public classTest {/***
*简单抓取当当图书分类中某一页指定信息输出到控制台并保存到文件中*/
public static void main(String[] args) throwsIOException {
CloseableHttpClient httpclient=HttpClients.createDefault();//创建一个文件，用来保存信息
BufferedWriter writer=new BufferedWriter(new FileWriter("D:\book.csv"));try{//发送请求URL填入当当网图书分类某一页面的地址
HttpGet httpget = new HttpGet("/cp01.36.04.08.00.00.html");
System.out.println("Executing request " +httpget.getRequestLine());//Create a custom response handler
ResponseHandler responseHandler = response ->{int status =response.getStatusLine().getStatusCode();if (status >= 200 && status < 300) {
HttpEntity entity=response.getEntity();return entity != null ? EntityUtils.toString(entity) : null;
}else{throw new ClientProtocolException("Unexpected response status: " +status);
}
};//得到请求体也就是页面源码responseBody
String responseBody =httpclient.execute(httpget, responseHandler);
System.out.println("----------------------------------------");//使用Jsoup解析得到一个document对象，代表这个页面
Document document=Jsoup.parse(responseBody);//这是人为分析源码中的数据后，取docunment中需要的元素
Element pos=document.getElementsByClass("bigimg").get(0);
Elements list=pos.children();for(Element e:list){
Element name= e.getElementsByClass("pic").get(0);
Element detail= e.getElementsByClass("detail").get(0);
Element author= e.getElementsByAttributeValue("name","itemlist-author").get(0);
Element press= e.getElementsByAttributeValue("name","P_cbs").get(0);
Element market= e.getElementsByClass("search_pre_price").get(0);
Element sale= e.getElementsByClass("search_now_price").get(0);
System.out.println("图书名："+name.attr("title"));
System.out.println("简介："+detail.text());
System.out.println("作者："+author.text());
System.out.println("出版社："+press.text());
System.out.println("市场价："+market.text());
System.out.println("惊喜价："+sale.text());
System.out.println("--------------------");//添加要写入文件的信息
writer.write(name.attr("title")+","+detail.text()+","+author.text()+","+press.text());
writer.newLine();
}
}finally{
writer.close();
httpclient.close();
}
}
}
