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 require_once __DIR__ . '/../LanguageConverter.php';
  25 require_once __DIR__ . '/LanguageZh_hans.php';
  26
  27 /**
  28  * @ingroup Language
  29  */
  30 class ZhConverter extends LanguageConverter {
  31         /**
  32          * @param Language $langobj
  33          * @param string $maincode
  34          * @param array $variants
  35          * @param array $variantfallbacks
  36          * @param array $flags
  37          * @param array $manualLevel
  38          */
  39         function __construct( $langobj, $maincode,
  40                                                                 $variants = array(),
  41                                                                 $variantfallbacks = array(),
  42                                                                 $flags = array(),
  43                                                                 $manualLevel = array() ) {
  44                 $this->mDescCodeSep = '：';
  45                 $this->mDescVarSep = '；';
  46                 parent::__construct( $langobj, $maincode,
  47                                                                         $variants,
  48                                                                         $variantfallbacks,
  49                                                                         $flags,
  50                                                                         $manualLevel );
  51                 $names = array(
  52                         'zh' => '原文',
  53                         'zh-hans' => '简体',
  54                         'zh-hant' => '繁體',
  55                         'zh-cn' => '大陆',
  56                         'zh-tw' => '台灣',
  57                         'zh-hk' => '香港',
  58                         'zh-mo' => '澳門',
  59                         'zh-sg' => '新加坡',
  60                         'zh-my' => '大马',
  61                 );
  62                 $this->mVariantNames = array_merge( $this->mVariantNames, $names );
  63         }
  64
  65         function loadDefaultTables() {
  66                 require __DIR__ . "/../../includes/ZhConversion.php";
  67                 $this->mTables = array(
  68                         'zh-hans' => new ReplacementArray( $zh2Hans ),
  69                         'zh-hant' => new ReplacementArray( $zh2Hant ),
  70                         'zh-cn' => new ReplacementArray( array_merge( $zh2Hans, $zh2CN ) ),
  71                         'zh-hk' => new ReplacementArray( array_merge( $zh2Hant, $zh2HK ) ),
  72                         'zh-mo' => new ReplacementArray( array_merge( $zh2Hant, $zh2HK ) ),
  73                         'zh-my' => new ReplacementArray( array_merge( $zh2Hans, $zh2SG ) ),
  74                         'zh-sg' => new ReplacementArray( array_merge( $zh2Hans, $zh2SG ) ),
  75                         'zh-tw' => new ReplacementArray( array_merge( $zh2Hant, $zh2TW ) ),
  76                         'zh' => new ReplacementArray
  77                 );
  78         }
  79
  80         function postLoadTables() {
  81                 $this->mTables['zh-cn']->merge( $this->mTables['zh-hans'] );
  82                 $this->mTables['zh-hk']->merge( $this->mTables['zh-hant'] );
  83                 $this->mTables['zh-mo']->merge( $this->mTables['zh-hant'] );
  84                 $this->mTables['zh-my']->merge( $this->mTables['zh-hans'] );
  85                 $this->mTables['zh-sg']->merge( $this->mTables['zh-hans'] );
  86                 $this->mTables['zh-tw']->merge( $this->mTables['zh-hant'] );
  87         }
  88
  89         /**
  90          * @param string $key
  91          * @return string
  92          */
  93         function convertCategoryKey( $key ) {
  94                 return $this->autoConvert( $key, 'zh' );
  95         }
  96 }
  97
  98 /**
  99  * class that handles both Traditional and Simplified Chinese
 100  * right now it only distinguish zh_hans, zh_hant, zh_cn, zh_tw, zh_sg and zh_hk.
 101  *
 102  * @ingroup Language
 103  */
 104 class LanguageZh extends LanguageZh_hans {
 105         function __construct() {
 106                 global $wgHooks;
 107
 108                 parent::__construct();
 109
 110                 $variants = array(
 111                         'zh',
 112                         'zh-hans',
 113                         'zh-hant',
 114                         'zh-cn',
 115                         'zh-hk',
 116                         'zh-mo',
 117                         'zh-my',
 118                         'zh-sg',
 119                         'zh-tw'
 120                 );
 121
 122                 $variantfallbacks = array(
 123                         'zh' => array( 'zh-hans', 'zh-hant', 'zh-cn', 'zh-tw', 'zh-hk', 'zh-sg', 'zh-mo', 'zh-my' ),
 124                         'zh-hans' => array( 'zh-cn', 'zh-sg', 'zh-my' ),
 125                         'zh-hant' => array( 'zh-tw', 'zh-hk', 'zh-mo' ),
 126                         'zh-cn' => array( 'zh-hans', 'zh-sg', 'zh-my' ),
 127                         'zh-sg' => array( 'zh-hans', 'zh-cn', 'zh-my' ),
 128                         'zh-my' => array( 'zh-hans', 'zh-sg', 'zh-cn' ),
 129                         'zh-tw' => array( 'zh-hant', 'zh-hk', 'zh-mo' ),
 130                         'zh-hk' => array( 'zh-hant', 'zh-mo', 'zh-tw' ),
 131                         'zh-mo' => array( 'zh-hant', 'zh-hk', 'zh-tw' ),
 132                 );
 133                 $ml = array(
 134                         'zh' => 'disable',
 135                         'zh-hans' => 'unidirectional',
 136                         'zh-hant' => 'unidirectional',
 137                 );
 138
 139                 $this->mConverter = new ZhConverter( $this, 'zh',
 140                                                                 $variants, $variantfallbacks,
 141                                                                 array(),
 142                                                                 $ml );
 143
 144                 $wgHooks['PageContentSaveComplete'][] = $this->mConverter;
 145         }
 146
 147         /**
 148          * this should give much better diff info
 149          *
 150          * @param string $text
 151          * @return string
 152          */
 153         function segmentForDiff( $text ) {
 154                 return preg_replace( '/[\xc0-\xff][\x80-\xbf]*/', ' $0', $text );
 155         }
 156
 157         /**
 158          * @param string $text
 159          * @return string
 160          */
 161         function unsegmentForDiff( $text ) {
 162                 return preg_replace( '/ ([\xc0-\xff][\x80-\xbf]*)/', '$1', $text );
 163         }
 164
 165         /**
 166          * auto convert to zh-hans and normalize special characters.
 167          *
 168          * @param string $string
 169          * @param string $autoVariant default to 'zh-hans'
 170          * @return string
 171          */
 172         function normalizeForSearch( $string, $autoVariant = 'zh-hans' ) {
 173                 wfProfileIn( __METHOD__ );
 174
 175                 // always convert to zh-hans before indexing. it should be
 176                 // better to use zh-hans for search, since conversion from
 177                 // Traditional to Simplified is less ambiguous than the
 178                 // other way around
 179                 $s = $this->mConverter->autoConvert( $string, $autoVariant );
 180                 // LanguageZh_hans::normalizeForSearch
 181                 $s = parent::normalizeForSearch( $s );
 182                 wfProfileOut( __METHOD__ );
 183                 return $s;
 184
 185         }
 186
 187         /**
 188          * @param array $termsArray
 189          * @return array
 190          */
 191         function convertForSearchResult( $termsArray ) {
 192                 $terms = implode( '|', $termsArray );
 193                 $terms = self::convertDoubleWidth( $terms );
 194                 $terms = implode( '|', $this->mConverter->autoConvertToAllVariants( $terms ) );
 195                 $ret = array_unique( explode( '|', $terms ) );
 196                 return $ret;
 197         }
 198 }