package com.test;
import java.io.IOException;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
public class TestJsoup {
public static void main(String[] args) throws IOException {
for(int i=1;i<10;i++){
String detailUrl = "http://bbs.9game.cn/forum-1096-"+i+".html";
Document doc = Jsoup.connect(detailUrl).userAgent("Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Trident/5.0)").timeout(15000).get();
// System.out.println(doc);
Elements es = doc.getElementsByClass("xst");
for (Element element : es) {
//System.out.println(element);
String linkHref = element.attr("href");
//System.out.println(linkHref);
String linkText = element.text();
//System.out.println(linkText);
String linkInnerH = element.html();
System.out.println(linkInnerH);
Document d = Jsoup.connect(linkHref).userAgent("Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; Trident/5.0)").timeout(15000).get();
Elements ets = d.getElementsByClass("t_f");
for (Element elt : ets) {
//System.out.println(elt);
String linkInnerH1 = elt.html();
//System.out.println(linkInnerH1);
}
}
}
}
}
java 采集论坛
最新推荐文章于 2017-07-19 21:14:19 发布