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 // ک و => ku -- ist richtig
  19 //  و ك=> ku -- ist auch richtig
  20
  21                 /* Doppel- und Halbvokale */
  22                 'ڵ' => 'll', #ll
  23                 'ڕ'  => 'rr', #rr
  24                 'ا'  => 'a',
  25                 # 'ئێ' => 'ê', # initial e
  26                 'ە'  => 'e',
  27                 'ه‌'  => 'e', # with one non-joiner
  28                 'ه‌‌'  => 'e', # with two non-joiner
  29                 'ة'  => 'e',
  30                 'ێ' => 'ê',
  31                 'ي'  => 'î',
  32                 'ی'  => 'î', # U+06CC  db 8c  ARABIC LETTER FARSI YEH
  33                 'ى'  => 'î', # U+0649  d9 89  ARABIC LETTER ALEF MAKSURA
  34                 'ۆ'  => 'o',
  35                 'و'  => 'w',
  36                 'ئ'  => '', # initial hemze should not be shown
  37                 '،'  => ',',
  38                 'ع'  => '\'', # ayn
  39                 '؟'  => '?',
  40
  41                 # digits
  42                 '٠' => '0', # &#x0660;
  43                 '١' => '1', # &#x0661;
  44                 '٢' => '2', # &#x0662;
  45                 '٣' => '3', # &#x0663;
  46                 '٤' => '4', # &#x0664;
  47                 '٥' => '5', # &#x0665;
  48                 '٦' => '6', # &#x0666;
  49                 '٧' => '7', # &#x0667;
  50                 '٨' => '8', # &#x0668;
  51                 '٩' => '9', # &#x0669;
  52         );
  53
  54         var $mLatinToArabic = array(
  55                 'b' => 'ب', 'c' => 'ج', 'ç' => 'چ', 'd' => 'د', 'f' => 'ف', 'g' => 'گ',
  56                 'h' => 'ه', 'j' => 'ژ', 'k' => 'ک', 'l' => 'ل',
  57                 'm' => 'م', 'n' => 'ن', 'p' => 'پ', 'q' => 'ق', 'r' => 'ر', 's' => 'س', 'ş' => 'ش',
  58                 't' => 'ت', 'v' => 'ڤ',
  59                 'x' => 'خ', 'y' => 'ی', 'z' => 'ز',
  60
  61
  62                 'B' => 'ب', 'C' => 'ج', 'Ç' => 'چ', 'D' => 'د', 'F' => 'ف', 'G' => 'گ', 'H' => 'ھ',
  63                 'H' => 'ہ', 'H' => 'ه', 'H' => 'ح', 'J' => 'ژ', 'K' => 'ك', 'K' => 'ک', 'L' => 'ل',
  64                 'M' => 'م', 'N' => 'ن', 'P' => 'پ', 'Q' => 'ق', 'R' => 'ر', 'S' => 'س', 'Ş' => 'ش',
  65                 'T' => 'ت', 'V' => 'ڤ', 'W' => 'و', 'X' => 'خ',
  66                 'Y' => 'ی', 'Z' => 'ز',
  67
  68                 /* Doppelkonsonanten */
  69                 # 'll' => 'ڵ', # wenn es geht, doppel-l und l getrennt zu behandeln
  70                 # 'rr' => 'ڕ', # selbiges für doppel-r
  71
  72                 /* Einzelne Großbuchstaben */
  73                 //' C' => 'ج',
  74
  75                 /* Vowels */
  76                 'a' => 'ا',
  77                 'e' => 'ە',
  78                 'ê' => 'ێ',
  79                 'i' => '',
  80                 'î' => 'ی',
  81                 'o' => 'ۆ',
  82                 'u' => 'و',
  83                 'û' => 'وو',
  84                 'w' => 'و',
  85                 ',' => '،',
  86                 '?' => '؟',
  87
  88                 # Try to replace the leading vowel
  89                 ' a' => 'ئا ',
  90                 ' e' => 'ئە ',
  91                 ' ê' => 'ئێ ',
  92                 ' î' => 'ئی ',
  93                 ' o' => 'ئۆ ',
  94                 ' u' => 'ئو ',
  95                 ' û' => 'ئوو ',
  96                 'A'  => 'ئا',
  97                 'E'  => 'ئە',
  98                 'Ê'  => 'ئێ',
  99                 'Î'  => 'ئی',
 100                 'O'  => 'ئۆ',
 101                 'U'  => 'ئو',
 102                 'Û'  => 'ئوو',
 103                 ' A' => 'ئا ',
 104                 ' E' => 'ئە ',
 105                 ' Ê' => 'ئێ ',
 106                 ' Î' => 'ئی ',
 107                 ' O' => 'ئۆ ',
 108                 ' U' => 'ئو ',
 109                 ' Û' => 'ئوو ',
 110                 # eyn erstmal deaktivieren, einfache Anführungsstriche sind einfach zu häufig, um sie als eyn zu interpretieren
 111                 # '\'' => 'ع',
 112
 113 /*              # deactivated for now, breaks links i.e. in header of Special:Recentchanges :-(
 114                 # digits
 115                 '0' => '٠', # &#x0660;
 116                 '1' => '١', # &#x0661;
 117                 '2' => '٢', # &#x0662;
 118                 '3' => '٣', # &#x0663;
 119                 '4' => '٤', # &#x0664;
 120                 '5' => '٥', # &#x0665;
 121                 '6' => '٦', # &#x0666;
 122                 '7' => '٧', # &#x0667;
 123                 '8' => '٨', # &#x0668;
 124                 '9' => '٩', # &#x0669;
 125 */
 126                 );
 127
 128         function loadDefaultTables() {
 129                 $this->mTables = array(
 130                         'ku-latn' => new ReplacementArray( $this->mArabicToLatin ),
 131                         'ku-arab' => new ReplacementArray( $this->mLatinToArabic ),
 132                         'ku'      => new ReplacementArray()
 133                 );
 134         }
 135
 136
 137         // Do not convert content on talk pages
 138         function parserConvert( $text, &$parser ){
 139                 if(is_object($parser->getTitle() ) && $parser->getTitle()->isTalkPage())
 140                         $this->mDoContentConvert=false;
 141                 else
