languages/classes/LanguageKk.php

   1 <?php
   2 /** Kazakh (Қазақша)
   3   * converter routines
   4   *
   5   * @addtogroup Language
   6   */
   7
   8 require_once( dirname(__FILE__).'/../LanguageConverter.php' );
   9 require_once( dirname(__FILE__).'/LanguageKk_cyrl.php' );
  10
  11 define( 'KK_C_UC', 'АӘБВГҒДЕЁЖЗИЙКҚЛМНҢОӨПРСТУҰҮФХҺЦЧШЩЪЫІЬЭЮЯ' ); # Kazakh Cyrillic uppercase
  12 define( 'KK_C_LC', 'аәбвгғдеёжзийкқлмнңоөпрстуұүфхһцчшщъыіьэюя' ); # Kazakh Cyrillic lowercase
  13 define( 'KK_L_UC', 'AÄBCÇDEÉFGĞHIİÏJKLMNÑOÖPQRSŞTUÜVWXYÝZ' ); # Kazakh Latin uppercase
  14 define( 'KK_L_LC', 'aäbcçdeéfgğhıiïjklmnñoöpqrsştuüvwxyýz' ); # Kazakh Latin lowercase
  15 define( 'KK_A', 'اٵبۆگعدەجزيكقلمنڭوٶپرستۋۇٷفحھچشىٸ' ); # Kazakh Arabic
  16
  17 class KkConverter extends LanguageConverter {
  18
  19         function loadDefaultTables() {
  20                 // require( dirname(__FILE__)."/../../includes/KkConversion.php" );
  21                 // Placeholder for future implementing. Remove variables declarations
  22                 // after generating KkConversion.php
  23                 $kk2Cyrl = array();
  24                 $kk2Latn = array();
  25                 $kk2Arab = array();
  26                 $kk2KZ = array();
  27                 $kk2TR = array();
  28                 $kk2CN = array();
  29
  30                 $this->mTables = array(
  31                         'kk-cyrl'       => new ReplacementArray( $kk2Cyrl ),
  32                         'kk-latn'       => new ReplacementArray( $kk2Latn ),
  33                         'kk-arab'       => new ReplacementArray( $kk2Arab ),
  34                         'kk-kz'         => new ReplacementArray( array_merge($kk2Cyrl, $kk2KZ) ),
  35                         'kk-tr'         => new ReplacementArray( array_merge($kk2Latn, $kk2TR) ),
  36                         'kk-cn'         => new ReplacementArray( array_merge($kk2Arab, $kk2CN) ),
  37                         'kk'            => new ReplacementArray()
  38                 );
  39
  40                 self::loadRegs();
  41         }
  42
  43         function postLoadTables() {
  44                 $this->mTables['kk-kz']->merge( $this->mTables['kk-cyrl'] );
  45                 $this->mTables['kk-tr']->merge( $this->mTables['kk-latn'] );
  46                 $this->mTables['kk-cn']->merge( $this->mTables['kk-arab'] );
  47         }
  48
  49         function loadRegs() {
  50
  51                 $this->mCyrl2Latn = array(
  52                         '/№/u' => 'No.',
  53                         ## Е after vowels
  54                         '/([АӘЕЁИОӨҰҮЭЮЯЪЬ])Е/u' => '$1YE',
  55                         '/([АӘЕЁИОӨҰҮЭЮЯЪЬ])е/ui' => '$1ye',
  56                         ## leading ЁЮЯЩ
  57                         '/^Ё(['.KK_C_UC.']|$)/u' => 'YO$1', '/^Ё(['.KK_C_LC.']|$)/u' => 'Yo$1',
  58                         '/^Ю(['.KK_C_UC.']|$)/u' => 'YU$1', '/^Ю(['.KK_C_LC.']|$)/u' => 'Yu$1',
