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 __construct($langobj, $maincode,
  20                                                                 $variants=array(),
  21                                                                 $variantfallbacks=array(),
  22                                                                 $markup=array(),
  23                                                                 $flags = array()) {
  24                 parent::__construct( $langobj, $maincode,
  25                         $variants, $variantfallbacks, $markup, $flags );
  26
  27                 // No point delaying this since they're in code.
  28                 // Waiting until loadDefaultTables() means they never get loaded
  29                 // when the tables themselves are loaded from cache.
  30                 $this->loadRegs();
  31         }
  32
  33         function loadDefaultTables() {
  34                 // require( dirname(__FILE__)."/../../includes/KkConversion.php" );
  35                 // Placeholder for future implementing. Remove variables declarations
  36                 // after generating KkConversion.php
  37                 $kk2Cyrl = array();
  38                 $kk2Latn = array();
  39                 $kk2Arab = array();
  40                 $kk2KZ = array();
  41                 $kk2TR = array();
  42                 $kk2CN = array();
  43
  44                 $this->mTables = array(
  45                         'kk-cyrl'       => new ReplacementArray( $kk2Cyrl ),
  46                         'kk-latn'       => new ReplacementArray( $kk2Latn ),
  47                         'kk-arab'       => new ReplacementArray( $kk2Arab ),
  48                         'kk-kz'         => new ReplacementArray( array_merge($kk2Cyrl, $kk2KZ) ),
  49                         'kk-tr'         => new ReplacementArray( array_merge($kk2Latn, $kk2TR) ),
  50                         'kk-cn'         => new ReplacementArray( array_merge($kk2Arab, $kk2CN) ),
  51                         'kk'            => new ReplacementArray()
  52                 );
  53         }
  54
  55         function postLoadTables() {
  56                 $this->mTables['kk-kz']->merge( $this->mTables['kk-cyrl'] );
  57                 $this->mTables['kk-tr']->merge( $this->mTables['kk-latn'] );
  58                 $this->mTables['kk-cn']->merge( $this->mTables['kk-arab'] );
  59         }
  60
  61         function loadRegs() {
  62
  63                 $this->mCyrl2Latn = array(
  64                         '/№/u' => 'No.',
  65                         ## Е after vowels
  66                         '/([АӘЕЁИОӨҰҮЭЮЯЪЬ])Е/u' => '$1YE',
  67                         '/([АӘЕЁИОӨҰҮЭЮЯЪЬ])е/ui' => '$1ye',
  68                         ## leading ЁЮЯЩ
  69                         '/^Ё(['.KK_C_UC.']|$)/u' => 'YO$1', '/^Ё(['.KK_C_LC.']|$)/u' => 'Yo$1',
  70                         '/^Ю(['.KK_C_UC.']|$)/u' => 'YU$1', '/^Ю(['.KK_C_LC.']|$)/u' => 'Yu$1',
