public static function toNgram($str, $n = 2)
{
$res = array();
$str = self::removeMarkdown($str);
$str = self::stripCrLf($str);
$len = mb_strlen($str, 'UTF8');
$index = 0;
while (true) {
$res[] = mb_substr($str, $index, $n, 'UTF8');
$index++;
if ($index + $n > $len) {
break;
}
}
return implode(' ', $res);
}