爬虫三：小案例模拟登陆慢慢买并获取用户积分

忘是亡心i 2024-04-19 13:02 56阅读 0赞

public class ReptileLogin {
    
        @Test
        public void Login() throws Exception{
            //确定url:
            String indexUrl = "http://home.manmanbuy.com/login.aspx";
            //2、发送请求，获取数据
            CloseableHttpClient httpClient = HttpClients.createDefault();
            HttpPost post = new HttpPost(indexUrl);
    
            //设置请求参数
            List<BasicNameValuePair> params = new ArrayList<BasicNameValuePair>();
            params.add(new BasicNameValuePair("__VIEWSTATE","/wEPDwULLTIwNjQ3Mzk2NDFkGAEFHl9fQ29udHJvbHNSZXF1aXJlUG9zdEJhY2tLZXlfXxYBBQlhdXRvTG9naW4voj01ABewCkGpFHsMsZvOn9mEZg=="));
            params.add(new BasicNameValuePair("__EVENTVALIDATION","/wEWBQLW+t7HAwLB2tiHDgLKw6LdBQKWuuO2AgKC3IeGDJ4BlQgowBQGYQvtxzS54yrOdnbC"));
            params.add(new BasicNameValuePair("txtUser","账户名"));
            params.add(new BasicNameValuePair("txtPass","密码"));
            params.add(new BasicNameValuePair("btnLogin","登录"));
    
            //设置请求头信息
            /*网站加了防盗链，必须设置是从哪个网页跳转过来的*/
            post.setHeader("Referer","http://home.manmanbuy.com/login.aspx");
            post.setHeader("User-Agent","Mozilla/5.0 (Windows NT 6.1; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/74.0.3729.131 Safari/537.36");
            //将参数加入请求体
            HttpEntity entity = new UrlEncodedFormEntity(params);
            post.setEntity(entity);
            //执行
            CloseableHttpResponse response = httpClient.execute(post);
    
            httpClient.close();
    
            //判断是否成功跳转
            if(response.getStatusLine().getStatusCode()==302){
                //获取响应头信息
                Header[] locations = response.getHeaders("Location");
                //System.out.println(locations[0].getValue());
                /*跳转之后的url：http://home.manmanbuy.com/usercenter.aspx
                * locations[0].getValue()：usercenter.aspx
                * */
                //拼接新的url
                String new_url = "http://home.manmanbuy.com"+locations[0].getValue();
                //拿到响应头中Set-Cookie的值
                Header[] cookies = response.getHeaders("Set-Cookie");
    
                //重新发送请求，获取数据
                httpClient = HttpClients.createDefault();
                HttpGet get = new HttpGet(new_url);
    
                //重新设置请求头信息（因为url变，网页重定向，需要手动设置用户cookie信息）
                //Cookie的默认写法，[0]为隐藏，[1]可以直接看到
                get.setHeader("Cookie",cookies[0].getValue()+" "+cookies[1].getValue());
    
                CloseableHttpResponse execute_info = httpClient.execute(get);
                String info = EntityUtils.toString(execute_info.getEntity(), "gb2312");
    
                /*-----------------------------------------------------------*/
                //解析数据
                Document parse = Jsoup.parse(info);
                Element element = parse.select("#aspnetForm > div.udivright > div:nth-child(2) > table > tbody > tr > td:nth-child(1) > table:nth-child(2) > tbody > tr > td:nth-child(2) > div:nth-child(1) > font").get(0);
               //输出当前用户账户积分
                System.out.println(element.text());
    
    
            }
        }
    
    }