  59                         '/^Я(['.KK_C_UC.']|$)/u' => 'YA$1', '/^Я(['.KK_C_LC.']|$)/u' => 'Ya$1',
  60                         '/^Щ(['.KK_C_UC.']|$)/u' => 'ŞÇ$1', '/^Щ(['.KK_C_LC.']|$)/u' => 'Şç$1',
  61                         ## other ЁЮЯ
  62                         '/Ё/u' => 'YO', '/ё/u' => 'yo',
  63                         '/Ю/u' => 'YU', '/ю/u' => 'yu',
  64                         '/Я/u' => 'YA', '/я/u' => 'ya',
  65                         '/Щ/u' => 'ŞÇ', '/щ/u' => 'şç',
  66                         ## soft and hard signs
  67                         '/[ъЪ]/u' => 'ʺ', '/[ьЬ]/u' => 'ʹ',
  68                         ## other characters
  69                         '/А/u' => 'A', '/а/u' => 'a', '/Ә/u' => 'Ä', '/ә/u' => 'ä',
  70                         '/Б/u' => 'B', '/б/u' => 'b', '/В/u' => 'V', '/в/u' => 'v',
  71                         '/Г/u' => 'G', '/г/u' => 'g', '/Ғ/u' => 'Ğ', '/ғ/u' => 'ğ',
  72                         '/Д/u' => 'D', '/д/u' => 'd', '/Е/u' => 'E', '/е/u' => 'e',
  73                         '/Ж/u' => 'J', '/ж/u' => 'j', '/З/u' => 'Z', '/з/u' => 'z',
  74                         '/И/u' => 'Ï', '/и/u' => 'ï', '/Й/u' => 'Ý', '/й/u' => 'ý',
  75                         '/К/u' => 'K', '/к/u' => 'k', '/Қ/u' => 'Q', '/қ/u' => 'q',
  76                         '/Л/u' => 'L', '/л/u' => 'l', '/М/u' => 'M', '/м/u' => 'm',
  77                         '/Н/u' => 'N', '/н/u' => 'n', '/Ң/u' => 'Ñ', '/ң/u' => 'ñ',
  78                         '/О/u' => 'O', '/о/u' => 'o', '/Ө/u' => 'Ö', '/ө/u' => 'ö',
  79                         '/П/u' => 'P', '/п/u' => 'p', '/Р/u' => 'R', '/р/u' => 'r',
  80                         '/С/u' => 'S', '/с/u' => 's', '/Т/u' => 'T', '/т/u' => 't',
  81                         '/У/u' => 'W', '/у/u' => 'w', '/Ұ/u' => 'U', '/ұ/u' => 'u',
  82                         '/Ү/u' => 'Ü', '/ү/u' => 'ü', '/Ф/u' => 'F', '/ф/u' => 'f',
  83                         '/Х/u' => 'X', '/х/u' => 'x', '/Һ/u' => 'H', '/һ/u' => 'h',
  84                         '/Ц/u' => 'C', '/ц/u' => 'c', '/Ч/u' => 'Ç', '/ч/u' => 'ç',
  85                         '/Ш/u' => 'Ş', '/ш/u' => 'ş', '/Ы/u' => 'I', '/ы/u' => 'ı',
  86                         '/І/u' => 'İ', '/і/u' => 'i', '/Э/u' => 'É', '/э/u' => 'é',
  87                 );
  88
  89                 $this->mLatn2Cyrl = array(
  90                         '/No\./u' => '№',
  91                         ## Şç
  92                         '/ŞÇʹ/u'=> 'ЩЬ', '/Şçʹ/u'=> 'Щь', '/Şçʹ/u'=> 'Щь',
  93                         '/Ş[Çç]/u' => 'Щ', '/şç/u' => 'щ',
  94                         ## soft and hard signs
  95                         '/(['.KK_L_UC.'])ʺ(['.KK_L_UC.'])/u' => '$1Ъ$2',
  96                         '/ʺ(['.KK_L_LC.'])/u' => 'ъ$1',
  97                         '/(['.KK_L_UC.'])ʹ(['.KK_L_UC.'])/u' => '$1Ь$2',
  98                         '/ʹ(['.KK_L_LC.'])/u' => 'ь$1',
  99                         '/ʺ/u' => 'ъ',
 100                         '/ʹ/u' => 'ь',
 101                         ## Ye Yo Yu Ya.
