice( $chars[0], -1 ) ) === 'д' ) {
  70                                                 $word = join( '', array_slice( $chars[0], 0, -1 ) ) . 'да';
  71                                         } elseif ( join( '', array_slice( $chars[0], -3 ) ) === 'ник' ) {
  72                                                 $word = join( '', array_slice( $chars[0], 0, -3 ) ) . 'ника';
  73                                         } elseif ( join( '', array_slice( $chars[0], -3 ) ) === 'ные' ) {
  74                                                 $word = join( '', array_slice( $chars[0], 0, -3 ) ) . 'ных';
  75                                         }
  76                                         break;
  77                                 case 'dative': # дательный падеж
  78                                         # stub
  79                                         break;
  80                                 case 'accusative': # винительный падеж
  81                                         # stub
  82                                         break;
  83                                 case 'instrumental': # творительный падеж
  84                                         # stub
  85                                         break;
  86                                 case 'prepositional': # предложный падеж
  87                                         if ( join( '', array_slice( $chars[0], -1 ) ) === 'ь' ) {
  88                                                 $word = join( '', array_slice( $chars[0], 0, -1 ) ) . 'е';
  89                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ия' ) {
  90                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'ии';
  91                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ка' ) {
  92                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'ке';
  93                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ти' ) {
  94                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'тях';
  95                                         } elseif ( join( '', array_slice( $chars[0], -2 ) ) === 'ды' ) {
  96                                                 $word = join( '', array_slice( $chars[0], 0, -2 ) ) . 'дах';
  97                                         } elseif ( join( '', array_slice( $chars[0], -1 ) ) === 'д' ) {
  98                                                 $word = join( '', array_slice( $chars[0], 0, -1 ) ) . 'де';
  99                                         } elseif ( join( '', array_slice( $chars[0], -3 ) ) === 'ник' ) {
 100                                                 $word = join( '', array_slice( $chars[0], 0, -3 ) ) . 'нике';
 101                                         } elseif ( join( '', array_slice( $chars[0], -3 ) ) === 'ные' ) {
 102                                                 $word = join( '', array_slice( $chars[0], 0, -3 ) ) . 'ных';
 103                                         }
 104                                         break;
 105                                 case 'languagegen': # язык в родительном падеже ("(с) русского")
 106                                         $suffix = join( '', array_slice( $chars[0], -3 ) );
 107                                         if ( $suffix === 'кий' ) {
 108                                                 $word = join(
 109                                                         '',
 110                                                         array_slice( $chars[0], 0, count( $chars[0] ) - 2 )
 111                                                 ) . 'ого';
 112
 113                                                 break;
 114                                         }
 115
 116                                         if ( in_array( $word, array( 'иврит', 'идиш' ) ) ) {
 117                                                 $word = $word . 'а';
 118
 119                                                 break;
 120                                         }
 121
 122                                         break;
 123                                 case 'languageprep': # язык в предложном падеже ("(на) русском")
 124                                         $suffix = join( '', array_slice( $chars[0], -3 ) );
 125                                         if ( $suffix === 'кий' ) {
 126                                                 $word = join(
 127                                                         '',
 128                                                         array_slice( $chars[0], 0, count( $chars[0] ) - 2 )
 129                                                 ) . 'ом';
 130
 131                                                 break;
 132                                         }
 133
 134                                         if ( in_array( $word, array( 'иврит', 'идиш' ) ) ) {
 135                                                 $word = $word . 'е';
 136
 137                                                 break;
 138                                         }
 139
 140                                         break;
 141                                 case 'languageadverb': # наречие с названием языка ("по-русски")
 142                                         $suffix = join( '', array_slice( $chars[0], -3 ) );
 143                                         if ( $suffix === 'кий' ) {
 144                                                 $word = 'по-' . join(
 145                                                         '',
 146                                                         array_slice( $chars[0], 0, count( $chars[0] ) - 1 )
 147                                                 );
 148
 149                                                 break;
 150                                         }
 151
 152                                         if ( in_array( $word, array( 'иврит', 'идиш' ) ) ) {
 153                                                 $word = 'на ' . $word . 'е';
 154
 155                                                 break;
 156                                         }
 157
 158                                         // Known particular cases of undeclinable names
 159                                         // Известные несклоняемые
 160                                         if ( in_array( $word, array( 'идо', 'урду', 'хинди', 'эсперанто' ) ) ) {
 161                                                 $word = "на $word";
 162
 163                                                 break;
 164                                         }
 165
 166                                         // Undeclinable
 167                                         // Остальные несклоняемые
 168                                         $word = "на языке $word";
 169
 170                                         break;
 171                         }
 172                 }
 173
 174                 return $word;
 175         }
 176
 177         /**
 178          * Four-digit number should be without group commas (spaces)
 179          * See manual of style at http://ru.wikipedia.org/wiki/Википедия:Оформление_статей
 180          * So "1 234 567", "12 345" but "1234"
 181          *
 182          * @param string $_
 183          *
 184          * @return string
 185          */
 186         function commafy( $_ ) {
 187                 if ( preg_match( '/^-?\d{1,4}(\.\d*)?$/', $_ ) ) {
 188                         return $_;
 189                 } else {
 190                         return strrev( (string)preg_replace( '/(\d{3})(?=\d)(?!\d*\.)/', '$1,', strrev( $_ ) ) );
 191                 }
 192         }
 193 }