有些关于URL去重的方面代码没有提供，须要自己去实现。主要这里提供思路
项目地址:http://git.oschina.net/woshidaniu/GuozhongCrawler/tree/master/example/changeProxyIp/
首先爬虫入口类：
public class PervadeSpider {
public static void main(String[] args) {
CrawTaskBuilder builder = CrawlManager.getInstance()
.prepareCrawlTask("看准网漫爬虫", DefaultPageDownloader.class)
.useThread(200)// 使用多个线程下载
.useDynamicEntrance(DynamicEntranceImpl.class)
.useProxyIpPool(KanzhunProxyIpPool.class, 800, 1000 * 60 * 20, 30)
.useQueueSimpleBlockingRequest()//採用广度优先策略，当然redis队列也是fifo。
假设想做分布式爬虫的话能够设置redis队列
.usePageEncoding(PageEncoding.UTF8);
CrawlTask spider = builder.build();
CrawlManager.getInstance().start(spider);
}
public static final class DynamicEntranceImpl extends DynamicEntrance{
@Override
public List<StartContext> loadStartContext() {
StartContext context = new StartContext();
context.injectSeed(context.createPageRequest("http://www.kanzhun.com/companyl/search/?
ka=banner-com", ExtractUrlPageProcessor.class));//公司
context.injectSeed(context.createPageRequest("http://www.kanzhun.com/salaryl/search/?stype=&ka=banner-salary", ExtractUrlPageProcessor.class));//工资
context.injectSeed(context.createPageRequest("http://www.kanzhun.com/jobl/p/?
ka=banner-recruit", ExtractUrlPageProcessor.class));//招聘
context.injectSeed(context.createPageRequest("http://www.kanzhun.com/interviewl/search/?stype=&ka=banner-interview", ExtractUrlPageProcessor.class));//面试
context.injectSeed(context.createPageRequest("http://www.kanzhun.com/topic/100.html?ka=com-topic-1", ExtractUrlPageProcessor.class));//公司之最
return Arrays.asList(context);
}
}
}
动态代理IP提供类:
public class KanzhunProxyIpPool extends ProxyIpPool {
public static final String IP_RESOURCE = "地址忽略";//地址请求的个数必须设置为initProxyIp(int size)中size的个数
public KanzhunProxyIpPool(int initSize, long pastTime, int max_use_count) {
super(initSize, pastTime, max_use_count);
}
private Pattern extractIp = Pattern.compile("([\\d]{1,3}\\.[\\d]{1,3}\\.[\\d]{1,3}\\.[\\d]{1,3}):(\\d+)");
@Override
protected List<ProxyIp> initProxyIp(int size) throws Exception {
List<ProxyIp> ip = new ArrayList<ProxyIp>();
URL url = null;
BufferedReader br = null;
StringBuffer buf = new StringBuffer();
try {
url = new URL(IP_RESOURCE);
InputStream in = url.openStream();
br = new BufferedReader(new InputStreamReader(in,"utf-8"));
String temp = null;
while((temp = br.readLine())!=null){
buf.append(temp).append("\n");
}
ProxyIp proxy = null;
Matcher matcher = extractIp.matcher(buf);
while(matcher.find()){
proxy = new ProxyIp(matcher.group(1), Integer.parseInt(matcher.group(2)));
ip.add(proxy);
}
} catch (Exception e) {
e.printStackTrace();
System.out.println(buf);
}finally{
if(br != null){
try {
br.close();
} catch (IOException e) {
e.printStackTrace();
}
}
}
return ip;
}
}
漫爬页面处理类：
public class ExtractUrlPageProcessor implements PageProcessor {
private String domain = "http://www.kanzhun.com";
private List<URLFilter> urlFilters = new ArrayList<URLFilter>();
private List<Pattern> urlRegulars = null;
public ExtractUrlPageProcessor(){
System.out.println("载入漫爬抽取URL正则");
urlRegulars = ConfigReader.getExtractRegular();
System.out.println("载入漫爬规则完成");
addURLFilter(new URLFilter() {
@Override
public boolean filter(String url) {
return !url.contains("javascript");//去除jquery标签
}
});
addURLFilter(new URLFilter() {
@Override
public boolean filter(String url) {
return url.contains("http://www.kanzhun.com");//保证域名内URL
}
});
addURLFilter(new URLFilter() {
@Override
public boolean filter(String url) {
for (Pattern pattern : urlRegulars) {
