HP内の中国語文字列をインターセプトする

2. 截取utf8编码的多字节字符串

< ?php
//截取utf8字符串
function utf8Substr($str, $from, $len)
{
    return preg_replace(&#39;#^(?:[\x00-\x7F]|[\xC0-\xFF][\x80-\xBF]+){0,&#39;.$from.&#39;}&#39;.
                       &#39;((?:[\x00-\x7F]|[\xC0-\xFF][\x80-\xBF]+){0,&#39;.$len.&#39;}).*#s&#39;,
                       &#39;$1&#39;,$str);
}
?>

3. UTF-8、GB2312都支持的汉字截取函数

< ?php
/*
Utf-8、gb2312都支持的汉字截取函数
cut_str(字符串, 截取长度, 开始长度, 编码);
编码默认为 utf-8
开始长度默认为 0
*/
 
function cut_str($string, $sublen, $start = 0, $code = &#39;UTF-8&#39;)
{
    if($code == &#39;UTF-8&#39;)
    {
        $pa = "/[\x01-\x7f]|[\xc2-\xdf][\x80-\xbf]|\xe0[\xa0-\xbf][\x80-\xbf]|[\xe1-\xef][\x80-\xbf][\x80-\xbf]|\xf0[\x90-\xbf][\x80-\xbf][\x80-\xbf]|[\xf1-\xf7][\x80-\xbf][\x80-\xbf][\x80-\xbf]/";
        preg_match_all($pa, $string, $t_string);
 
        if(count($t_string[0]) - $start > $sublen) return join(&#39;&#39;, array_slice($t_string[0], $start, $sublen))."...";
        return join(&#39;&#39;, array_slice($t_string[0], $start, $sublen));
    }
    else
    {
        $start = $start*2;
        $sublen = $sublen*2;
        $strlen = strlen($string);
        $tmpstr = &#39;&#39;;
 
        for($i=0; $i< $strlen; $i++)
        {
            if($i>=$start && $i< ($start+$sublen))
            {
                if(ord(substr($string, $i, 1))>129)
                {
                    $tmpstr.= substr($string, $i, 2);
                }
                else
                {
                    $tmpstr.= substr($string, $i, 1);
                }
            }
            if(ord(substr($string, $i, 1))>129) $i++;
        }
        if(strlen($tmpstr)< $strlen ) $tmpstr.= "...";
        return $tmpstr;
    }
}
 
$str = "abcd需要截取的字符串";
echo cut_str($str, 8, 0, &#39;gb2312&#39;);
?>

4. BugFree 的字符截取函数

< ?php
/**
 * @package     BugFree
 * @version     $Id: FunctionsMain.inc.php,v 1.32 2005/09/24 11:38:37 wwccss Exp $
 *
 *
 * Return part of a string(Enhance the function substr())
 *
 * @author                  Chunsheng Wang <wwccss@263.net>
 * @param string  $String  the string to cut.
 * @param int     $Length  the length of returned string.
 * @param booble  $Append  whether append "...": false|true
 * @return string           the cutted string.
 */
function sysSubStr($String,$Length,$Append = false)
{
    if (strlen($String) < = $Length )
    {
        return $String;
    }
    else
    {
        $I = 0;
        while ($I < $Length)
        {
            $StringTMP = substr($String,$I,1);
            if ( ord($StringTMP) >=224 )
            {
                $StringTMP = substr($String,$I,3);
                $I = $I + 3;
            }
            elseif( ord($StringTMP) >=192 )
            {
                $StringTMP = substr($String,$I,2);
                $I = $I + 2;
            }
            else
            {
                $I = $I + 1;
            }
            $StringLast[] = $StringTMP;
        }
        $StringLast = implode("",$StringLast);
        if($Append)
        {
            $StringLast .= "...";
        }
        return $StringLast;
    }
}
 
$String = "17test.info 走在中国自动化测试的前沿";
$Length = "18";
$Append = false;
echo sysSubStr($String,$Length,$Append);
?>