  71                         '/^Я(['.KK_C_UC.']|$)/u' => 'YA$1', '/^Я(['.KK_C_LC.']|$)/u' => 'Ya$1',
  72                         '/^Щ(['.KK_C_UC.']|$)/u' => 'ŞÇ$1', '/^Щ(['.KK_C_LC.']|$)/u' => 'Şç$1',
  73                         ## other ЁЮЯ
  74                         '/Ё/u' => 'YO', '/ё/u' => 'yo',
  75                         '/Ю/u' => 'YU', '/ю/u' => 'yu',
  76                         '/Я/u' => 'YA', '/я/u' => 'ya',
  77                         '/Щ/u' => 'ŞÇ', '/щ/u' => 'şç',
  78                         ## soft and hard signs
  79                         '/[ъЪ]/u' => 'ʺ', '/[ьЬ]/u' => 'ʹ',
  80                         ## other characters
  81                         '/А/u' => 'A', '/а/u' => 'a', '/Ә/u' => 'Ä', '/ә/u' => 'ä',
  82                         '/Б/u' => 'B', '/б/u' => 'b', '/В/u' => 'V', '/в/u' => 'v',
  83                         '/Г/u' => 'G', '/г/u' => 'g', '/Ғ/u' => 'Ğ', '/ғ/u' => 'ğ',
  84                         '/Д/u' => 'D', '/д/u' => 'd', '/Е/u' => 'E', '/е/u' => 'e',
  85                         '/Ж/u' => 'J', '/ж/u' => 'j', '/З/u' => 'Z', '/з/u' => 'z',
  86                         '/И/u' => 'Ï', '/и/u' => 'ï', '/Й/u' => 'Ý', '/й/u' => 'ý',
  87                         '/К/u' => 'K', '/к/u' => 'k', '/Қ/u' => 'Q', '/қ/u' => 'q',
  88                         '/Л/u' => 'L', '/л/u' => 'l', '/М/u' => 'M', '/м/u' => 'm',
  89                         '/Н/u' => 'N', '/н/u' => 'n', '/Ң/u' => 'Ñ', '/ң/u' => 'ñ',
  90                         '/О/u' => 'O', '/о/u' => 'o', '/Ө/u' => 'Ö', '/ө/u' => 'ö',
  91                         '/П/u' => 'P', '/п/u' => 'p', '/Р/u' => 'R', '/р/u' => 'r',
  92                         '/С/u' => 'S', '/с/u' => 's', '/Т/u' => 'T', '/т/u' => 't',
  93                         '/У/u' => 'W', '/у/u' => 'w', '/Ұ/u' => 'U', '/ұ/u' => 'u',
  94                         '/Ү/u' => 'Ü', '/ү/u' => 'ü', '/Ф/u' => 'F', '/ф/u' => 'f',
  95                         '/Х/u' => 'X', '/х/u' => 'x', '/Һ/u' => 'H', '/һ/u' => 'h',
  96                         '/Ц/u' => 'C', '/ц/u' => 'c', '/Ч/u' => 'Ç', '/ч/u' => 'ç',
  97                         '/Ш/u' => 'Ş', '/ш/u' => 'ş', '/Ы/u' => 'I', '/ы/u' => 'ı',
  98                         '/І/u' => 'İ', '/і/u' => 'i', '/Э/u' => 'É', '/э/u' => 'é',
  99                 );
 100
 101                 $this->mLatn2Cyrl = array(
 102                         '/No\./u' => '№',
 103                         ## Şç
 104                         '/ŞÇʹ/u'=> 'ЩЬ', '/Şçʹ/u'=> 'Щь', '/Şçʹ/u'=> 'Щь',
 105                         '/Ş[Çç]/u' => 'Щ', '/şç/u' => 'щ',
 106                         ## soft and hard signs
 107                         '/(['.KK_L_UC.'])ʺ(['.KK_L_UC.'])/u' => '$1Ъ$2',
 108                         '/ʺ(['.KK_L_LC.'])/u' => 'ъ$1',
 109                         '/(['.KK_L_UC.'])ʹ(['.KK_L_UC.'])/u' => '$1Ь$2',
 110                         '/ʹ(['.KK_L_LC.'])/u' => 'ь$1',
 111                         '/ʺ/u' => 'ъ',
 112                         '/ʹ/u' => 'ь',
 113                         ## Ye Yo Yu Ya.
