X-Git-Url: https://scripts.mit.edu/gitweb/autoinstallsdev/mediawiki.git/blobdiff_plain/d7967d5e4460e08b6b258307afbca0596b18a3dd..18a6620945d02687fbcfc4c27355d952fd748b41:/languages/classes/LanguageZh.php

diff --git a/languages/classes/LanguageZh.php b/languages/classes/LanguageZh.php
index bcdf7dd8..0055a33b 100644
--- a/languages/classes/LanguageZh.php
+++ b/languages/classes/LanguageZh.php
@@ -1,102 +1,156 @@
 <?php
-/**
-  * @addtogroup Language
-  */
-require_once( dirname(__FILE__).'/../LanguageConverter.php' );
-require_once( dirname(__FILE__).'/LanguageZh_cn.php' );
 
+require_once( dirname( __FILE__ ) . '/../LanguageConverter.php' );
+require_once( dirname( __FILE__ ) . '/LanguageZh_hans.php' );
+
+/**
+ * @ingroup Language
+ */
 class ZhConverter extends LanguageConverter {
+
+	function __construct( $langobj, $maincode,
+								$variants = array(),
+								$variantfallbacks = array(),
+								$flags = array(),
+								$manualLevel = array() ) {
+		$this->mDescCodeSep = '：';
+		$this->mDescVarSep = '；';
+		parent::__construct( $langobj, $maincode,
+									$variants,
+									$variantfallbacks,
+									$flags,
+									$manualLevel );
+		$names = array(
+			'zh'      => '原文',
+			'zh-hans' => '简体',
+			'zh-hant' => '繁體',
+			'zh-cn'   => '大陆',
+			'zh-tw'   => '台灣',
+			'zh-hk'   => '香港',
+			'zh-mo'   => '澳門',
+			'zh-sg'   => '新加坡',
+			'zh-my'   => '大马',
+		);
+		$this->mVariantNames = array_merge( $this->mVariantNames, $names );
+	}
+
 	function loadDefaultTables() {
-		require( dirname(__FILE__)."/../../includes/ZhConversion.php" );
+		require( dirname( __FILE__ ) . "/../../includes/ZhConversion.php" );
 		$this->mTables = array(
-			'zh-cn' => new ReplacementArray( $zh2CN ),
-			'zh-tw' => new ReplacementArray( $zh2TW ),
-			'zh-sg' => new ReplacementArray( array_merge($zh2CN, $zh2SG) ),
-			'zh-hk' => new ReplacementArray( array_merge($zh2TW, $zh2HK) ),
-			'zh' => new ReplacementArray
+			'zh-hans' => new ReplacementArray( $zh2Hans ),
+			'zh-hant' => new ReplacementArray( $zh2Hant ),
+			'zh-cn'   => new ReplacementArray( array_merge( $zh2Hans, $zh2CN ) ),
+			'zh-hk'   => new ReplacementArray( array_merge( $zh2Hant, $zh2HK ) ),
+			'zh-mo'   => new ReplacementArray( array_merge( $zh2Hant, $zh2HK ) ),
+			'zh-my'   => new ReplacementArray( array_merge( $zh2Hans, $zh2SG ) ),
+			'zh-sg'   => new ReplacementArray( array_merge( $zh2Hans, $zh2SG ) ),
+			'zh-tw'   => new ReplacementArray( array_merge( $zh2Hant, $zh2TW ) ),
+			'zh'      => new ReplacementArray
 		);
 	}
 
 	function postLoadTables() {
-		$this->mTables['zh-sg']->merge( $this->mTables['zh-cn'] );
-		$this->mTables['zh-hk']->merge( $this->mTables['zh-tw'] );
-    }
+		$this->mTables['zh-cn']->merge( $this->mTables['zh-hans'] );
+		$this->mTables['zh-hk']->merge( $this->mTables['zh-hant'] );
+		$this->mTables['zh-mo']->merge( $this->mTables['zh-hant'] );
+		$this->mTables['zh-my']->merge( $this->mTables['zh-hans'] );
+		$this->mTables['zh-sg']->merge( $this->mTables['zh-hans'] );
+		$this->mTables['zh-tw']->merge( $this->mTables['zh-hant'] );
+	}
 
