package com.yeshi.fanli.util.goods.jd;
|
|
import java.io.IOException;
|
import java.util.ArrayList;
|
import java.util.List;
|
|
import org.jsoup.Jsoup;
|
import org.jsoup.nodes.Document;
|
import org.jsoup.nodes.Element;
|
import org.jsoup.select.Elements;
|
|
/**
|
* 京东内优惠商品爬取 https://www.n-youhui.com
|
*
|
* @author Administrator
|
*
|
*/
|
public class NYouHuiUtil {
|
public static List<String> getClasses() {
|
List<String> classList = new ArrayList<>();
|
try {
|
Document doc = Jsoup.connect("https://www.n-youhui.com").timeout(20000)
|
.userAgent(
|
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/78.0.3904.108 Safari/537.36")
|
.get();
|
Element classesRoot = doc.getElementsByClass("widget_categories").get(0);
|
Elements level0 = classesRoot.getElementsByClass("level-0");
|
for (int i = 0; i < level0.size(); i++) {
|
String name = level0.get(i).html().replace(" ", "").replaceAll("\\(([0-9]|,)*\\)", "").trim();
|
classList.add(name);
|
}
|
} catch (IOException e) {
|
e.printStackTrace();
|
}
|
return classList;
|
}
|
|
}
|