/**
* 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 $s;
}
function stripForSearch($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-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
$t = $this->mConverter->autoConvert($t, 'zh-hans');
$t = parent::stripForSearch($t);
wfProfileOut($fname);
return $t;
}
function stripForSearch($string)
{
wfProfileIn(__METHOD__);
// eventually this should be a word segmentation
// for now just treat each character as a word
// @fixme only do this for Han characters...
$t = preg_replace("/([\\xc0-\\xff][\\x80-\\xbf]*)/", " \$1", $string);
//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
$t = $this->mConverter->autoConvert($t, 'zh-hans');
$t = parent::stripForSearch($t);
wfProfileOut(__METHOD__);
return $t;
}