summaryrefslogtreecommitdiff
path: root/languages/LanguageZh.php
diff options
context:
space:
mode:
authorPierre Schmitz <pierre@archlinux.de>2006-10-11 18:12:39 +0000
committerPierre Schmitz <pierre@archlinux.de>2006-10-11 18:12:39 +0000
commit183851b06bd6c52f3cae5375f433da720d410447 (patch)
treea477257decbf3360127f6739c2f9d0ec57a03d39 /languages/LanguageZh.php
MediaWiki 1.7.1 wiederhergestellt
Diffstat (limited to 'languages/LanguageZh.php')
-rw-r--r--languages/LanguageZh.php103
1 files changed, 103 insertions, 0 deletions
diff --git a/languages/LanguageZh.php b/languages/LanguageZh.php
new file mode 100644
index 00000000..b24c08e6
--- /dev/null
+++ b/languages/LanguageZh.php
@@ -0,0 +1,103 @@
+<?php
+/**
+ * @package MediaWiki
+ * @subpackage Language
+ */
+require_once( "LanguageConverter.php" );
+require_once( "LanguageZh_cn.php");
+require_once( "LanguageZh_tw.php");
+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'], $this->mTables['zh-hk']);
+ }
+
+ /* there shouldn't be any latin text in Chinese conversion, so no need
+ to mark anything
+ */
+ function markNoConversion($text) {
+ return $text;
+ }
+
+ function convertCategoryKey( $key ) {
+ return $this->autoConvert( $key, 'zh-cn' );
+ }
+}
+
+
+/* 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 __construct() {
+ 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;
+ }
+
+
+ # this should give much better diff info
+ function segmentForDiff( $text ) {
+ return preg_replace(
+ "/([\\xc0-\\xff][\\x80-\\xbf]*)/e",
+ "' ' .\"$1\"", $text);
+ }
+
+ function unsegmentForDiff( $text ) {
+ return preg_replace(
+ "/ ([\\xc0-\\xff][\\x80-\\xbf]*)/e",
+ "\"$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 = LanguageUtf8::stripForSearch( $t );
+ wfProfileOut( $fname );
+ return $t;
+
+ }
+
+ function convertForSearchResult( $termsArray ) {
+ $terms = implode( '|', $termsArray );
+ $terms = implode( '|', $this->mConverter->autoConvertToAllVariants( $terms ) );
+ $ret = array_unique( explode('|', $terms) );
+ return $ret;
+ }
+
+}
+?>