本文共 6102 字，大约阅读时间需要 20 分钟。
无论使用任何语言做模拟登陆或者抓取访问页面，无外乎以下思路： 第一 启用一个web访问会话方法或者实例化一个web访问类，如.net中的HttpWebRequest； 第二 模拟POST或者GET方式提交的数据； 第三 模拟请求的头； 第四 提交请求并获得响应，及对响应做我们所需要的处理。 这里我们以人人网的登录为例，将涉及到POST以及GET两种请求方式。 在之前的文章《免费网页抓包工具，火狐插件FireBug的抓包使用教程》中我们知道，登陆人人网的时候，一共做了一个POST请求以及两个GET请求，如下图： 人人网登录请求 观察这三个请求的详细信息，不难看出第一个GET请求的地址可以由POST的响应得到，而第二个GET请求的地址又由第一个GET的响应得到。 先来模拟第一个POST请求 view plaincopy to clipboardprint? HttpWebRequest request = null; HttpWebResponse response = null; string gethost = string.Empty; CookieContainer cc = new CookieContainer(); string Cookiesstr = string.Empty; try { //第一次POST请求 string postdata = "email=" + UserName.Replace("@", "%40") + "&password=" + PassWord + "&origURL=" + HostUrl + "&domain=renren.com";//模拟请求数据，数据样式可以用FireBug插件得到。人人网POST数据时，用户名邮箱中的“@”变为“%40”，所以我们也要作此变化 string LoginUrl="http://www.renren.com/PLogin.do"; request = (HttpWebRequest)WebRequest.Create(LoginUrl);//实例化web访问类 request.Method = "POST";//数据提交方式为POST //模拟头 request.ContentType = "application/x-www-form-urlencoded"; byte[] postdatabytes = Encoding.UTF8.GetBytes(postdata); request.ContentLength = postdatabytes.Length; //request.Referer = "http://www.renren.com/Login.do?rf=r&domain=renren.com&origURL=" + HostUrl; request.AllowAutoRedirect = false; request.CookieContainer = cc; request.KeepAlive = true; //提交请求 Stream stream; stream = request.GetRequestStream(); stream.Write(postdatabytes, 0, postdatabytes.Length); stream.Close(); //接收响应 response = (HttpWebResponse)request.GetResponse(); //保存返回cookie response.Cookies = request.CookieContainer.GetCookies(request.RequestUri); CookieCollection cook = response.Cookies; string strcrook = request.CookieContainer.GetCookieHeader(request.RequestUri); Cookiesstr = strcrook; //取第一次GET跳转地址 StreamReader sr = new StreamReader(response.GetResponseStream(), Encoding.UTF8); string content = sr.ReadToEnd(); response.Close(); string[] substr = content.Split(new char[] { '"' }); gethost = substr[1]; } catch (Exception) { //第一次POST出错； } HttpWebRequest request = null; HttpWebResponse response = null; string gethost = string.Empty; CookieContainer cc = new CookieContainer(); string Cookiesstr = string.Empty; try { //第一次POST请求 string postdata = "email=" + UserName.Replace("@", "%40") + "&password=" + PassWord + "&origURL=" + HostUrl + "&domain=renren.com";//模拟请求数据，数据样式可以用FireBug插件得到。人人网POST数据时，用户名邮箱中的“@”变为“%40”，所以我们也要作此变化 string LoginUrl="http://www.renren.com/PLogin.do"; request = (HttpWebRequest)WebRequest.Create(LoginUrl);//实例化web访问类 request.Method = "POST";//数据提交方式为POST //模拟头 request.ContentType = "application/x-www-form-urlencoded"; byte[] postdatabytes = Encoding.UTF8.GetBytes(postdata); request.ContentLength = postdatabytes.Length; //request.Referer = "http://www.renren.com/Login.do?rf=r&domain=renren.com&origURL=" + HostUrl; request.AllowAutoRedirect = false; request.CookieContainer = cc; request.KeepAlive = true; //提交请求 Stream stream; stream = request.GetRequestStream(); stream.Write(postdatabytes, 0, postdatabytes.Length); stream.Close(); //接收响应 response = (HttpWebResponse)request.GetResponse(); //保存返回cookie