日期:2014-05-17 浏览次数:21136 次
通过jsoup解析页面html获取优酷页面视频列表
作者: javaboy2012
Email:yanek@163.com
qq:    1046011462
代码如下:
package com.yanek;
import java.io.IOException;
import java.util.HashMap;
import org.jsoup.Jsoup;
import org.jsoup.nodes.Document;
import org.jsoup.nodes.Element;
import org.jsoup.select.Elements;
public class Tool {
	/**
	 * @param args
	 */
	public static void main(String[] args) {
		
		String url="http://www.youku.com";
		//url="http://movie.youku.com";
		
		HashMap list=new HashMap();
		
		Document doc;
		try {
			doc = Jsoup.connect(url).get();
			
			Elements links = doc.select("a[href]");
			
			int s=0;
			
			for (Element link : links) {
				String v_url=link.attr("abs:href");
				
				if 	(link.ownText().length()==0)
				{
					continue;
				}
				if (list.containsKey(v_url))
				{
					continue;
				}
				
				if (v_url.startsWith("http://v.youku.com/v_show"))
				{
					System.out.println(link.attr("abs:href")+"-"+link.ownText());
					list.put(v_url, v_url);
					s++;
				}
			}
			System.out.println("total:"+ s);
			
			
		} catch (IOException e) {
			e.printStackTrace();
		}
		
		
	}
}
 