ViewSource

viewsource.cs

using System;
using System.Text;
using System.Net;
using System.Text.RegularExpressions;
using System.IO;

class Program
{
    static void Main()
    {
        ViewSource("http://www.baidu.com/");
        ViewSource("http://www.google.com/");
        ViewSource("http://www.163.com/");
    }

    static String ViewSource(string url)
    {
        string source = string.Empty, charset = string.Empty;
        HttpWebRequest httpreq = (HttpWebRequest)WebRequest.Create(url);
        using (HttpWebResponse httpresp = (HttpWebResponse)httpreq.GetResponse())
        {
            using (MemoryStream stream = new MemoryStream())
            {
                httpresp.GetResponseStream().CopyTo(stream);
                charset = httpresp.CharacterSet;
                if (string.IsNullOrEmpty(charset) || charset == "ISO-8859-1")
                {
                    source = Encoding.GetEncoding("ISO-8859-1").GetString(stream.GetBuffer(), 0, (int)stream.Length);
                    stream.Position = 0;
                    if (Regex.IsMatch(source, @"^(?:[\x00-\x7f]|[\xfc-\xff][\x80-\xbf]{5}|[\xf8-\xfb][\x80-\xbf]{4}|[\xf0-\xf7][\x80-\xbf]{3}|[\xe0-\xef][\x80-\xbf]{2}|[\xc0-\xdf][\x80-\xbf])+$"))
                    {
                        charset = "utf-8";
                    }
                    else
                    {
                        charset = "gb2312";
                    }
                }
                Console.WriteLine(httpresp.CharacterSet + " => " + charset);
                source = Encoding.GetEncoding(charset).GetString(stream.GetBuffer(), 0, (int)stream.Length);
            }
        }
        Console.WriteLine(source.Substring(0, source.Length > 500 ? 500 : source.Length));
        return source;
    }
}