 102                         '/Y[Ee]/u' => 'Е', '/ye/u' => 'е',
 103                         '/Y[Oo]/u' => 'Ё', '/yo/u' => 'ё',
 104                         '/Y[UWuw]/u' => 'Ю', '/y[uw]/u' => 'ю',
 105                         '/Y[Aa]/u' => 'Я', '/ya/u' => 'я',
 106                         ## other characters
 107                         '/A/u' => 'А', '/a/u' => 'а', '/Ä/u' => 'Ә', '/ä/u' => 'ә',
 108                         '/B/u' => 'Б', '/b/u' => 'б', '/C/u' => 'Ц', '/c/u' => 'ц',
 109                         '/Ç/u' => 'Ч', '/ç/u' => 'ч', '/D/u' => 'Д', '/d/u' => 'д',
 110                         '/E/u' => 'Е', '/e/u' => 'е', '/É/u' => 'Э', '/é/u' => 'э',
 111                         '/F/u' => 'Ф', '/f/u' => 'ф', '/G/u' => 'Г', '/g/u' => 'г',
 112                         '/Ğ/u' => 'Ғ', '/ğ/u' => 'ғ', '/H/u' => 'Һ', '/h/u' => 'һ',
 113                         '/I/u' => 'Ы', '/ı/u' => 'ы', '/İ/u' => 'І', '/i/u' => 'і',
 114                         '/Ï/u' => 'И', '/ï/u' => 'и', '/J/u' => 'Ж', '/j/u' => 'ж',
 115                         '/K/u' => 'К', '/k/u' => 'к', '/L/u' => 'Л', '/l/u' => 'л',
 116                         '/M/u' => 'М', '/m/u' => 'м', '/N/u' => 'Н', '/n/u' => 'н',
 117                         '/Ñ/u' => 'Ң', '/ñ/u' => 'ң', '/O/u' => 'О', '/o/u' => 'о',
 118                         '/Ö/u' => 'Ө', '/ö/u' => 'ө', '/P/u' => 'П', '/p/u' => 'п',
 119                         '/Q/u' => 'Қ', '/q/u' => 'қ', '/R/u' => 'Р', '/r/u' => 'р',
 120                         '/S/u' => 'С', '/s/u' => 'с', '/Ş/u' => 'Ш', '/ş/u' => 'ш',
 121                         '/T/u' => 'Т', '/t/u' => 'т', '/U/u' => 'Ұ', '/u/u' => 'ұ',
 122                         '/Ü/u' => 'Ү', '/ü/u' => 'ү', '/V/u' => 'В', '/v/u' => 'в',
 123                         '/W/u' => 'У', '/w/u' => 'у', '/Ý/u' => 'Й', '/ý/u' => 'й',
 124                         '/X/u' => 'Х', '/x/u' => 'х', '/Z/u' => 'З', '/z/u' => 'з',
 125                 );
 126
 127                 $this->mCyLa2Arab = array(
 128                         ## Cyrillic -> Arabic
 129                         '/\№/u' => 'نٶ.',
 130                         '/([АӘЕЁИОӨҰҮЭЮЯЪЬ])е/ui' => '$1يە',
 131                         '/а/ui' => 'ا', '/ә/ui' => 'ٵ', '/б/ui' => 'ب', '/в/ui' => 'ۆ',
 132                         '/г/ui' => 'گ', '/ғ/ui' => 'ع', '/д/ui' => 'د', '/[еэ]/ui' => 'ە',
 133                         '/ё/ui' => 'يو', '/ж/ui' => 'ج', '/з/ui' => 'ز', '/и/ui' => 'ي',
 134                         '/й/ui' => 'ي', '/к/ui' => 'ك', '/қ/ui' => 'ق', '/л/ui' => 'ل',
 135                         '/м/ui' => 'م', '/н/ui' => 'ن', '/ң/ui' => 'ڭ', '/о/ui' => 'و',
 136                         '/ө/ui' => 'ٶ', '/п/ui' => 'پ', '/р/ui' => 'ر', '/с/ui' => 'س',
 137                         '/т/ui' => 'ت', '/у/ui' => 'ۋ', '/ұ/ui' => 'ۇ', '/ү/ui' => 'ٷ',
 138                         '/ф/ui' => 'ف', '/х/ui' => 'ح', '/һ/ui' => 'ھ', '/ц/ui' => 'تس',
 139                         '/ч/ui' => 'چ', '/ш/ui' => 'ش', '/щ/ui' => 'شش', '/[ъь]/ui' => '',
 140                         '/ы/ui' => 'ى', '/і/ui' => 'ٸ', '/ю/ui' => 'يۋ', '/я/ui' => 'يا',
 141
 142                         ## Latin -> Arabic // commented for now...
