爬虫简单案例

1.原生jdk发送get请求

//演示: 原生jdk发送get请求
public class JDKRequest {

    public static void main(String[] args) throws Exception {
        //1.确定url
        String indexUrl = "http://www.baidu.com";

        //2.发送请求获取数据。
        //封装URL对象
        URL url = new URL(indexUrl);
        //获取连接
        HttpURLConnection urlConnection = (HttpURLConnection)url.openConnection();

        //设置请求方式
        urlConnection.setRequestMethod("GET");

        InputStream in = urlConnection.getInputStream();

        InputStreamReader inputStreamReader = new InputStreamReader(in, "GB2312");
        //字符缓冲流
        BufferedReader bufferedReader = new BufferedReader(inputStreamReader);
        //字符串缓冲区
        StringBuilder stringBuilder = new StringBuilder();

        String len = null;
        //按行读
        while((len=bufferedReader.readLine())!=null){
            //追加到字符串缓冲区存放
            stringBuilder.append(len);
        }
        System.out.println(stringBuilder.toString());
        in.close();
    }
}

2.原生jdk发送post请求

// 模拟 原生jdk发送post请求
public class JDKPost {


    public static void main(String[] args) throws Exception {
        //确定URL
        String indexUrl = "https://www.douban.com/";
        //发送请求获取数据
        //封装URL对象
        URL url = new URL(indexUrl);
        //获取连接
        HttpURLConnection connection = (HttpURLConnection) url.openConnection();
        //设置请求方式
        connection.setRequestMethod("POST");
        //打开输出流
        connection.setDoOutput(true);
        //设置请求参数
        OutputStream out = connection.getOutputStream();
        out.write("username=liang&password=123".getBytes());


        //获取字节流
        InputStream in = connection.getInputStream();


        InputStreamReader inputStreamReader = new InputStreamReader(in, "UTF-8");
        //字符缓冲流
        BufferedRe