languages/classes/LanguageZh.php

   1 <?php
   2 /**
   3  * Chinese 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  * @ingroup Language
  26  */
  27 class ZhConverter extends LanguageConverter {
  28         /**
  29          * @param Language $langobj
  30          * @param string $maincode
  31          * @param array $variants
  32          * @param array $variantfallbacks
  33          * @param array $flags
  34          * @param array $manualLevel
  35          */
  36         function __construct( $langobj, $maincode,
  37                                                                 $variants = array(),
  38                                                                 $variantfallbacks = array(),
  39                                                                 $flags = array(),
  40                                                                 $manualLevel = array() ) {
  41                 $this->mDescCodeSep = '：';
  42                 $this->mDescVarSep = '；';
  43                 parent::__construct( $langobj, $maincode,
  44                                                                         $variants,
  45                                                                         $variantfallbacks,
  46                                                                         $flags,
  47                                                                         $manualLevel );
  48                 $names = array(
  49                         'zh' => '原文',
  50                         'zh-hans' => '简体',
  51                         'zh-hant' => '繁體',
  52                         'zh-cn' => '大陆',
  53                         'zh-tw' => '台灣',
  54                         'zh-hk' => '香港',
  55                         'zh-mo' => '澳門',
  56                         'zh-sg' => '新加坡',
  57                         'zh-my' => '大马',
  58                 );
  59                 $this->mVariantNames = array_merge( $this->mVariantNames, $names );
  60         }
  61
  62         function loadDefaultTables() {
  63                 require __DIR__ . "/../../includes/ZhConversion.php";
  64                 $this->mTables = array(
  65                         'zh-hans' => new ReplacementArray( $zh2Hans ),
  66                         'zh-hant' => new ReplacementArray( $zh2Hant ),
  67                         'zh-cn' => new ReplacementArray( $zh2CN ),
  68                         'zh-hk' => new ReplacementArray( $zh2HK ),
  69                         'zh-mo' => new ReplacementArray( $zh2HK ),
  70                         'zh-my' => new ReplacementArray( $zh2CN ),
  71                         'zh-sg' => new ReplacementArray( $zh2CN ),
  72                         'zh-tw' => new ReplacementArray( $zh2TW ),
  73                         'zh' => new ReplacementArray
  74                 );
  75         }
  76
  77         function postLoadTables() {
  78                 $this->mTables['zh-cn']->setArray(
  79                         $this->mTables['zh-cn']->getArray() + $this->mTables['zh-hans']->getArray()
  80                 );
  81                 $this->mTables['zh-hk']->setArray(
  82                         $this->mTables['zh-hk']->getArray() + $this->mTables['zh-hant']->getArray()
  83                 );
  84                 $this->mTables['zh-mo']->setArray(
  85                         $this->mTables['zh-mo']->getArray() + $this->mTables['zh-hant']->getArray()
  86                 );
  87                 $this->mTables['zh-my']->setArray(
  88                         $this->mTables['zh-my']->getArray() + $this->mTables['zh-hans']->getArray()
  89                 );
  90                 $this->mTables['zh-sg']->setArray(
  91                         $this->mTables['zh-sg']->getArray() + $this->mTables['zh-hans']->getArray()
  92                 );
  93                 $this->mTables['zh-tw']->setArray(
  94                         $this->mTables['zh-tw']->getArray() + $this->mTables['zh-hant']->getArray()
  95                 );
  96         }
  97
  98         /**
  99          * @param string $key
 100          * @return string
 101          */
 102         function convertCategoryKey( $key ) {
 103                 return $this->autoConvert( $key, 'zh' );
 104         }
 105 }
 106
 107 /**
 108  * class that handles both Traditional and Simplified Chinese
 109  * right now it only distinguish zh_hans, zh_hant, zh_cn, zh_tw, zh_sg and zh_hk.
 110  *
 111  * @ingroup Language
 112  */
 113 class LanguageZh extends LanguageZh_hans {
 114         function __construct() {
 115                 global $wgHooks;
 116
 117                 parent::__construct();
 118
 119                 $variants = array(
 120                         'zh',
 121                         'zh-hans',
 122                         'zh-hant',
 123                         'zh-cn',
 124                         'zh-hk',
 125                         'zh-mo',
 126                         'zh-my',
 127                         'zh-sg',
 128                         'zh-tw'
 129                 );
 130
 131                 $variantfallbacks = array(
 132                         'zh' => array( 'zh-hans', 'zh-hant', 'zh-cn', 'zh-tw', 'zh-hk', 'zh-sg', 'zh-mo', 'zh-my' ),
 133                         'zh-hans' => array( 'zh-cn', 'zh-sg', 'zh-my' ),
 134                         'zh-hant' => array( 'zh-tw', 'zh-hk', 'zh-mo' ),
 135                         'zh-cn' => array( 'zh-hans', 'zh-sg', 'zh-my' ),
 136                         'zh-sg' => array( 'zh-hans', 'zh-cn', 'zh-my' ),
 137                         'zh-my' => array( 'zh-hans', 'zh-sg', 'zh-cn' ),
 138                         'zh-tw' => array( 'zh-hant', 'zh-hk', 'zh-mo' ),
 139                         'zh-hk' => array( 'zh-hant', 'zh-mo', 'zh-tw' ),
 140                         'zh-mo' => array( 'zh-hant', 'zh-hk', 'zh-tw' ),
 141                 );
 142                 $ml = array(
 143                         'zh' => 'disable',
 144                         'zh-hans' => 'unidirectional',
 145                         'zh-hant' => 'unidirectional',
 146                 );
 147
 148                 $this->mConverter = new ZhConverter( $this, 'zh',
 149                                                                 $variants, $variantfallbacks,
 150                                                                 array(),
 151                                                                 $ml );
 152
 153                 $wgHooks['PageContentSaveComplete'][] = $this->mConverter;
 154         }
 155
 156         /**
 157          * this should give much better diff info
 158          *
 159          * @param string $text
 160          * @return string
 161          */
 162         function segmentForDiff( $text ) {
 163                 return preg_replace( '/[\xc0-\xff][\x80-\xbf]*/', ' $0', $text );
 164         }
 165
 166         /**
 167          * @param string $text
 168          * @return string
 169          */
 170         function unsegmentForDiff( $text ) {
 171                 return preg_replace( '/ ([\xc0-\xff][\x80-\xbf]*)/', '$1', $text );
 172         }
 173
 174         /**
 175          * auto convert to zh-hans and normalize special characters.
 176          *
 177          * @param string $string
 178          * @param string $autoVariant Defaults to 'zh-hans'
 179          * @return string
 180          */
 181         function normalizeForSearch( $string, $autoVariant = 'zh-hans' ) {
 182
 183                 // always convert to zh-hans before indexing. it should be
 184                 // better to use zh-hans for search, since conversion from
 185                 // Traditional to Simplified is less ambiguous than the
 186                 // other way around
 187                 $s = $this->mConverter->autoConvert( $string, $autoVariant );
 188                 // LanguageZh_hans::normalizeForSearch
 189                 $s = parent::normalizeForSearch( $s );
 190                 return $s;
 191
 192         }
 193
 194         /**
 195          * @param array $termsArray
 196          * @return array
 197          */
 198         function convertForSearchResult( $termsArray ) {
 199                 $terms = implode( '|', $termsArray );
 200                 $terms = self::convertDoubleWidth( $terms );
 201                 $terms = implode( '|', $this->mConverter->autoConvertToAllVariants( $terms ) );
 202                 $ret = array_unique( explode( '|', $terms ) );
 203                 return $ret;
 204         }
 205 }