boolean result = pattern.matcher(url).find();
if(result)
return true;
}
return false;//保证url符合urlRegulars里随意一个正则
}
});
}
/**
* kanzhunId抽取正则
*/
final static Pattern[] idExtracts = new Pattern[]{
Pattern.compile("gso(\\d+)[^\\d]+"),//简单介绍抽取公司id
Pattern.compile("gsr(\\d+)[^\\d]+"),//点评抽取公司id
Pattern.compile("gsm(\\d+)[^\\d]+"),//面试抽取公司id
Pattern.compile("gsx(\\d+)[^\\d]+"),//工资抽取公司id
Pattern.compile("g(\\d+)[^\\d]+"),//招聘抽取公司id
Pattern.compile("gsp(\\d+)[^\\d]+"),//照片抽取公司id
Pattern.compile("gsl(\\d+)[^\\d]+")//员工抽取公司id
};
@Override
public PageScript getJavaScript() {
// TODO Auto-generated method stub
return null;
}
private Pattern normalContain = Pattern.compile("看准网");
@Override
public Pattern getNormalContain() {
return normalContain;
}
@Override
public void process(OkPage page,StartContext context,List<BasicRequest> queue,List<Proccessable> objectContainer) throws Exception {
// TODO Auto-generated method stub
/**
* 每一个页面抽取的符合urlRegulars规则的url
*/
Set<String> extractUrls = new HashSet<String>();
/**
* 每一个页面全部的kanzhunId
* 这里解释下比方阿里巴巴的首页是http://www.kanzhun.com/gso9012.html?ka=com1-title
* 那个阿里巴巴的kanzhunId就是9012
* 我们能够依据这个推导出阿里巴巴的
* 点评页：http://www.kanzhun.com/gsr9012.html?ka=com-blocker1-review
* 面试页：http://www.kanzhun.com/gsm9012.html?ka=com-blocker1-interview
* 工资页：http://www.kanzhun.com/gsx9012.html?ka=com-blocker1-salary
* 招聘页：http://www.kanzhun.com/job/g9012.html?ka=com-blocker1-job
* 照片页：http://www.kanzhun.com/gsp9012.html?ka=com-blocker1-photo
* 员工页：http://www.kanzhun.com/gsl9012.html?ka=com-blocker1-employee
*
*/
Set<String> kanzhunIds = new HashSet<String>();
Document doc = Jsoup.parse(page.getContent());
Elements allLinks = doc.select("a");
String href = null;
for (Element link : allLinks) {
href = link.attr("href");
if(href.startsWith("/")){
href = domain+href;
}
if(pass(href)){
extractUrls.add(href);
}
//抽取页面全部包括的kanzhunID
for (Pattern pattern : idExtracts) {
Matcher matcher = pattern.matcher(href);
if(matcher.find()){
kanzhunIds.add(matcher.group(1));
}
}
}
//step1
System.out.println(page.getRequest().getUrl()+"抽取了URL"+extractUrls.size()+"个:");
//对url去重（这个须要你们自己实现这里用伪代码表示）
System.out.println("去出反复url...");
//step2
System.out.println(page.getRequest().getUrl()+"抽取了kanzhunId"+kanzhunIds.size()+"个:");
//对抓过的kanzhunId进行去重（这个须要你们自己实现这里用伪代码表示）
System.out.println("kanzhunId进行去重...");
//将抽取的URL增加到队列
for (String extractUrl:extractUrls) {
PageRequest pageRequest = context.createPageRequest(extractUrl, ExtractUrlPageProcessor.class);
queue.add(pageRequest);//加到队列
}
//将抽取的kanzhunId封装成每一个企业的主页URL。
抓取企业信息
for (String kanzhunId:kanzhunIds) {
PageRequest pageRequest = context.createPageRequest("http://www.kanzhun.com/gso"+kanzhunId+".html?
ka=com1-title", CompanyPageProcessor.class);
queue.add(pageRequest);//加到队列
}
}
@Override
public void processErrorPage(Page arg0, StartContext arg1) throws Exception {
// TODO Auto-generated method stub
}
/**
* 对每一个URL进行filter
* @param url
* @return
*/
private boolean pass(String url){
for (URLFilter filter : urlFilters) {
if(!filter.filter(url)){
return false;
}
}
return true;
}
public void addURLFilter(URLFilter urlFilter){
urlFilters.add(urlFilter);
}
}
最后URL抽取规则贴上。
<ExtractRegular>
<!-- 行业URL 或者 公司标签 或者 城市-->
<Regular>/pl[act][a-z0-9]+\.html</Regular>
<!-- 行业URL 或者 城市-->
<Regular>/xs[ac][a-z0-9]+\.html</Regular>
<!-- 招聘分类URL-->
<Regular>/jobli_.+</Regular>
<!-- 面试分类URL-->
<Regular>/ms[ac].+</Regular>
<!-- 公司之最-->
<Regular>/topic/[a-z0-9]+</Regular>
<!-- 热门职位-->
<Regular>/salary/(\d+)/</Regular>
<Regular>/interview/(\d+)/</Regular>
</ExtractRegular>
搞定。