languages/classes/LanguageKu.php

   1 <?php
   2 /**
   3  * Kurdish 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  * Kurdish converter routines
  26  *
  27  * @ingroup Language
  28  */
  29 class KuConverter extends LanguageConverter {
  30         public $mArabicToLatin = [
  31                 'ب' => 'b', 'ج' => 'c', 'چ' => 'ç', 'د' => 'd', 'ف' => 'f', 'گ' => 'g', 'ھ' => 'h',
  32                 'ہ' => 'h', 'ه' => 'h', 'ح' => 'h', 'ژ' => 'j', 'ك' => 'k', 'ک' => 'k', 'ل' => 'l',
  33                 'م' => 'm', 'ن' => 'n', 'پ' => 'p', 'ق' => 'q', 'ر' => 'r', 'س' => 's', 'ش' => 'ş',
  34                 'ت' => 't', 'ڤ' => 'v', 'خ' => 'x', 'غ' => 'x', 'ز' => 'z',
  35
  36 // ک و => ku -- ist richtig
  37 //  و ك=> ku -- ist auch richtig
  38
  39                 /* Doppel- und Halbvokale */
  40                 'ڵ' => 'll', # ll
  41                 'ڕ' => 'rr', # rr
  42                 'ا' => 'a',
  43                 # 'ئێ' => 'ê', # initial e
  44                 'ە' => 'e',
  45                 'ه‌' => 'e', # with one non-joiner
  46                 'ه‌‌' => 'e', # with two non-joiner
  47                 'ة' => 'e',
  48                 'ێ' => 'ê',
  49                 'ي' => 'î',
  50                 'ی' => 'î', # U+06CC  db 8c  ARABIC LETTER FARSI YEH
  51                 'ى' => 'î', # U+0649  d9 89  ARABIC LETTER ALEF MAKSURA
  52                 'ۆ' => 'o',
  53                 'و' => 'w',
  54                 'ئ' => '', # initial hemze should not be shown
  55                 '،' => ',',
  56                 'ع' => '\'', # ayn
  57                 '؟' => '?',
  58
  59                 # digits
  60                 '٠' => '0', # &#x0660;
  61                 '١' => '1', # &#x0661;
  62                 '٢' => '2', # &#x0662;
  63                 '٣' => '3', # &#x0663;
  64                 '٤' => '4', # &#x0664;
  65                 '٥' => '5', # &#x0665;
  66                 '٦' => '6', # &#x0666;
  67                 '٧' => '7', # &#x0667;
  68                 '٨' => '8', # &#x0668;
  69                 '٩' => '9', # &#x0669;
  70         ];
  71
  72         public $mLatinToArabic = [
  73                 'b' => 'ب', 'c' => 'ج', 'ç' => 'چ', 'd' => 'د', 'f' => 'ف', 'g' => 'گ',
  74                 'h' => 'ه', 'j' => 'ژ', 'k' => 'ک', 'l' => 'ل',
  75                 'm' => 'م', 'n' => 'ن', 'p' => 'پ', 'q' => 'ق', 'r' => 'ر', 's' => 'س', 'ş' => 'ش',
  76                 't' => 'ت', 'v' => 'ڤ',
  77                 'x' => 'خ', 'y' => 'ی', 'z' => 'ز',
  78
  79                 'B' => 'ب', 'C' => 'ج', 'Ç' => 'چ', 'D' => 'د', 'F' => 'ف', 'G' => 'گ', 'H' => 'ھ',
  80                 'H' => 'ہ', 'H' => 'ه', 'H' => 'ح', 'J' => 'ژ', 'K' => 'ك', 'K' => 'ک', 'L' => 'ل',
  81                 'M' => 'م', 'N' => 'ن', 'P' => 'پ', 'Q' => 'ق', 'R' => 'ر', 'S' => 'س', 'Ş' => 'ش',
  82                 'T' => 'ت', 'V' => 'ڤ', 'W' => 'و', 'X' => 'خ',
  83                 'Y' => 'ی', 'Z' => 'ز',
  84
  85                 /* Doppelkonsonanten */
  86                 # 'll' => 'ڵ', # wenn es geht, doppel-l und l getrennt zu behandeln
  87                 # 'rr' => 'ڕ', # selbiges für doppel-r
  88
  89                 /* Einzelne Großbuchstaben */
  90                 // ' C' => 'ج',
  91
  92                 /* Vowels */
  93                 'a' => 'ا',
  94                 'e' => 'ە',
  95                 'ê' => 'ێ',
  96                 'i' => '',
  97                 'î' => 'ی',
  98                 'o' => 'ۆ',
  99                 'u' => 'و',
 100                 'û' => 'وو',
 101                 'w' => 'و',
 102                 ',' => '،',
 103                 '?' => '؟',
 104
 105                 # Try to replace the leading vowel
 106                 ' a' => 'ئا ',
 107                 ' e' => 'ئە ',
 108                 ' ê' => 'ئێ ',
 109                 ' î' => 'ئی ',
 110                 ' o' => 'ئۆ ',
 111                 ' u' => 'ئو ',
 112                 ' û' => 'ئوو ',
 113                 'A' => 'ئا',
 114                 'E' => 'ئە',
 115                 'Ê' => 'ئێ',
 116                 'Î' => 'ئی',
 117                 'O' => 'ئۆ',
 118                 'U' => 'ئو',
 119                 'Û' => 'ئوو',
 120                 ' A' => 'ئا ',
 121                 ' E' => 'ئە ',
 122                 ' Ê' => 'ئێ ',
 123                 ' Î' => 'ئی ',
 124                 ' O' => 'ئۆ ',
 125                 ' U' => 'ئو ',
 126                 ' Û' => 'ئوو ',
 127                 # eyn erstmal deaktivieren, einfache Anführungsstriche sind einfach zu
 128                 # häufig, um sie als eyn zu interpretieren.
