使用Heritrix爬取国内某最火的电子商城的笔记本信息遇到的问题 -

awenhaowenchao

浏览: 72144 次
性别:
来自: 北京

最近访客更多访客>>

c_kh_zfb

maochong2009

vanber

belllab

博主相关

博客

微博

相册

留言

关于我

文章分类

社区版块

存档分类

使用Heritrix爬取国内某最火的电子商城的笔记本信息遇到的问题

博客分类：

搜索引擎

360 HTML

    网上的某些资料也说了对Heritrix进行扩展定制即可个性化的从网上爬取自己需要的内容，为建立垂直搜索作做好准备。
    这里主要扩展FrontierScheduler和Extractor两个class，由于是个demo所以比较简单，由于这些代码仅仅是周日下午和今天晚上抽时间些的，故效率上没有可考虑，比如太多的正则，反射方法构造后应该缓存等，仅供参考

    扩展如下：

    class FrontierSchedulerFor360Buy extends FrontierScheduler

    //至于为什么要扩展schedule大家应该都知道，CandidateURI表示代选的链接，那么意思就是对将要进入候选的链接进行业务上的过滤

    注意:这里有个问题，就是如果需要过滤符合某个格式的链接，但是这个格式的链接不是全部过滤，而且也找不到一定的业务规则去过滤，那么就不能在这里过滤了，后面在Extractor中讲解


	protected void schedule(CandidateURI caUri) {
		// TODO Auto-generated method stub
		String url = caUri.toString();
		System.err.println(">>>++++++ " + url);

		if(url.indexOf("dns:www.360buy.com") != -1){
			System.err.println(">>>>>>++++++ " + url);
			this.getController().getFrontier().schedule(caUri);
		}
		if(url.indexOf("http://www.360buy.com/robots.txt") != -1){
			System.err.println(">>>>>>++++++ " + url);
			this.getController().getFrontier().schedule(caUri);
		}

		if(url.indexOf(PATTERN_360_NOTEBOOK)!=-1){
			System.err.println(">>>>>>++++++ " + url);
			this.getController().getFrontier().schedule(caUri);
		}
		
		if(url.matches(this.PATTERN_360_NOTEBOOK_BRAND)){
			if(url.indexOf("670-671-672-0")!=-1){
				return;
			}
			System.err.println(">>>>>>++++++ " + url);
			this.getController().getFrontier().schedule(caUri);
		}
		
	}

    class ExtractorFor360Buy extends Extractor
    这里扩展的话基本是对crawlUri的解析，以及对其中的一部分符合业务规则的链接加入到候选的链接（已经加入就需要过滤了，这就遇到了上面红色注意部分说道的问题），这样我们需要直接将链接交由 BdbFrontier 处理，那么根据源码查看，是因为在schedul候选链接CadicateUri的时候，会getPathFromSeed()，所以需要通过反射设置pathFromSeed


以下为具体的异常：

java.lang.NullPointerException
	at org.archive.crawler.datamodel.CandidateURI.getTransHops(CandidateURI.java:382)
	at org.archive.crawler.frontier.AbstractFrontier.applySpecialHandling(AbstractFrontier.java:727)
	at org.archive.crawler.frontier.WorkQueueFrontier.receive(WorkQueueFrontier.java:442)
	at org.archive.crawler.util.SetBasedUriUniqFilter.add(SetBasedUriUniqFilter.java:90)
	at org.archive.crawler.frontier.WorkQueueFrontier.schedule(WorkQueueFrontier.java:427)
	at com.awen.heritrix.ExtractorFor360Buy.extract(ExtractorFor360Buy.java:124)
	at org.archive.crawler.extractor.Extractor.innerProcess(Extractor.java:67)
	at org.archive.crawler.framework.Processor.process(Processor.java:109)
	at org.archive.crawler.framework.ToeThread.processCrawlUri(ToeThread.java:306)
	at org.archive.crawler.framework.ToeThread.run(ToeThread.java:154)


protected void extract(CrawlURI curi) {
		// TODO Auto-generated method stub
		String currentUrl = curi.toString();
		//如果是笔记本品牌的话，直接解析，进行数据的
		if(currentUrl.matches(this.PATTERN_360_NOTEBOOK_BRAND_ABSOLUTE)){
			
			NodeFilter reputation_clild_filter = new AndFilter(new TagNameFilter("span"), 
				      new HasAttributeFilter("class", "reputation"));
			/**
			 * 此filter为笔记本产品的基本信息，标题，价格，评论链接，图片等
			 */
			NodeFilter div_child_filter = new HasChildFilter(reputation_clild_filter);
			NodeFilter product_filter = new HasChildFilter(div_child_filter);
			try {
				parser.setURL(currentUrl);
				//获取品牌的编码http://www.360buy.com/products/670-671-672-[1-9][\\d]*+-0-0-0-0-0-0-1-1-[\\d].html
				String brandNo ="";
				brandNo = currentUrl.substring(currentUrl.indexOf("670-671-672-") + 12, currentUrl.indexOf("-", "http://www.360buy.com/products/670-671-672-".length()));
				NodeList nodeList = parser.parse(product_filter);
				for (int i = 0; i < nodeList.size(); ++i){
					Node node_li = nodeList.elementAt(i);					
					//String node_li_html = node_li.toHtml();使用正则匹配的话效率太低
					
					try {
						
						String content = node_li.toHtml();
						
						String detail="";
						String img="";
						String price="";
						String remark="";
						String review = "";
						
						//产品详情链接
						Pattern pattern = Pattern.compile(this.PATTERN_360_PRODUCT,Pattern.CASE_INSENSITIVE);
						Matcher matcher = pattern.matcher(content);
						if(matcher.find()){
							String temp = matcher.group();
							
