Handle unicode strings when normalizing for search (#1487866)

thomascube

2011-04-12 12dac4911b91e80cf1d8c85ee8ad1ef191b630cb

Handle unicode strings when normalizing for search (#1487866)

 program/include/rcube_addressbook.php

@@ -397,14 +397,24 @@
     */
    protected static function normalize_string($str)
    {
        $norm = strtolower(strtr(utf8_decode($str),
            'ÇçäâàåéêëèïîìÅÉöôòüûùÿøØáíóúñÑÁÂÀãÃÊËÈÍÎÏÓÔõÕÚÛÙýÝ',
            'ccaaaaeeeeiiiaeooouuuyooaiounnaaaaaeeeiiioooouuuyy'));

        return preg_replace(
            array('/[\s;\+\-\/]+/i', '/(\d)\s+(\d)/', '/\s\w{1,3}\s/'),
        // split by words
        $arr = explode(" ", preg_replace(
            array('/[\s;\+\-\/]+/i', '/(\d)[-.\s]+(\d)/', '/\s\w{1,3}\s/'),
            array(' ', '\\1\\2', ' '),
            $norm);
            $str));
        
        foreach ($arr as $i => $part) {
            if (utf8_encode(utf8_decode($part)) == $part) {  // is latin-1 ?
                $arr[$i] = strtr(strtolower(strtr(utf8_decode($part),
                    'ÇçäâàåéêëèïîìÅÉöôòüûùÿøØáíóúñÑÁÂÀãÃÊËÈÍÎÏÓÔõÕÚÛÙýÝ',
                    'ccaaaaeeeeiiiaeooouuuyooaiounnaaaaaeeeiiioooouuuyy')),
                    array('ß' => 'ss', 'ae' => 'a', 'oe' => 'o', 'ue' => 'u'));
            }
            else
                $arr[$i] = strtolower($part);
        }
        
        return join(" ", $arr);
    }
    
}