java document的用法_Java Document.getElementsByAttribute方法代码示例

import org.jsoup.nodes.Document; //导入方法依赖的package包/类

public List call() {

lstHouseInfo = new ArrayList();

CloseableHttpClient client = HttpClients.createDefault();

Lock lock = new ReentrantLock();

HttpGet get = new HttpGet(url);

//设置请求头

get.setHeader("Accept", "text/html, application/xhtml+xml, image/jxr, */*");

get.setHeader("Accept-Language", "zh-Hans-CN, zh-Hans; q=0.8, en-US; q=0.5, en; q=0.3");

get.setHeader("User-Agent", "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36");

try{

//获取响应

CloseableHttpResponse response = client.execute(get);

HttpEntity entity = response.getEntity();

String html = EntityUtils.toString(entity, "UTF-8");

//解析html

Document doc = Jsoup.parse(html);

Elements lists = doc.getElementsByAttribute("logr");

lock.lock();

try{

for (Element list : lists) {

//将关键信息提取出来

HouseInfo houseInfo = new HouseInfo();

String[] houseInfoArray = list.getElementsByTag("h2").first().text().split(" ");

houseInfo.setHouseTitle(list.getElementsByTag("h2").first().text());

houseInfo.setHouseURL("http://" + "bj" + ".58.com" + list.getElementsByTag("a").first().attributes().get("href"));

houseInfo.setMoney(list.getElementsByClass("money").tagName("b").text());

houseInfo.setHouseLocation(houseInfoArray[1]);

lstHouseInfo.add(houseInfo);

}

}finally {

lock.unlock();

}

}catch (IOException e){

System.out.println( e.getMessage());

}

return lstHouseInfo;

}