languages/classes/LanguageKu.php

   1 <?php
   2 /** Kurdish
   3   * converter routines
   4   *
   5   * @addtogroup Language
   6   */
   7
   8 require_once( dirname(__FILE__).'/../LanguageConverter.php' );
   9 require_once( dirname(__FILE__).'/LanguageKu_ku.php' );
  10
  11 class KuConverter extends LanguageConverter {
  12         var $mArabicToLatin = array(
  13                 'ب' => 'b', 'ج' => 'c', 'چ' => 'ç', 'د' => 'd', 'ف' => 'f', 'گ' => 'g', 'ھ' => 'h',
  14                 'ہ' => 'h', 'ه' => 'h', 'ح' => 'h', 'ژ' => 'j', 'ك' => 'k', 'ک' => 'k', 'ل' => 'l',
  15                 'م' => 'm', 'ن' => 'n', 'پ' => 'p', 'ق' => 'q', 'ر' => 'r', 'س' => 's', 'ش' => 'ş',
  16                 'ت' => 't', 'ڤ' => 'v', 'خ' => 'x', 'غ' => 'x', 'ز' => 'z',
  17
  18                 /* Doppel- und Halbvokale */
  19                 'ڵ' => 'll', #ll
  20                 'ڕ'  => 'rr', #rr
  21                 'ا'  => 'a',
  22                 # 'ئێ' => 'ê', # Anfangs-e
  23                 'ە'  => 'e',
  24                 'ه‌'  => 'e', # mit einem non-joiner
  25                 'ه‌‌'  => 'e', # mit zwei non-joinern
  26                 'ة'  => 'e',
  27                 'ێ' => 'ê',
  28                 'ي'  => 'î',
  29                 'ی'  => 'î',
  30                 'ۆ'  => 'o',
  31                 'و'  => 'w',
  32                 'ئ'  => '', # initiales hemze soll verschluckt werden
  33                 '،'  => ',',
  34                 'ع'  => '\'', # ayn
  35                 '؟'  => '?',
  36         );
  37
  38         var $mLatinToArabic = array(
  39                 'b' => 'ب', 'c' => 'ج', 'ç' => 'چ', 'd' => 'د', 'f' => 'ف', 'g' => 'گ',
  40                 'h' => 'ه', 'j' => 'ژ', 'k' => 'ک', 'l' => 'ل',
  41                 'm' => 'م', 'n' => 'ن', 'p' => 'پ', 'q' => 'ق', 'r' => 'ر', 's' => 'س', 'ş' => 'ش',
  42                 't' => 'ت', 'v' => 'ڤ',
  43                 'x' => 'خ', 'y' => 'ی', 'z' => 'ز',
  44
  45
  46                 'B' => 'ب', 'C' => 'ج', 'Ç' => 'چ', 'D' => 'د', 'F' => 'ف', 'G' => 'گ', 'H' => 'ھ',
  47                 'H' => 'ہ', 'H' => 'ه', 'H' => 'ح', 'J' => 'ژ', 'K' => 'ك', 'K' => 'ک', 'L' => 'ل',
  48                 'M' => 'م', 'N' => 'ن', 'P' => 'پ', 'Q' => 'ق', 'R' => 'ر', 'S' => 'س', 'Ş' => 'ش',
  49                 'T' => 'ت', 'V' => 'ڤ', 'W' => 'و', 'X' => 'خ',
  50                 'Y' => 'ی', 'Z' => 'ز',
  51
  52                 /* Doppelkonsonanten */
  53                 # 'll' => 'ڵ', # wenn es geht, doppel-l und l getrennt zu behandeln
  54                 # 'rr' => 'ڕ', # selbiges für doppel-r
  55
  56                 /* Einzelne Großbuchstaben */
  57                 //' C' => 'ج',
  58
  59                 /* Vowels */
  60                 'a' => 'ا',
  61                 'e' => 'ە',
  62                 'ê' => 'ێ',
  63                 'i' => '',
  64                 'î' => 'ی',
  65                 'o' => 'ۆ',
  66                 'u' => 'و',
  67                 'û' => 'وو',
  68                 'w' => 'و',
  69                 ',' => '،',
  70                 '?' => '؟',
  71
  72                 # Try to replace the leading vowel
  73                 ' a' => 'ئا ',
  74                 ' e' => 'ئە ',
  75                 ' ê' => 'ئێ ',
  76                 ' î' => 'ئی ',
  77                 ' o' => 'ئۆ ',
  78                 ' u' => 'ئو ',
  79                 ' û' => 'ئوو ',
  80                 'A'  => 'ئا',
  81                 'E'  => 'ئە',
  82                 'Ê'  => 'ئێ',
  83                 'Î'  => 'ئی',
  84                 'O'  => 'ئۆ',
  85                 'U'  => 'ئو',
  86                 'Û'  => 'ئوو',
  87                 ' A' => 'ئا ',
  88                 ' E' => 'ئە ',
  89                 ' Ê' => 'ئێ ',
  90                 ' Î' => 'ئی ',
  91                 ' O' => 'ئۆ ',
  92                 ' U' => 'ئو ',
  93                 ' Û' => 'ئوو ',
  94                 # eyn erstmal deaktivieren, einfache Anführungsstriche sind einfach zu häufig, um sie als eyn zu interpretieren
  95                 # '\'' => 'ع',
  96
  97         );
  98
  99         function loadDefaultTables() {
 100                 $this->mTables = array(
 101                         'ku-latn' => new ReplacementArray( $this->mArabicToLatin ),
 102                         'ku-arab' => new ReplacementArray( $this->mLatinToArabic ),
 103                         'ku'      => new ReplacementArray()
 104                 );
 105         }
 106
 107
 108         // Do not convert content on talk pages
 109         function parserConvert( $text, &$parser ){
 110                 if(is_object($parser->getTitle() ) && $parser->getTitle()->isTalkPage())
 111                         $this->mDoContentConvert=false;
 112                 else
 113                         $this->mDoContentConvert=true;
 114
 115                 return parent::parserConvert($text, $parser );
 116         }
 117
 118         /*
 119          * A function wrapper:
 120          *   - if there is no selected variant, leave the link
