1. 程式人生 > >php實現gbk和uft8編碼中英文字元擷取函式的應用

php實現gbk和uft8編碼中英文字元擷取函式的應用

<?php
/*
utf-8、gb2312都支援的漢字擷取函式
cut_str(www.3ppt.com字串, 擷取長度, 開始長度, 編碼);
編碼預設為 utf-8
開始長度預設為 0
*/

function cut_str($string, $sublen, $start = 0, $code = 'utf-8')
{
    if($code == 'utf-8')
    {
        $pa = "/[x01-x7f]|[xc2-xdf][x80-xbf]|xe0[xa0-xbf][x80-xbf]|[xe1-xef][x80-xbf][x80-xbf]|xf0[x90-xbf][x80-xbf][x80-xbf]|[xf1-xf7][x80-xbf][x80-xbf][x80-xbf]/";
        preg_match_all($pa, $string, $t_string);

        if(count($t_string[0]) - $start > $sublen) return join('', array_slice($t_string[0], $start, $sublen))."...";
        return join('', array_slice($t_string[0], $start, $sublen));
    }
    else
    {
        $start = $start*2;
        $sublen = $sublen*2;
        $strlen = strlen($string);
        $tmpstr = '';

        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;
    }
}

$str = "abcd需要擷取的字串";
echo cut_str($str, 8, 0, 'gb2312');
?>