 143                         /*'/No\./u' => 'نٶ.',
 144                         '/a/ui' => 'ا',  '/ä/ui' => 'ٵ',  '/b/ui' => 'ب', '/c/ui' => 'تس',
 145                         '/ç/ui' => 'چ', '/d/ui' => 'د', '/[eé]/ui' => 'ە', '/f/ui' => 'ف',
 146                         '/g/ui' => 'گ',  '/ğ/ui' => 'ع', '/h/ui' => 'ھ', '/[ıI]/u' => 'ى',
 147                         '/[iİ]/u' => 'ٸ', '/ï/ui' => 'ي', '/j/ui' => 'ج', '/k/ui' => 'ك',
 148                         '/l/ui' => 'ل', '/m/ui' => 'م', '/n/ui' => 'ن', '/ñ/ui' => 'ڭ',
 149                         '/o/ui' => 'و', '/ö/ui' => 'ٶ', '/p/ui' => 'پ', '/q/ui' => 'ق',
 150                         '/r/ui' => 'ر', '/s/ui' => 'س', '/ş/ui' => 'ش', '/t/ui' => 'ت',
 151                         '/u/ui' => 'ۇ', '/ü/ui' => 'ٷ', '/v/ui' => 'ۆ', '/w/ui' => 'ۋ',
 152                         '/x/ui' => 'ح', '/[yý]/ui' => 'ي', '/z/ui' => 'ز', '/[ʺʹ]/ui' => '',*/
 153
 154                         ## Punctuation -> Arabic
 155                         '/\?/' => '؟', # &#x061F;
 156                         '/\,/' => '،', # &#x060C;
 157                         '/\;/' => '؛' , # &#x061B;
 158                         '/\%/' => '٪', # &#x066a;
 159                         ## Digits -> Arabic
 160                         '/0/' => '۰',  # &#x06f0;
 161                         '/1/' => '۱', # &#x06f1;
 162                         '/2/' => '۲', # &#x06f2;
 163                         '/3/' => '۳', # &#x06f3;
 164                         '/4/' => '۴', # &#x06f4;
 165                         '/5/' => '۵', # &#x06f5;
 166                         '/6/' => '۶', # &#x06f6;
 167                         '/7/' => '۷', # &#x06f7;
 168                         '/8/' => '۸', # &#x06f8;
 169                         '/9/' => '۹', # &#x06f9;
 170                 );
 171
 172         }
 173
 174         /* rules should be defined as -{ekavian | iyekavian-} -or-
 175                 -{code:text | code:text | ...}-
 176                 update: delete all rule parsing because it's not used
 177                                 currently, and just produces a couple of bugs
 178         */
 179         function parseManualRule($rule, $flags=array()) {
 180                 if(in_array('T',$flags)){
 181                         return parent::parseManualRule($rule, $flags);
 182                 }
 183
 184                 // otherwise ignore all formatting
 185                 foreach($this->mVariants as $v) {
 186                         $carray[$v] = $rule;
 187                 }
 188
 189                 return $carray;
 190         }
 191
 192         // Do not convert content on talk pages
 193         function parserConvert( $text, &$parser ){
 194                 if(is_object($parser->getTitle() ) && $parser->getTitle()->isTalkPage())
 195                         $this->mDoContentConvert=false;
 196                 else
 197                         $this->mDoContentConvert=true;
 198
 199                 return parent::parserConvert($text, $parser );
 200         }
 201
 202         /*
 203          * A function wrapper:
 204          *  - if there is no selected variant, leave the link
 205          *    names as they were
 206          *  - do not try to find variants for usernames
 207          */
 208         function findVariantLink( &$link, &$nt ) {
 209                 // check for user namespace
 210                 if(is_object($nt)){
 211                         $ns = $nt->getNamespace();
 212                         if($ns==NS_USER || $ns==NS_USER_TALK)
 213                                 return;
 214                 }
 215
 216                 $oldlink=$link;
 217                 parent::findVariantLink($link,$nt);
 218                 if($this->getPreferredVariant()==$this->mMainLanguageCode)
 219                         $link=$oldlink;
 220         }
 221
 222         /*
 223          * An ugly function wrapper for parsing Image titles
 224          * (to prevent image name conversion)
 225          */
 226         function autoConvert($text, $toVariant=false) {
 227                 global $wgTitle;
 228                 if(is_object($wgTitle) && $wgTitle->getNameSpace()==NS_IMAGE){
 229                         $imagename = $wgTitle->getNsText();
 230                         if(preg_match("/^$imagename:/",$text)) return $text;
 231                 }
 232                 return parent::autoConvert($text,$toVariant);
 233         }
 234
 235         /**