 114                         '/Y[Ee]/u' => 'Е', '/ye/u' => 'е',
 115                         '/Y[Oo]/u' => 'Ё', '/yo/u' => 'ё',
 116                         '/Y[UWuw]/u' => 'Ю', '/y[uw]/u' => 'ю',
 117                         '/Y[Aa]/u' => 'Я', '/ya/u' => 'я',
 118                         ## other characters
 119                         '/A/u' => 'А', '/a/u' => 'а', '/Ä/u' => 'Ә', '/ä/u' => 'ә',
 120                         '/B/u' => 'Б', '/b/u' => 'б', '/C/u' => 'Ц', '/c/u' => 'ц',
 121                         '/Ç/u' => 'Ч', '/ç/u' => 'ч', '/D/u' => 'Д', '/d/u' => 'д',
 122                         '/E/u' => 'Е', '/e/u' => 'е', '/É/u' => 'Э', '/é/u' => 'э',
 123                         '/F/u' => 'Ф', '/f/u' => 'ф', '/G/u' => 'Г', '/g/u' => 'г',
 124                         '/Ğ/u' => 'Ғ', '/ğ/u' => 'ғ', '/H/u' => 'Һ', '/h/u' => 'һ',
 125                         '/I/u' => 'Ы', '/ı/u' => 'ы', '/İ/u' => 'І', '/i/u' => 'і',
 126                         '/Ï/u' => 'И', '/ï/u' => 'и', '/J/u' => 'Ж', '/j/u' => 'ж',
 127                         '/K/u' => 'К', '/k/u' => 'к', '/L/u' => 'Л', '/l/u' => 'л',
 128                         '/M/u' => 'М', '/m/u' => 'м', '/N/u' => 'Н', '/n/u' => 'н',
 129                         '/Ñ/u' => 'Ң', '/ñ/u' => 'ң', '/O/u' => 'О', '/o/u' => 'о',
 130                         '/Ö/u' => 'Ө', '/ö/u' => 'ө', '/P/u' => 'П', '/p/u' => 'п',
 131                         '/Q/u' => 'Қ', '/q/u' => 'қ', '/R/u' => 'Р', '/r/u' => 'р',
 132                         '/S/u' => 'С', '/s/u' => 'с', '/Ş/u' => 'Ш', '/ş/u' => 'ш',
 133                         '/T/u' => 'Т', '/t/u' => 'т', '/U/u' => 'Ұ', '/u/u' => 'ұ',
 134                         '/Ü/u' => 'Ү', '/ü/u' => 'ү', '/V/u' => 'В', '/v/u' => 'в',
 135                         '/W/u' => 'У', '/w/u' => 'у', '/Ý/u' => 'Й', '/ý/u' => 'й',
 136                         '/X/u' => 'Х', '/x/u' => 'х', '/Z/u' => 'З', '/z/u' => 'з',
 137                 );
 138
 139                 $this->mCyLa2Arab = array(
 140                         ## Cyrillic -> Arabic
 141                         '/\№/u' => 'نٶ.',
 142                         '/([АӘЕЁИОӨҰҮЭЮЯЪЬ])е/ui' => '$1يە',
 143                         '/а/ui' => 'ا', '/ә/ui' => 'ٵ', '/б/ui' => 'ب', '/в/ui' => 'ۆ',
 144                         '/г/ui' => 'گ', '/ғ/ui' => 'ع', '/д/ui' => 'د', '/[еэ]/ui' => 'ە',
 145                         '/ё/ui' => 'يو', '/ж/ui' => 'ج', '/з/ui' => 'ز', '/и/ui' => 'ي',
 146                         '/й/ui' => 'ي', '/к/ui' => 'ك', '/қ/ui' => 'ق', '/л/ui' => 'ل',
 147                         '/м/ui' => 'م', '/н/ui' => 'ن', '/ң/ui' => 'ڭ', '/о/ui' => 'و',
 148                         '/ө/ui' => 'ٶ', '/п/ui' => 'پ', '/р/ui' => 'ر', '/с/ui' => 'س',
 149                         '/т/ui' => 'ت', '/у/ui' => 'ۋ', '/ұ/ui' => 'ۇ', '/ү/ui' => 'ٷ',
 150                         '/ф/ui' => 'ف', '/х/ui' => 'ح', '/һ/ui' => 'ھ', '/ц/ui' => 'تس',
 151                         '/ч/ui' => 'چ', '/ш/ui' => 'ش', '/щ/ui' => 'شش', '/[ъь]/ui' => '',
 152                         '/ы/ui' => 'ى', '/і/ui' => 'ٸ', '/ю/ui' => 'يۋ', '/я/ui' => 'يا',
 153
 154                         ## Latin -> Arabic // commented for now...