 142                         $this->mDoContentConvert=true;
 143
 144                 return parent::parserConvert($text, $parser );
 145         }
 146
 147         /*
 148          * A function wrapper:
 149          *   - if there is no selected variant, leave the link
 150          *     names as they were
 151          *   - do not try to find variants for usernames
 152          */
 153         function findVariantLink( &$link, &$nt ) {
 154                 // check for user namespace
 155                 if(is_object($nt)){
 156                         $ns = $nt->getNamespace();
 157                         if($ns==NS_USER || $ns==NS_USER_TALK)
 158                                 return;
 159                 }
 160
 161                 $oldlink=$link;
 162                 parent::findVariantLink($link,$nt);
 163                 if($this->getPreferredVariant()==$this->mMainLanguageCode)
 164                         $link=$oldlink;
 165         }
 166
 167         /*
 168          * We want our external link captions to be converted in variants,
 169          * so we return the original text instead -{$text}-, except for URLs
 170          */
 171         function markNoConversion($text, $noParse=false) {
 172                 if($noParse || preg_match("/^https?:\/\/|ftp:\/\/|irc:\/\//",$text))
 173                     return parent::markNoConversion($text);
 174                 return $text;
 175         }
 176
 177         /*
 178          * An ugly function wrapper for parsing Image titles
 179          * (to prevent image name conversion)
 180          */
 181         function autoConvert($text, $toVariant=false) {
 182                 global $wgTitle;
 183                 if(is_object($wgTitle) && $wgTitle->getNameSpace()==NS_IMAGE){
 184                         $imagename = $wgTitle->getNsText();
 185                         if(preg_match("/^$imagename:/",$text)) return $text;
 186                 }
 187                 return parent::autoConvert($text,$toVariant);
 188         }
 189
 190         /**
 191          *  It translates text into variant, specials:
 192          *    - ommiting roman numbers
 193          */
 194         function translate($text, $toVariant){
 195                 $breaks = '[^\w\x80-\xff]';
 196
 197                 /* From Kazakh interface, maybe we need it later
 198                  *
 199                 // regexp for roman numbers
 200                 $roman = 'M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3})';
 201                 $roman = '';
 202
 203                 $reg = '/^'.$roman.'$|^'.$roman.$breaks.'|'.$breaks.$roman.'$|'.$breaks.$roman.$breaks.'/';
 204
 205                 $matches = preg_split($reg, $text, -1, PREG_SPLIT_OFFSET_CAPTURE);
 206
 207                 $m = array_shift($matches);
 208                 if( !isset( $this->mTables[$toVariant] ) ) {
 209                         throw new MWException( "Broken variant table: " . implode( ',', array_keys( $this->mTables ) ) );
 210                 }
 211                 $ret = $this->mTables[$toVariant]->replace( $m[0] );
 212                 $mstart = $m[1]+strlen($m[0]);
 213                 foreach($matches as $m) {
 214                         $ret .= substr($text, $mstart, $m[1]-$mstart);
 215                         $ret .= parent::translate($m[0], $toVariant);
 216                         $mstart = $m[1] + strlen($m[0]);
 217                 }
 218
 219                 return $ret;
 220                 */
 221
 222                 if( !isset( $this->mTables[$toVariant] ) ) {
 223                         throw new MWException( "Broken variant table: " . implode( ',', array_keys( $this->mTables ) ) );
 224                 }
 225
 226                 return parent::translate( $text, $toVariant );
 227         }
 228 }
 229
 230 class LanguageKu extends LanguageKu_ku {
 231
 232         function __construct() {
 233                 global $wgHooks;
 234                 parent::__construct();
 235
 236                 $variants = array( 'ku', 'ku-arab', 'ku-latn' );
 237                 $variantfallbacks = array(
 238                         'ku'      => 'ku-latn',
 239                         'ku-arab' => 'ku-latn',
 240                         'ku-latn' => 'ku-arab',
 241                 );
 242
 243                 $this->mConverter = new KuConverter( $this, 'ku', $variants, $variantfallbacks );
 244                 $wgHooks['ArticleSaveComplete'][] = $this->mConverter;
 245         }
 246
 247 /*   From Kazakh interface, not needed for the moment
 248
 249         function convertGrammar( $word, $case ) {
 250                 $fname="LanguageKu::convertGrammar";
 251                 wfProfileIn( $fname );
 252
 253                 //always convert to ku-latn before convertGrammar
 254                 $w1 = $word;
 255                 $word = $this->mConverter->autoConvert( $word, 'ku-latn' );
 256                 $w2 = $word;
 257                 $word = parent::convertGrammar( $word, $case );
 258                 //restore encoding
 259                 if( $w1 != $w2 ) {
 260                         $word = $this->mConverter->translate( $word, 'ku-latn' );
 261                 }
 262                 wfProfileOut( $fname );
 263                 return $word;
 264         }
 265 */
 266 }