<?php
-require_once( "LanguageUtf8.php" );
+/**
+ * @package MediaWiki
+ * @subpackage Language
+ */
+require_once( "LanguageConverter.php" );
require_once( "LanguageZh_cn.php");
require_once( "LanguageZh_tw.php");
-require_once( "ZhConversion.php");
-
-/* class that handles both Traditional and Simplified Chinese
- right now it only distinguish zh_cn and zh_tw (actuall, zh_cn and
- non-zh_cn), will add support for zh_sg, zh_hk, etc, later.
-*/
-class LanguageZh extends LanguageUtf8 {
-
- var $mZhLang=false, $mZhLanguageCode=false;
-
- function LanguageZh() {
- $this->mZhLanguageCode = $this->getPreferredLanguage();
- if($this->mZhLanguageCode == "cn") {
- $this->mZhLang = new LanguageZh_cn();
- }
- else {
- $this->mZhLang = new LanguageZh_tw();
- }
+require_once( "LanguageZh_sg.php");
+require_once( "LanguageZh_hk.php");
+
+class ZhConverter extends LanguageConverter {
+ function loadDefaultTables() {
+ require( "includes/ZhConversion.php" );
+ $this->mTables = array();
+ $this->mTables['zh-cn'] = $zh2CN;
+ $this->mTables['zh-tw'] = $zh2TW;
+ $this->mTables['zh-sg'] = array_merge($zh2CN, $zh2SG);
+ $this->mTables['zh-hk'] = array_merge($zh2TW, $zh2HK);
+ $this->mTables['zh'] = array();
+ }
+
+ function postLoadTables() {
+ $this->mTables['zh-sg'] = array_merge($this->mTables['zh-cn'], $this->mTables['zh-sg']);
+ $this->mTables['zh-hk'] = array_merge($this->mTables['zh-tw'], $this->mTables['zh-hk']);
}
- /*
- get preferred language variants. eventually this will check the
- user's preference setting as well, once the language option in
- the setting pages is finalized.
+ /* there shouldn't be any latin text in Chinese conversion, so no need
+ to mark anything
*/
- function getPreferredLanguage() {
- if($this->mZhLanguageCode)
- return $this->mZhLanguageCode;
-
- $this->mZhLanguageCode="cn";
- $value = $_SERVER["HTTP_ACCEPT_LANGUAGE"];
- $zh = explode("zh-", $value);
- array_shift($zh);
- $l = array_shift($zh);
- if($l != NULL) {
- $this->mZhLanguageCode = strtolower(substr($l,0,2));
- }
-
- return $this->mZhLanguageCode;
- }
-
-
- /* the Simplified/Traditional conversion stuff */
-
- function simp2trad($text) {
- global $wgZhSimp2Trad;
- return strtr($text, $wgZhSimp2Trad);
- }
-
- function trad2simp($text) {
- global $wgZhTrad2Simp;
- return strtr($text, $wgZhTrad2Simp);
- }
-
- function convert($text) {
-
- // no conversion if redirecting
- if(substr($text,0,9) == "#REDIRECT") {
- return $text;
- }
- // determine the preferred language from the request header
- $tolang = $this->getPreferredLanguage();
-
- $ltext = explode("-{", $text);
- $lfirst = array_shift($ltext);
-
- if($tolang == "cn") {
- $text = $this->trad2simp($lfirst);
- }
- else {
- $text = $this->simp2trad($lfirst);
- }
-
- foreach ($ltext as $i => $txt) {
- $a = explode("}-", $txt);
- $b = explode("zh-", $a{0});
- if($b{1}==NULL) {
- $text = $text.$b{0};
- }
- else {
- foreach ($b as $j => $lang) {
- if(substr($lang,0,2) == $tolang) {
- $text = $text.substr($lang, 2);
- break;
- }
- }
- }
- if($tolang == "cn") {
- $text = $text.$this->trad2simp($a{1});
- }
- else {
- $text = $text.$this->simp2trad($a{1});
- }
- }
+ function markNoConversion($text) {
return $text;
}
-
-
-
-
- /* these just calls the method of the corresponding class */
-
- function getDefaultUserOptions () {
- return $this->mZhLang->getDefaultUserOptions();
- }
-
- function getBookstoreList () {
- return $this->mZhLang->getBookstoreList() ;
- }
- function getNamespaces() {
- return $this->mZhLang->getNamespaces();
+ function convertCategoryKey( $key ) {
+ return $this->autoConvert( $key, 'zh-cn' );
}
+}
- function getNsText( $index ) {
- return $this->mZhLang->getNsText($index);
- }
- function getNsIndex( $text ) {
- return $this->mZhLang->getNsIndex($text);
- }
-
- function getQuickbarSettings() {
- return $this->mZhLang->getQuickbarSettings();
- }
-
- function getSkinNames() {
- return $this->mZhLang->getSkinNames();
+/* class that handles both Traditional and Simplified Chinese
+ right now it only distinguish zh_cn, zh_tw, zh_sg and zh_hk.
+*/
+class LanguageZh extends LanguageZh_cn {
+
+ function LanguageZh() {
+ global $wgHooks;
+ $this->mConverter = new ZhConverter($this, 'zh',
+ array('zh', 'zh-cn', 'zh-tw', 'zh-sg', 'zh-hk'),
+ array('zh'=>'zh-cn',
+ 'zh-cn'=>'zh-sg',
+ 'zh-sg'=>'zh-cn',
+ 'zh-tw'=>'zh-hk',
+ 'zh-hk'=>'zh-tw'));
+ $wgHooks['ArticleSaveComplete'][] = $this->mConverter;
}
- function date( $ts, $adj = false )
- {
- return $this->mZhLang->date($ts,$adj);
+
+ # this should give much better diff info
+ function segmentForDiff( $text ) {
+ return preg_replace(
+ "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
+ "' ' .\"$1\"", $text);
}
- function timeanddate( $ts, $adj = false )
- {
- return $this->mZhLang->timeanddate($ts, $adj);
+ function unsegmentForDiff( $text ) {
+ return preg_replace(
+ "/ ([\\xc0-\\xff][\\x80-\\xbf]*)/e",
+ "\"$1\"", $text);
}
- function getValidSpecialPages()
- {
- return $this->mZhLang->getValidSpecialPages();
- }
+ // word segmentation
+ function stripForSearch( $string ) {
+ $fname="LanguageZh::stripForSearch";
+ wfProfileIn( $fname );
- function getSysopSpecialPages()
- {
- return $this->mZhLang->getSysopSpecialPages();
- }
+ // eventually this should be a word segmentation
+ // for now just treat each character as a word
+ $t = preg_replace(
+ "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
+ "' ' .\"$1\"", $string);
- function getDeveloperSpecialPages()
- {
- return $this->mZhLang->getDeveloperSpecialPages();
+ //always convert to zh-cn before indexing. it should be
+ //better to use zh-cn for search, since conversion from
+ //Traditional to Simplified is less ambiguous than the
+ //other way around
- }
+ $t = $this->mConverter->autoConvert($t, 'zh-cn');
+ $t = LanguageUtf8::stripForSearch( $t );
+ wfProfileOut( $fname );
+ return $t;
- function getMessage( $key )
- {
- return $this->mZhLang->getMessage($key);
}
- function stripForSearch( $string ) {
- return $this->mZhLang->stripForSearch($string);
+ function convertForSearchResult( $termsArray ) {
+ $terms = implode( '|', $termsArray );
+ $terms = implode( '|', $this->mConverter->autoConvertToAllVariants( $terms ) );
+ $ret = array_unique( explode('|', $terms) );
+ return $ret;
}
-
}
-
-
?>