							CandidateURI cUri = new CandidateURI(UURIFactory.getInstance(temp));
							//在往BdbFrontier直接加入CandidateURI过程中会调用CandidateURI的getPathFromSeed，由于为保护方法，所以需要反射设置
							Method m = cUri.getClass().getDeclaredMethod("setPathFromSeed", String.class);
							m.setAccessible(true);
							m.invoke(cUri, temp);
							m.setAccessible(false);
							this.getController().getFrontier().schedule(cUri);
							
							detail = temp.substring(temp.lastIndexOf("/")+1, temp.indexOf(".html"));
							System.err.println(detail);
						}
						
						//产品图片链接
						Pattern pattern1 = Pattern.compile(this.PATTERN_360_PRODUCT_IMG,Pattern.CASE_INSENSITIVE);
						Matcher matcher1 = pattern1.matcher(content);
						if(matcher1.find()){
							String temp = matcher1.group();
							this.addLindFromString(curi, temp, "", Link.EMBED_HOP);
							img = temp.substring(temp.lastIndexOf("/")+1, temp.indexOf(".jpg"));
							System.err.println(img);
							this.getController().getFrontier().schedule(new org.archive.crawler.datamodel.CandidateURI(UURIFactory.getInstance(temp)));
						}
						
						//产品评论链接
						Pattern pattern2 = Pattern.compile(this.PATTERN_360_PRODUCT_EVALUATE,Pattern.CASE_INSENSITIVE);
						Matcher matcher2 = pattern2.matcher(content);
						if(matcher2.find()){
							String temp = matcher2.group();
							//this.addLindFromString(curi, temp, "", Link.NAVLINK_HOP);
							review = temp;
						}
						
						
						//价格
						NodeFilter price_filter = new AndFilter(new TagNameFilter("div"), 
							      new HasAttributeFilter("class", "p-price"));
						//名称
						NodeFilter name_filter = new AndFilter(new TagNameFilter("div"), 
							      new HasAttributeFilter("class", "p-name"));			
						
						//获取价格
						Parser parser = new Parser();
						parser.setInputHTML(content);
						NodeList price_list = parser.parse(price_filter);
						TextExtractingVisitor vistior = new TextExtractingVisitor();
						new Parser(price_list.elementAt(0).toHtml()).visitAllNodesWith(vistior);
						price =vistior.getExtractedText();
						
						System.err.println("price = " + price);
						
						//获取名称描述
						Parser parser1 = new Parser();
						parser1.setInputHTML(content);
						NodeList name_list = parser1.parse(name_filter);
						TextExtractingVisitor vistior1 = new TextExtractingVisitor();
						new Parser(name_list.elementAt(0).toHtml()).visitAllNodesWith(vistior1);
						remark = vistior1.getExtractedText();
						
						System.err.println("name = " + remark);
						try {
							File file = new File("/home/awen/360/" + brandNo + "_" + detail + ".txt");
							if(!file.exists()){
								file.createNewFile();
							}
							BufferedWriter bw = new BufferedWriter(new FileWriter(file));
							bw.write(brandNo + ls);//品牌
							bw.write(detail + ls);//详情链接后缀
							bw.write(img + ls);//图片链接后缀
							bw.write(price + ls);//价格
							bw.write(review + ls);//评价链接							
							bw.write(remark + ls);//描述
							bw.flush();
							bw.close();
						} catch (IOException e) {
							// TODO Auto-generated catch block
							e.printStackTrace();
						}
					
						
					}  catch (ParserException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					} catch (URIException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					} catch (SecurityException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					} catch (NoSuchMethodException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					} catch (IllegalArgumentException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					} catch (IllegalAccessException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					} catch (InvocationTargetException e) {
						// TODO Auto-generated catch block
						e.printStackTrace();
					}
								
					

		        }
			} catch (ParserException e) {
				// TODO Auto-generated catch block
				e.printStackTrace();
			}
		}
		
		//如果不是笔记本品牌的话，直接解析
		HttpRecorder recorder = curi.getHttpRecorder();
		ReplayCharSequence rcs = null;
		try {
			rcs = recorder.getReplayCharSequence();
		} catch (IOException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}

		if(rcs==null){
			return;
		}
		
		String content = rcs.toString();
		
		Pattern pattern = Pattern.compile(this.PATTERN_A_HREF, Pattern.CASE_INSENSITIVE);
		Matcher matcher = pattern.matcher(content);
		
		while(matcher.find()){
			String newUrl = matcher.group(2);
			//符合笔记本
			if(newUrl.matches(this.PATTERN_360_NOTEBOOK)){
				System.err.println("————————————————————————————————符合笔记本——————————————————————————");
				//对newUrl进行修改
				newUrl = "http://www.360buy.com/products" + newUrl;
				this.addLindFromString(curi, newUrl, "", Link.NAVLINK_HOP);
			}
			
			//符合笔记本品牌
			if(newUrl.matches(this.PATTERN_360_NOTEBOOK_BRAND)){
				//对newUrl进行修改
				System.err.println("————————————————————————————————符合笔记本品牌——————————————————————————");
				newUrl = "http://www.360buy.com/products/" + newUrl;
				this.addLindFromString(curi, newUrl, "", Link.NAVLINK_HOP);
			}
		}

	}
	
	//将链接加入到候选链接中，如果在Frontier中也对这种格式的链接需要过滤却不能分辨哪些是否需要过滤，而这些链接
	//又必须要抓取，那么可以直接交给BdbFrontier依次上级调用
	private void addLindFromString(CrawlURI curi, String uri, CharSequence context, char hopType){
		try {
			curi.createAndAddLinkRelativeToBase(uri, context, hopType);
		} catch (URIException e) {
			// TODO Auto-generated catch block
			e.printStackTrace();
		}
	}