 121          *     names as they were
 122          *   - do not try to find variants for usernames
 123          */
 124         function findVariantLink( &$link, &$nt ) {
 125                 // check for user namespace
 126                 if(is_object($nt)){
 127                         $ns = $nt->getNamespace();
 128                         if($ns==NS_USER || $ns==NS_USER_TALK)
 129                                 return;
 130                 }
 131
 132                 $oldlink=$link;
 133                 parent::findVariantLink($link,$nt);
 134                 if($this->getPreferredVariant()==$this->mMainLanguageCode)
 135                         $link=$oldlink;
 136         }
 137
 138         /*
 139          * We want our external link captions to be converted in variants,
 140          * so we return the original text instead -{$text}-, except for URLs
 141          */
 142         function markNoConversion($text, $noParse=false) {
 143                 if($noParse || preg_match("/^https?:\/\/|ftp:\/\/|irc:\/\//",$text))
 144                     return parent::markNoConversion($text);
 145                 return $text;
 146         }
 147
 148         /*
 149          * An ugly function wrapper for parsing Image titles
 150          * (to prevent image name conversion)
 151          */
 152         function autoConvert($text, $toVariant=false) {
 153                 global $wgTitle;
 154                 if(is_object($wgTitle) && $wgTitle->getNameSpace()==NS_IMAGE){
 155                         $imagename = $wgTitle->getNsText();
 156                         if(preg_match("/^$imagename:/",$text)) return $text;
 157                 }
 158                 return parent::autoConvert($text,$toVariant);
 159         }
 160
 161         /**
 162          *  It translates text into variant, specials:
 163          *    - ommiting roman numbers
 164          */
 165         function translate($text, $toVariant){
 166                 $breaks = '[^\w\x80-\xff]';
 167
 168                 /* From Kazakh interface, maybe we need it later
 169                  *
 170                 // regexp for roman numbers
 171                 $roman = 'M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})';
 172                 $roman = '';
 173
 174                 $reg = '/^'.$roman.'$|^'.$roman.$breaks.'|'.$breaks.$roman.'$|'.$breaks.$roman.$breaks.'/';
 175
 176                 $matches = preg_split($reg, $text, -1, PREG_SPLIT_OFFSET_CAPTURE);
 177
 178                 $m = array_shift($matches);
 179                 if( !isset( $this->mTables[$toVariant] ) ) {
 180                         throw new MWException( "Broken variant table: " . implode( ',', array_keys( $this->mTables ) ) );
 181                 }
 182                 $ret = $this->mTables[$toVariant]->replace( $m[0] );
 183                 $mstart = $m[1]+strlen($m[0]);
 184                 foreach($matches as $m) {
 185                         $ret .= substr($text, $mstart, $m[1]-$mstart);
 186                         $ret .= parent::translate($m[0], $toVariant);
 187                         $mstart = $m[1] + strlen($m[0]);
 188                 }
 189
 190                 return $ret;
 191                 */
 192
 193                 if( !isset( $this->mTables[$toVariant] ) ) {
 194                         throw new MWException( "Broken variant table: " . implode( ',', array_keys( $this->mTables ) ) );
 195                 }
 196
 197                 return parent::translate( $text, $toVariant );
 198         }
 199 }
 200
 201 class LanguageKu extends LanguageKu_ku {
 202
 203         function __construct() {
 204                 global $wgHooks;
 205                 parent::__construct();
 206
 207                 $variants = array( 'ku', 'ku-arab', 'ku-latn' );
 208                 $variantfallbacks = array(
 209                         'ku'      => 'ku-latn',
 210                         'ku-arab' => 'ku-latn',
 211                         'ku-latn' => 'ku-arab',
 212                 );
 213
 214                 $this->mConverter = new KuConverter( $this, 'ku', $variants, $variantfallbacks );
 215                 $wgHooks['ArticleSaveComplete'][] = $this->mConverter;
 216         }
 217
 218 /*   From Kazakh interface, not needed for the moment
 219
 220         function convertGrammar( $word, $case ) {
 221                 $fname="LanguageKu::convertGrammar";
 222                 wfProfileIn( $fname );
 223
 224                 //always convert to ku-latn before convertGrammar
 225                 $w1 = $word;
 226                 $word = $this->mConverter->autoConvert( $word, 'ku-latn' );
 227                 $w2 = $word;
 228                 $word = parent::convertGrammar( $word, $case );
 229                 //restore encoding
 230                 if( $w1 != $w2 ) {
 231                         $word = $this->mConverter->translate( $word, 'ku-latn' );
 232                 }
 233                 wfProfileOut( $fname );
 234                 return $word;
 235         }
 236 */
 237 }
 238
 239