htmlpaser打造个性化的爬虫程序第二天

本文链接：https://blog.csdn.net/hymcn/article/details/7449775

任务1：

抽取页面中的所有链接

  LinkBean lb = new LinkBean ();
              lb.setURL ("http://sthaboutme.sinaapp.com/");
              URL[] urls = lb.getLinks ();
              for (int i = 0; i < urls.length; i++)
                  System.out.println (urls[i]);

任务2：

抽取页面中满足既定条件的链接

try {
			Parser parser = new Parser("http://sthaboutme.sinaapp.com");
			String matchPattern = "http://sthaboutme.sinaapp.com/?";
			NodeFilter filter = new LinkRegexFilter(matchPattern);
			NodeList nlist = parser.extractAllNodesThatMatch(filter);
			System.out.println(nlist.size());
			for(int i= 0 ;i < nlist.size();i++)
			{
				LinkTag link =(LinkTag)nlist.elementAt(i);
				System.out.println(link.getLink());
			}
			
		} catch (ParserException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}

任务3：

抽取页面中满足多条件的链接

	try {
			Parser parser = new Parser("http://sthaboutme.sinaapp.com");
			String StrContain = "http://";
			String StrNotContain ="#";
			NodeFilter filter1 = new LinkRegexFilter(StrContain);
			NodeFilter filter2 = new StringFilter(StrNotContain)
			{
				 public boolean accept (Node node)
				    {
				        boolean ret = true;			    
				        if (LinkTag.class.isAssignableFrom (node.getClass ()))
				        {
				            String link = ((LinkTag)node).getLink ();
				                if (link.indexOf (mPattern) > -1)
				                {
				                	 ret = false; 
				                	// System.out.print(mPattern);
				                }
				                       
				        }
				        return ret;
				    };
			};
			
			AndFilter andFilter = new AndFilter(filter1,filter2);	
			NodeList nlist = parser.extractAllNodesThatMatch(andFilter);
			System.out.println(nlist.size());
			for(int i= 0 ;i < nlist.size();i++)
			{
				LinkTag link =(LinkTag)nlist.elementAt(i);
				System.out.println(link.getLink());
			}
			
		} catch (ParserException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}