 155                         /*'/No\./u' => 'نٶ.',
 156                         '/a/ui' => 'ا',  '/ä/ui' => 'ٵ',  '/b/ui' => 'ب', '/c/ui' => 'تس',
 157                         '/ç/ui' => 'چ', '/d/ui' => 'د', '/[eé]/ui' => 'ە', '/f/ui' => 'ف',
 158                         '/g/ui' => 'گ',  '/ğ/ui' => 'ع', '/h/ui' => 'ھ', '/[ıI]/u' => 'ى',
 159                         '/[iİ]/u' => 'ٸ', '/ï/ui' => 'ي', '/j/ui' => 'ج', '/k/ui' => 'ك',
 160                         '/l/ui' => 'ل', '/m/ui' => 'م', '/n/ui' => 'ن', '/ñ/ui' => 'ڭ',
 161                         '/o/ui' => 'و', '/ö/ui' => 'ٶ', '/p/ui' => 'پ', '/q/ui' => 'ق',
 162                         '/r/ui' => 'ر', '/s/ui' => 'س', '/ş/ui' => 'ش', '/t/ui' => 'ت',
 163                         '/u/ui' => 'ۇ', '/ü/ui' => 'ٷ', '/v/ui' => 'ۆ', '/w/ui' => 'ۋ',
 164                         '/x/ui' => 'ح', '/[yý]/ui' => 'ي', '/z/ui' => 'ز', '/[ʺʹ]/ui' => '',*/
 165
 166                         ## Punctuation -> Arabic
 167                         '/\?/' => '؟', # &#x061F;
 168                         '/\,/' => '،', # &#x060C;
 169                         '/\;/' => '؛' , # &#x061B;
 170                         '/\%/' => '٪', # &#x066a;
 171                         ## Digits -> Arabic
 172                         '/0/' => '۰',  # &#x06f0;
 173                         '/1/' => '۱', # &#x06f1;
 174                         '/2/' => '۲', # &#x06f2;
 175                         '/3/' => '۳', # &#x06f3;
 176                         '/4/' => '۴', # &#x06f4;
 177                         '/5/' => '۵', # &#x06f5;
 178                         '/6/' => '۶', # &#x06f6;
 179                         '/7/' => '۷', # &#x06f7;
 180                         '/8/' => '۸', # &#x06f8;
 181                         '/9/' => '۹', # &#x06f9;
 182                 );
 183
 184         }
 185
 186         /* rules should be defined as -{ekavian | iyekavian-} -or-
 187                 -{code:text | code:text | ...}-
 188                 update: delete all rule parsing because it's not used
 189                                 currently, and just produces a couple of bugs
 190         */
 191         function parseManualRule($rule, $flags=array()) {
 192                 if(in_array('T',$flags)){
 193                         return parent::parseManualRule($rule, $flags);
 194                 }
 195
 196                 // otherwise ignore all formatting
 197                 foreach($this->mVariants as $v) {
 198                         $carray[$v] = $rule;
 199                 }
 200
 201                 return $carray;
 202         }
 203
 204         // Do not convert content on talk pages
 205         function parserConvert( $text, &$parser ){
 206                 if(is_object($parser->getTitle() ) && $parser->getTitle()->isTalkPage())
 207                         $this->mDoContentConvert=false;
 208                 else
 209                         $this->mDoContentConvert=true;
 210
 211                 return parent::parserConvert($text, $parser );
 212         }
 213
 214         /*
 215          * A function wrapper:
 216          *  - if there is no selected variant, leave the link
 217          *    names as they were
 218          *  - do not try to find variants for usernames
 219          */
 220         function findVariantLink( &$link, &$nt ) {
 221                 // check for user namespace
 222                 if(is_object($nt)){
 223                         $ns = $nt->getNamespace();
 224                         if($ns==NS_USER || $ns==NS_USER_TALK)
 225                                 return;
 226                 }
 227
 228                 $oldlink=$link;
 229                 parent::findVariantLink($link,$nt);
 230                 if($this->getPreferredVariant()==$this->mMainLanguageCode)
 231                         $link=$oldlink;
 232         }
 233
 234         /*
 235          * An ugly function wrapper for parsing Image titles
 236          * (to prevent image name conversion)
 237          */
 238         function autoConvert($text, $toVariant=false) {
 239                 global $wgTitle;
 240                 if(is_object($wgTitle) && $wgTitle->getNameSpace()==NS_IMAGE){
 241                         $imagename = $wgTitle->getNsText();
 242                         if(preg_match("/^$imagename:/",$text)) return $text;
 243                 }
 244                 return parent::autoConvert($text,$toVariant);
 245         }
 246
 247         /**