 129                 # '\'' => 'ع',
 130
 131 /*              # deactivated for now, breaks links i.e. in header of Special:Recentchanges :-(
 132                 # digits
 133                 '0' => '٠', # &#x0660;
 134                 '1' => '١', # &#x0661;
 135                 '2' => '٢', # &#x0662;
 136                 '3' => '٣', # &#x0663;
 137                 '4' => '٤', # &#x0664;
 138                 '5' => '٥', # &#x0665;
 139                 '6' => '٦', # &#x0666;
 140                 '7' => '٧', # &#x0667;
 141                 '8' => '٨', # &#x0668;
 142                 '9' => '٩', # &#x0669;
 143 */
 144                 ];
 145
 146         function loadDefaultTables() {
 147                 $this->mTables = [
 148                         'ku-latn' => new ReplacementArray( $this->mArabicToLatin ),
 149                         'ku-arab' => new ReplacementArray( $this->mLatinToArabic ),
 150                         'ku' => new ReplacementArray()
 151                 ];
 152         }
 153
 154         /**
 155          * A function wrapper:
 156          *   - if there is no selected variant, leave the link
 157          *     names as they were
 158          *   - do not try to find variants for usernames
 159          *
 160          * @param string &$link
 161          * @param Title &$nt
 162          * @param bool $ignoreOtherCond
 163          */
 164         function findVariantLink( &$link, &$nt, $ignoreOtherCond = false ) {
 165                 // check for user namespace
 166                 if ( is_object( $nt ) ) {
 167                         $ns = $nt->getNamespace();
 168                         if ( $ns == NS_USER || $ns == NS_USER_TALK ) {
 169                                 return;
 170                         }
 171                 }
 172
 173                 $oldlink = $link;
 174                 parent::findVariantLink( $link, $nt, $ignoreOtherCond );
 175                 if ( $this->getPreferredVariant() == $this->mMainLanguageCode ) {
 176                         $link = $oldlink;
 177                 }
 178         }
 179
 180         /**
 181          *  It translates text into variant, specials:
 182          *    - ommiting roman numbers
 183          *
 184          * @param string $text
 185          * @param bool $toVariant
 186          *
 187          * @throws MWException
 188          * @return string
 189          */
 190         function translate( $text, $toVariant ) {
 191                 $this->loadTables();
 192                 /* From Kazakh interface, maybe we need it later
 193                 $breaks = '[^\w\x80-\xff]';
 194                 // regexp for roman numbers
 195                 $roman = 'M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})';
 196                 $roman = '';
 197
 198                 $reg = '/^'.$roman.'$|^'.$roman.$breaks.'|'.$breaks.$roman.'$|'.$breaks.$roman.$breaks.'/';
 199
 200                 $matches = preg_split($reg, $text, -1, PREG_SPLIT_OFFSET_CAPTURE);
 201
 202                 $m = array_shift($matches);
 203                 if( !isset( $this->mTables[$toVariant] ) ) {
 204                         throw new MWException( "Broken variant table: " . implode( ',', array_keys( $this->mTables ) ) );
 205                 }
 206                 $ret = $this->mTables[$toVariant]->replace( $m[0] );
 207                 $mstart = $m[1]+strlen($m[0]);
 208                 foreach($matches as $m) {
 209                         $ret .= substr($text, $mstart, $m[1]-$mstart);
 210                         $ret .= parent::translate($m[0], $toVariant);
 211                         $mstart = $m[1] + strlen($m[0]);
 212                 }
 213
 214                 return $ret;
 215                 */
 216
 217                 if ( !isset( $this->mTables[$toVariant] ) ) {
 218                         throw new MWException( "Broken variant table: " . implode( ',', array_keys( $this->mTables ) ) );
 219                 }
 220
 221                 return parent::translate( $text, $toVariant );
 222         }
 223 }
 224
 225 /**
 226  * Kurdish (Kurdî / كوردی)
 227  *
 228  * @ingroup Language
 229  */
 230 class LanguageKu extends Language {
 231
 232         function __construct() {
 233                 parent::__construct();
 234
 235                 $variants = [ 'ku', 'ku-arab', 'ku-latn' ];
 236                 $variantfallbacks = [
 237                         'ku' => 'ku-latn',
 238                         'ku-arab' => 'ku-latn',
 239                         'ku-latn' => 'ku-arab',
 240                 ];
 241
 242                 $this->mConverter = new KuConverter( $this, 'ku', $variants, $variantfallbacks );
 243         }
 244 }