[轉貼] 獲取網頁編碼，自動區分網頁編碼

出處：http://www.piaoyi.org/c-sharp/378.html

        protected void Page_Load(object sender, EventArgs e)
        {
            Response.Write(getHtml("http://tw.yahoo.com/", null));
        }

        //獲取網頁編碼，自動區分網頁編碼
        private string getHtml(string url, string charSet)
        //url是要訪問的網站地址，charSet是目標網頁的編碼
        //如果傳入的是null或者""，就自動分析網頁的編碼
        {
            WebClient myWebClient = new WebClient();
            myWebClient.Proxy = null;
            byte[] myDataBuffer = myWebClient.DownloadData(url);
            string strWebData = Encoding.Default.GetString(myDataBuffer);

            //獲取網頁編碼描述訊息
            Match charSetMatch = Regex.Match(strWebData, " ]*)charset=(\")?(.*)?\"", RegexOptions.IgnoreCase | RegexOptions.Multiline);
            string webCharSet = charSetMatch.Groups[3].Value;
            if (charSet == null || charSet == "") charSet = webCharSet;

            if (charSet != null && charSet != "" && Encoding.GetEncoding(charSet) != Encoding.Default)
            {
                strWebData = Encoding.GetEncoding(charSet).GetString(myDataBuffer);
            }
            return strWebData;
        }