 248          *  It translates text into variant
 249          */
 250         function translate( $text, $toVariant ){
 251                 global $wgContLanguageCode;
 252                 $text = parent::translate( $text, $toVariant );
 253
 254                 $letters = '';
 255                 switch( $toVariant ) {
 256                         case 'kk-cyrl':
 257                         case 'kk-kz':
 258                                 $letters = KK_L_UC . KK_L_LC . 'ʺʹ0123456789';
 259                                 $wgContLanguageCode = 'kk';
 260                                 break;
 261                         case 'kk-latn':
 262                         case 'kk-tr':
 263                                 $letters = KK_C_UC . KK_C_LC . '№0123456789';
 264                                 $wgContLanguageCode = 'kk-Latn';
 265                                 break;
 266                         case 'kk-arab':
 267                         case 'kk-cn':
 268                                 // $letters = KK_C_UC.KK_C_LC.KK_L_UC.KK_L_LC.'ʺʹ%№0123456789?,;';
 269                                 $letters = KK_C_UC . KK_C_LC . '%№0123456789?,;';
 270                                 $wgContLanguageCode = 'kk-Arab';
 271                                 break;
 272                         default:
 273                                 $wgContLanguageCode = 'kk';
 274                                 return $text;
 275                 }
 276                 // disable conversion variables like $1, $2...
 277                 $varsfix = '\$[0-9]';
 278
 279                 $matches = preg_split( '/' . $varsfix . '[^' . $letters . ']+/u', $text, -1, PREG_SPLIT_OFFSET_CAPTURE);
 280                 $mstart = 0;
 281                 $ret = '';
 282                 foreach( $matches as $m ) {
 283                         $ret .= substr( $text, $mstart, $m[1]-$mstart );
 284                         $ret .= $this->regsConverter( $m[0], $toVariant );
 285                         $mstart = $m[1] + strlen($m[0]);
 286                 }
 287                 return $ret;
 288         }
 289
 290         function regsConverter( $text, $toVariant ) {
 291                 if ($text == '') return $text;
 292
 293                 $pat = array();
 294                 $rep = array();
 295                 switch( $toVariant ) {
 296                         case 'kk-arab':
 297                         case 'kk-cn':
 298                                 foreach( $this->mCyLa2Arab as $pat => $rep ) {
 299                                         $text = preg_replace( $pat, $rep, $text );
 300                                 }
 301                                 return $text;
 302                                 break;
 303                         case 'kk-latn':
 304                         case 'kk-tr':
 305                                 foreach( $this->mCyrl2Latn as $pat => $rep ) {
 306                                         $text = preg_replace( $pat, $rep, $text );
 307                                 }
 308                                 return $text;
 309                                 break;
 310                         case 'kk-cyrl':
 311                         case 'kk-cn':
 312                                 foreach( $this->mLatn2Cyrl as $pat => $rep ) {
 313                                         $text = preg_replace( $pat, $rep, $text );
 314                                 }
 315                                 return $text;
 316                                 break;
 317                         default:
 318                                 return $text;
 319                 }
 320         }
 321
 322         /*
 323          * We want our external link captions to be converted in variants,
 324          * so we return the original text instead -{$text}-, except for URLs
 325          */
 326         function markNoConversion( $text, $noParse=false ) {
 327                 if( $noParse || preg_match( "/^https?:\/\/|ftp:\/\/|irc:\/\//", $text ) )
 328                         return parent::markNoConversion( $text );
 329                 return $text;
 330         }
 331
 332         function convertCategoryKey( $key ) {
 333                 return $this->autoConvert( $key, 'kk' );
 334         }
 335
 336 }
 337
 338 /* class that handles Cyrillic, Latin and Arabic scripts for Kazakh
 339    right now it only distinguish kk_cyrl, kk_latn, kk_arab and kk_kz, kk_tr, kk_cn.
