在做网络爬虫项目的时候，有些网站设置了权限，只有在登录了之后才能爬取网站的内容，如何模拟登录呢？请看下面的代码！

后面持续更新腾讯微博、新浪微博模拟登陆等，请关注！

// 获取登陆后的 client 就可以访问明星的任何页面数据
	@SuppressWarnings("deprecation")
	public static DefaultHttpClient login(String u, String p) {

		DefaultHttpClient client = new DefaultHttpClient();
		HttpClientParams.setCookiePolicy(client.getParams(),
				CookiePolicy.BROWSER_COMPATIBILITY);

		try {
			/** 获得rsaPubkey,rsakv,servertime等参数值 **/
			HashMap params = preLogin(encodeAccount(u), client);

			/******** 登录操作 *********/
			HttpPost post = new HttpPost(
					"http://login.sina.com.cn/sso/login.php?client=ssologin.js(v1.4.5)");
			post.setHeader("Accept",
					"text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8");
			post.setHeader("User-Agent",
					"Mozilla/5.0 (Windows NT 5.1; rv:9.0.1) Gecko/20100101 Firefox/9.0.1");

			post.setHeader("Accept-Language", "zh-cn,zh;q=0.5");
			post.setHeader("Accept-Charset", "GB2312,utf-8;q=0.7,*;q=0.7");
			post.setHeader("Referer",
					"http://weibo.com/?c=spr_web_sq_firefox_weibo_t001");
			post.setHeader("Content-Type", "application/x-www-form-urlencoded");

			String nonce = makeNonce(6);

			Listnvps = new ArrayList();
			nvps.add(new BasicNameValuePair("encoding", "UTF-8"));
			nvps.add(new BasicNameValuePair("entry", "weibo"));
			nvps.add(new BasicNameValuePair("from", ""));
			nvps.add(new BasicNameValuePair("gateway", "1"));
			nvps.add(new BasicNameValuePair("nonce", nonce));
			nvps.add(new BasicNameValuePair("pagerefer",
					"http://i.firefoxchina.cn/old/"));
			nvps.add(new BasicNameValuePair("prelt", "111"));
			nvps.add(new BasicNameValuePair("pwencode", "rsa2"));
			nvps.add(new BasicNameValuePair("returntype", "META"));
			nvps.add(new BasicNameValuePair("rsakv", params.get("rsakv")));
			nvps.add(new BasicNameValuePair("savestate", "0"));
			nvps.add(new BasicNameValuePair("servertime", params
					.get("servertime")));

			nvps.add(new BasicNameValuePair("service", "miniblog"));

			/******************** *加密密码 ***************************/
			ScriptEngineManager sem = new ScriptEngineManager();
			ScriptEngine se = sem.getEngineByName("javascript");

			se.eval(getJs());
			String pass = "";

			if (se instanceof Invocable) {
				Invocable invoke = (Invocable) se;
				// 调用preprocess方法，并传入两个参数密码和验证码

				pass = invoke.invokeFunction("getpass", p,
						params.get("servertime"), nonce, params.get("pubkey"))
						.toString();

				System.out.println("c = " + pass);
			}

			nvps.add(new BasicNameValuePair("sp", pass));
			nvps.add(new BasicNameValuePair("su", encodeAccount(u)));
			nvps.add(new BasicNameValuePair(
					"url",
					"http://weibo.com/ajaxlogin.php?framelogin=1&callback=parent.sinaSSOController.feedBackUrlCallBack"));

			nvps.add(new BasicNameValuePair("useticket", "1"));

			nvps.add(new BasicNameValuePair("vsnf", "1"));

			post.setEntity(new UrlEncodedFormEntity(nvps, HTTP.UTF_8));

			HttpResponse response = client.execute(post);

			String entity = EntityUtils.toString(response.getEntity());
			if (entity.replace("\"", "").indexOf("retcode=0") > -1) {
				// http://passport.weibo.com/wbsso/login? 新变化的
				// retcode=0 新变化的
				String url = entity.substring(entity
						.indexOf("http://passport.weibo.com/wbsso/login?"),
						entity.indexOf("retcode=0") + 9);
				String nick = "chenkun"; // 昵称

				// 获取到实际url进行连接
				HttpGet getMethod = new HttpGet(url);
				response = client.execute(getMethod);
				entity = EntityUtils.toString(response.getEntity());

				// nick = entity.substring(entity.indexOf("displayname") + 14,
				// entity.lastIndexOf("userdomain") - 3).trim();

				url = entity.substring(entity.indexOf("userdomain") + 13,
						entity.lastIndexOf("\""));
				String original = "p/1035051087770692/weibo?profile_ftype=0&is_ori=1&from=page_103505_home&wvr=5.1&mod=originalweibo#place";
				// 登陆后可以请求 任何明星微博页面数据
				getMethod = new HttpGet("http://weibo.com/" + original);
				response = client.execute(getMethod);
				// 返回页面数据内容
				entity = EntityUtils.toString(response.getEntity());
				// System.out.println(entity);
			}

		} catch (Exception e) {
			e.printStackTrace();

		}
		return client;

	}