 236          *  It translates text into variant
 237          */
 238         function translate( $text, $toVariant ){
 239                 global $wgContLanguageCode;
 240                 $text = parent::translate( $text, $toVariant );
 241
 242                 $letters = '';
 243                 switch( $toVariant ) {
 244                         case 'kk-cyrl':
 245                         case 'kk-kz':
 246                                 $letters = KK_L_UC . KK_L_LC . 'ʺʹ0123456789';
 247                                 $wgContLanguageCode = 'kk';
 248                                 break;
 249                         case 'kk-latn':
 250                         case 'kk-tr':
 251                                 $letters = KK_C_UC . KK_C_LC . '№0123456789';
 252                                 $wgContLanguageCode = 'kk-Latn';
 253                                 break;
 254                         case 'kk-arab':
 255                         case 'kk-cn':
 256                                 // $letters = KK_C_UC.KK_C_LC.KK_L_UC.KK_L_LC.'ʺʹ%№0123456789?,;';
 257                                 $letters = KK_C_UC . KK_C_LC . '%№0123456789?,;';
 258                                 $wgContLanguageCode = 'kk-Arab';
 259                                 break;
 260                         default:
 261                                 $wgContLanguageCode = 'kk';
 262                                 return $text;
 263                 }
 264                 // disable conversion variables like $1, $2...
 265                 $varsfix = '\$[0-9]';
 266
 267                 $matches = preg_split( '/' . $varsfix . '[^' . $letters . ']+/u', $text, -1, PREG_SPLIT_OFFSET_CAPTURE);
 268                 $mstart = 0;
 269                 $ret = '';
 270                 foreach( $matches as $m ) {
 271                         $ret .= substr( $text, $mstart, $m[1]-$mstart );
 272                         $ret .= $this->regsConverter( $m[0], $toVariant );
 273                         $mstart = $m[1] + strlen($m[0]);
 274                 }
 275                 return $ret;
 276         }
 277
 278         function regsConverter( $text, $toVariant ) {
 279                 if ($text == '') return $text;
 280
 281                 $pat = array();
 282                 $rep = array();
 283                 switch( $toVariant ) {
 284                         case 'kk-arab':
 285                         case 'kk-cn':
 286                                 foreach( $this->mCyLa2Arab as $pat => $rep ) {
 287                                         $text = preg_replace( $pat, $rep, $text );
 288                                 }
 289                                 return $text;
 290                                 break;
 291                         case 'kk-latn':
 292                         case 'kk-tr':
 293                                 foreach( $this->mCyrl2Latn as $pat => $rep ) {
 294                                         $text = preg_replace( $pat, $rep, $text );
 295                                 }
 296                                 return $text;
 297                                 break;
 298                         case 'kk-cyrl':
 299                         case 'kk-cn':
 300                                 foreach( $this->mLatn2Cyrl as $pat => $rep ) {
 301                                         $text = preg_replace( $pat, $rep, $text );
 302                                 }
 303                                 return $text;
 304                                 break;
 305                         default:
 306                                 return $text;
 307                 }
 308         }
 309
 310         /*
 311          * We want our external link captions to be converted in variants,
 312          * so we return the original text instead -{$text}-, except for URLs
 313          */
 314         function markNoConversion( $text, $noParse=false ) {
 315                 if( $noParse || preg_match( "/^https?:\/\/|ftp:\/\/|irc:\/\//", $text ) )
 316                         return parent::markNoConversion( $text );
 317                 return $text;
 318         }
 319
 320         function convertCategoryKey( $key ) {
 321                 return $this->autoConvert( $key, 'kk' );
 322         }
 323
 324 }
 325
 326 /* class that handles Cyrillic, Latin and Arabic scripts for Kazakh
 327    right now it only distinguish kk_cyrl, kk_latn, kk_arab and kk_kz, kk_tr, kk_cn.