 340 */
 341 class LanguageKk extends LanguageKk_cyrl {
 342
 343         function __construct() {
 344                 global $wgHooks;
 345                 parent::__construct();
 346
 347                 $variants = array( 'kk', 'kk-cyrl', 'kk-latn', 'kk-arab', 'kk-kz', 'kk-tr', 'kk-cn' );
 348                 $variantfallbacks = array(
 349                         'kk'      => 'kk-kz',
 350                         'kk-cyrl' => 'kk',
 351                         'kk-latn' => 'kk',
 352                         'kk-arab' => 'kk',
 353                         'kk-kz'   => 'kk-cyrl',
 354                         'kk-tr'   => 'kk-latn',
 355                         'kk-cn'   => 'kk-arab'
 356                 );
 357
 358                 $this->mConverter = new KkConverter( $this, 'kk', $variants, $variantfallbacks );
 359
 360                 $wgHooks['ArticleSaveComplete'][] = $this->mConverter;
 361         }
 362
 363         /**
 364          * Work around for right-to-left direction support in kk-arab and kk-cn
 365          *
 366          * @return bool
 367          */
 368         function isRTL() {
 369                 $variant = $this->getPreferredVariant();
 370                 if ( $variant == 'kk-arab' || $variant == 'kk-cn' ) {
 371                         return true;
 372                 } else {
 373                         return parent::isRTL();
 374                 }
 375         }
 376
 377         /*
 378          * It fixes issue with ucfirst for transforming 'i' to 'İ'
 379          *
 380          */
 381         function ucfirst ( $string ) {
 382                 $variant = $this->getPreferredVariant();
 383                 if ( ($variant == 'kk-latn' || $variant == 'kk-tr') && $string[0] == 'i' ) {
 384                         $string = 'İ' . substr( $string, 1 );
 385                 } else {
 386                         $string = parent::ucfirst( $string );
 387                 }
 388                 return $string;
 389         }
 390
 391         /*
 392          * It fixes issue with  lcfirst for transforming 'I' to 'ı'
 393          *
 394          */
 395         function lcfirst ( $string ) {
 396                 $variant = $this->getPreferredVariant();
 397                 if ( ($variant == 'kk-latn' || $variant == 'kk-tr') && $string[0] == 'I' ) {
 398                         $string = 'ı' . substr( $string, 1 );
 399                 } else {
 400                         $string = parent::lcfirst( $string );
 401                 }
 402                 return $string;
 403         }
 404
 405         function convertGrammar( $word, $case ) {
 406                 $fname="LanguageKk::convertGrammar";
 407                 wfProfileIn( $fname );
 408
 409                 $variant = $this->getPreferredVariant();
 410                 switch ( $variant ) {
 411                         case 'kk-arab':
 412                         case 'kk-cn':
 413                                 $word = parent::convertGrammar( $word, $case, $variant = 'kk-arab' );
 414                                 break;
 415                         case 'kk-latn':
 416                         case 'kk-tr':
 417                                 $word = parent::convertGrammar( $word, $case, $variant = 'kk-latn' );
 418                                 break;
 419                         case 'kk-cyrl':
 420                         case 'kk-kz':
 421                         case 'kk':
 422                         default:
 423                                 $word = parent::convertGrammar( $word, $case, $variant = 'kk-cyrl' );
 424                 }
 425
 426                 wfProfileOut( $fname );
 427                 return $word;
 428         }
 429
 430 }