response.Cookies = request.CookieContainer.GetCookies(request.RequestUri); CookieCollection cook = response.Cookies; string strcrook = request.CookieContainer.GetCookieHeader(request.RequestUri); Cookiesstr = strcrook; //取第一次GET跳转地址 StreamReader sr = new StreamReader(response.GetResponseStream(), Encoding.UTF8); string content = sr.ReadToEnd(); response.Close(); string[] substr = content.Split(new char[] { '"' }); gethost = substr[1]; } catch (Exception) { //第一次POST出错； } 注释写的很详细了，在这就不再分析，也许有人对request = (HttpWebRequest)WebRequest.Create(LoginUrl)有疑问，可以去google一下HttpWebRequest和WebRequest的区别，简单来说WebRequest是一个抽象类，不能直接实例化，需要被继承，而HttpWebRequest继承自WebRequest。 再模拟第一个和第二个GET请求 view plaincopy to clipboardprint? try { request = (HttpWebRequest)WebRequest.Create(gethost); request.Method = "GET"; request.KeepAlive = true; request.Headers.Add("Cookie:" + Cookiesstr); request.CookieContainer = cc; request.AllowAutoRedirect = false; response = (HttpWebResponse)request.GetResponse(); //设置cookie Cookiesstr = request.CookieContainer.GetCookieHeader(request.RequestUri); //取再次跳转链接 StreamReader sr = new StreamReader(response.GetResponseStream(), Encoding.UTF8); string ss = sr.ReadToEnd(); string[] substr = ss.Split(new char[] { '"' }); gethost = substr[1]; request.Abort(); sr.Close(); response.Close(); } catch (Exception) { //第一次GET出错 } try { //第二次GET请求 request = (HttpWebRequest)WebRequest.Create(gethost); request.Method = "GET"; request.KeepAlive = true; request.Headers.Add("Cookie:" + Cookiesstr); request.CookieContainer = cc; request.AllowAutoRedirect = false; response = (HttpWebResponse)request.GetResponse(); //设置cookie Cookiesstr = request.CookieContainer.GetCookieHeader(request.RequestUri); request.Abort(); response.Close(); } catch (Exception) { //第二次GET出错 } try { request = (HttpWebRequest)WebRequest.Create(gethost); request.Method = "GET"; request.KeepAlive = true; request.Headers.Add("Cookie:" + Cookiesstr); request.CookieContainer = cc; request.AllowAutoRedirect = false; response = (HttpWebResponse)request.GetResponse(); //设置cookie Cookiesstr = request.CookieContainer.GetCookieHeader(request.RequestUri); //取再次跳转链接 StreamReader sr = new StreamReader(response.GetResponseStream(), Encoding.UTF8); string ss = sr.ReadToEnd(); string[] substr = ss.Split(new char[] { '"' }); gethost = substr[1]; request.Abort(); sr.Close(); response.Close(); } catch (Exception) { //第一次GET出错 } try { //第二次GET请求 request = (HttpWebRequest)WebRequest.Create(gethost); request.Method = "GET"; request.KeepAlive = true; request.Headers.Add("Cookie:" + Cookiesstr); request.CookieContainer = cc; request.AllowAutoRedirect = false; response = (HttpWebResponse)request.GetResponse(); //设置cookie Cookiesstr = request.CookieContainer.GetCookieHeader(request.RequestUri); request.Abort(); response.Close(); } catch (Exception) { //第二次GET出错 } GET与POST请求大同小异，这里便不再累述。三次请求结束，保存好你的cookie string，每次请求的时候都赋给请求的头部，你就处于登录状态了。 人人网的HttpWebRequest登陆模拟很简单，但是POST及GET涉及到了，是个不错的案例。 当然，在.net想做自动访问的操作还可以使用WebBrowser控件，而且还能够和HttpWebRequest共用cookie，抛砖引玉一下不在本篇文章的讨论范围。转载地址：http://pyunb.baihongyu.com/