 	/* there shouldn't be any latin text in Chinese conversion, so no need
 	   to mark anything.
 	   $noParse is there for compatibility with LanguageConvert::markNoConversion
-    */
-	function markNoConversion($text, $noParse = false) {
+	 */
+	function markNoConversion( $text, $noParse = false ) {
 		return $text;
 	}
 
 	function convertCategoryKey( $key ) {
-		return $this->autoConvert( $key, 'zh-cn' );
+		return $this->autoConvert( $key, 'zh' );
 	}
 }
 
-
-/* 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 {
+/**
+ * class that handles both Traditional and Simplified Chinese
+ * right now it only distinguish zh_hans, zh_hant, zh_cn, zh_tw, zh_sg and zh_hk.
+ *
+ * @ingroup Language
+ */
+class LanguageZh extends LanguageZh_hans {
 
 	function __construct() {
 		global $wgHooks;
 		parent::__construct();
-		$this->mConverter = new ZhConverter($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'));
+
+		$variants = array( 'zh', 'zh-hans', 'zh-hant', 'zh-cn', 'zh-hk', 'zh-mo', 'zh-my', 'zh-sg', 'zh-tw' );
+
+		$variantfallbacks = array(
+			'zh'      => array( 'zh-hans', 'zh-hant', 'zh-cn', 'zh-tw', 'zh-hk', 'zh-sg', 'zh-mo', 'zh-my' ),
+			'zh-hans' => array( 'zh-cn', 'zh-sg', 'zh-my' ),
+			'zh-hant' => array( 'zh-tw', 'zh-hk', 'zh-mo' ),
+			'zh-cn'   => array( 'zh-hans', 'zh-sg', 'zh-my' ),
+			'zh-sg'   => array( 'zh-hans', 'zh-cn', 'zh-my' ),
+			'zh-my'   => array( 'zh-hans', 'zh-sg', 'zh-cn' ),
+			'zh-tw'   => array( 'zh-hant', 'zh-hk', 'zh-mo' ),
+			'zh-hk'   => array( 'zh-hant', 'zh-mo', 'zh-tw' ),
+			'zh-mo'   => array( 'zh-hant', 'zh-hk', 'zh-tw' ),
+		);
+		$ml = array(
+			'zh'      => 'disable',
+			'zh-hans' => 'unidirectional',
+			'zh-hant' => 'unidirectional',
+		);
+
+		$this->mConverter = new ZhConverter( $this, 'zh',
+								$variants, $variantfallbacks,
+								array(),
+								$ml );
+
 		$wgHooks['ArticleSaveComplete'][] = $this->mConverter;
 	}
 
-
 	# this should give much better diff info
 	function segmentForDiff( $text ) {
 		return preg_replace(
 			"/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
-			"' ' .\"$1\"", $text);
+			"' ' .\"$1\"", $text );
 	}
 
 	function unsegmentForDiff( $text ) {
 		return preg_replace(
 			"/ ([\\xc0-\\xff][\\x80-\\xbf]*)/e",
-			"\"$1\"", $text);
+			"\"$1\"", $text );
 	}
 
-	// word segmentation
-	function stripForSearch( $string ) {
-		$fname="LanguageZh::stripForSearch";
-		wfProfileIn( $fname );
-
-		// 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);
-
-        //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 = parent::stripForSearch( $t );
-		wfProfileOut( $fname );
-		return $t;
+	/**
+	 * auto convert to zh-hans and normalize special characters.
+	 *
+	 * @param $string String
+	 * @param $autoVariant String, default to 'zh-hans'
+	 * @return String
+	 */
+	function normalizeForSearch( $string, $autoVariant = 'zh-hans' ) {
+		wfProfileIn( __METHOD__ );
+
+		// always convert to zh-hans before indexing. it should be
+		// better to use zh-hans for search, since conversion from
+		// Traditional to Simplified is less ambiguous than the
+		// other way around
+		$s = $this->mConverter->autoConvert( $string, $autoVariant );
+		// LanguageZh_hans::normalizeForSearch
+		$s = parent::normalizeForSearch( $s );
+		wfProfileOut( __METHOD__ );
+		return $s;
 
 	}
 
 	function convertForSearchResult( $termsArray ) {
 		$terms = implode( '|', $termsArray );
+		$terms = self::convertDoubleWidth( $terms );
 		$terms = implode( '|', $this->mConverter->autoConvertToAllVariants( $terms ) );
-		$ret = array_unique( explode('|', $terms) );
+		$ret = array_unique( explode( '|', $terms ) );
 		return $ret;
 	}
-
 }