 328 */
 329 class LanguageKk extends LanguageKk_cyrl {
 330
 331         function __construct() {
 332                 global $wgHooks;
 333                 parent::__construct();
 334
 335                 $variants = array( 'kk', 'kk-cyrl', 'kk-latn', 'kk-arab', 'kk-kz', 'kk-tr', 'kk-cn' );
 336                 $variantfallbacks = array(
 337                         'kk'      => 'kk-kz',
 338                         'kk-cyrl' => 'kk',
 339                         'kk-latn' => 'kk',
 340                         'kk-arab' => 'kk',
 341                         'kk-kz'   => 'kk-cyrl',
 342                         'kk-tr'   => 'kk-latn',
 343                         'kk-cn'   => 'kk-arab'
 344                 );
 345
 346                 $this->mConverter = new KkConverter( $this, 'kk', $variants, $variantfallbacks );
 347
 348                 $wgHooks['ArticleSaveComplete'][] = $this->mConverter;
 349         }
 350
 351         /**
 352          * Work around for right-to-left direction support in kk-arab and kk-cn
 353          *
 354          * @return bool
 355          */
 356         function isRTL() {
 357                 $variant = $this->getPreferredVariant();
 358                 if ( $variant == 'kk-arab' || $variant == 'kk-cn' ) {
 359                         return true;
 360                 } else {
 361                         return parent::isRTL();
 362                 }
 363         }
 364
 365         /*
 366          * It fixes issue with ucfirst for transforming 'i' to 'İ'
 367          *
 368          */
 369         function ucfirst ( $string ) {
 370                 $variant = $this->getPreferredVariant();
 371                 if ( ($variant == 'kk-latn' || $variant == 'kk-tr') && $string[0] == 'i' ) {
 372                         $string = 'İ' . substr( $string, 1 );
 373                 } else {
 374                         $string = parent::ucfirst( $string );
 375                 }
 376                 return $string;
 377         }
 378
 379         /*
 380          * It fixes issue with  lcfirst for transforming 'I' to 'ı'
 381          *
 382          */
 383         function lcfirst ( $string ) {
 384                 $variant = $this->getPreferredVariant();
 385                 if ( ($variant == 'kk-latn' || $variant == 'kk-tr') && $string[0] == 'I' ) {
 386                         $string = 'ı' . substr( $string, 1 );
 387                 } else {
 388                         $string = parent::lcfirst( $string );
 389                 }
 390                 return $string;
 391         }
 392
 393         function convertGrammar( $word, $case ) {
 394                 $fname="LanguageKk::convertGrammar";
 395                 wfProfileIn( $fname );
 396
 397                 $variant = $this->getPreferredVariant();
 398                 switch ( $variant ) {
 399                         case 'kk-arab':
 400                         case 'kk-cn':
 401                                 $word = parent::convertGrammar( $word, $case, $variant = 'kk-arab' );
 402                                 break;
 403                         case 'kk-latn':
 404                         case 'kk-tr':
 405                                 $word = parent::convertGrammar( $word, $case, $variant = 'kk-latn' );
 406                                 break;
 407                         case 'kk-cyrl':
 408                         case 'kk-kz':
 409                         case 'kk':
 410                         default:
 411                                 $word = parent::convertGrammar( $word, $case, $variant = 'kk-cyrl' );
 412                 }
 413
 414                 wfProfileOut( $fname );
 415                 return $word;
 416         }
 417
 418 }