自己写的php中文截取函数mb_strlen和mb_substr

众所周知，php 自带的 strlen 与 substr 函数没法处理中文字符，于是，我们会用 mb_ 系列函数替代。但是，没有 mbstring 库怎么办？这就需要我们自己写一个来替代了，废话不多说，先上代码:

if ( !function_exists('mb_strlen') ) {
 function mb_strlen ($text, $encode) {
  if ($encode=='utf-8') {
   return preg_match_all('%(?:
       [\x09\x0a\x0d\x20-\x7e]           # ascii
     | [\xc2-\xdf][\x80-\xbf]            # non-overlong 2-byte
     |  \xe0[\xa0-\xbf][\x80-\xbf]       # excluding overlongs
     | [\xe1-\xec\xee\xef][\x80-\xbf]{2} # straight 3-byte
     |  \xed[\x80-\x9f][\x80-\xbf]       # excluding surrogates
     |  \xf0[\x90-\xbf][\x80-\xbf]{2}    # planes 1-3
     | [\xf1-\xf3][\x80-\xbf]{3}         # planes 4-15
     |  \xf4[\x80-\x8f][\x80-\xbf]{2}    # plane 16
     )%xs',$text,$out);
  }else{
   return strlen($text);
  }
 }
}

/* from internet, author unknown */
if (!function_exists('mb_substr')) {
    function mb_substr($str, $start, $len = '', $encoding="utf-8"){
        $limit = strlen($str);
 
        for ($s = 0; $start > 0;--$start) {// found the real start
            if ($s >= $limit)
                break;
 
            if ($str[$s] <= "\x7f")
                ++$s;
            else {
                ++$s; // skip length
 
                while ($str[$s] >= "\x80" && $str[$s] <= "\xbf")
                    ++$s;
            }
        }
 
        if ($len == '')
            return substr($str, $s);
        else
            for ($e = $s; $len > 0; --$len) {//found the real end
                if ($e >= $limit)
                    break;
 
                if ($str[$e] <= "\x7f")
                    ++$e;
                else {
                    ++$e;//skip length
 
                    while ($str[$e] >= "\x80" && $str[$e] <= "\xbf" && $e < $limit)
                        ++$e;
                }
            }
 
        return substr($str, $s, $e - $s);
    }
}