remoteurlfilter.java

java写的搜索引擎网络爬虫 java写的搜索引擎网络爬虫 java写的搜索引擎网络爬虫

/* 
 * FileName FilterRemoteUrl.java
 * Create Time 2006-5-16 16:18:35
 * Author shiwei
 * Descript 判断是否是远程URL
 * Version 
 */

package com.snoics.reptile.regex.filter;

import java.util.Iterator;
import java.util.List;
import java.util.regex.Pattern;

import com.snoics.base.util.StringClass;
import com.snoics.base.util.regex.Regex;
import com.snoics.reptile.system.common.Common;
import com.snoics.reptile.system.common.CommonObject;

public class RemoteUrlFilter implements IRegexFilter{
	private String url="";
	private Regex regex=new Regex();
	private CommonObject commonObject=new CommonObject();
	
	/**
	 * 获取当前URL
	 * @return Returns the url.
	 */
	public String getUrl() {
		return url;
	}
	/**
	 * 设置当前URL
	 * @param url The url to set.
	 */
	public void setUrl(String url) {
		this.url = url;
	}
	
	/**
	 * 判断是否属于远程URL
	 * @return boolean
	 */
	public boolean filter() {
		String remoteUrlFlags=commonObject.getConfigInfo(Common.CONFIGFILE_NODE_REMOTEURLFLAG);
		String website=commonObject.getConfigInfo(Common.CONFIGFILE_NODE_WEBSITE);
		List remoteUrlFlagList=null;
		remoteUrlFlagList=StringClass.getInterString(Common.STRING_SEPARATE_FLAG,remoteUrlFlags);
		boolean flag=false;
		if((remoteUrlFlagList!=null)&&(remoteUrlFlagList.size()>0)){
			Iterator iterator=remoteUrlFlagList.iterator();
			while(iterator.hasNext()) {
				String remoteUrlFlag=(String)iterator.next();
				if(regex.find(url,remoteUrlFlag,Pattern.CASE_INSENSITIVE)) {
					if(!regex.find(url,website,Pattern.CASE_INSENSITIVE)){
						return true;
					}
				}
			}
		}
		return flag;
	}
	
	public static void main(String[] args) {
		RemoteUrlFilter filterRemoteUrl=new RemoteUrlFilter();
		String url="HTTPs://fjdsklfjkdslfjldsf";
		filterRemoteUrl.setUrl(url);
		System.out.println(filterRemoteUrl.filter());
		

		url="HTTPs://fjdsklfjkdslfjldsf";
		filterRemoteUrl.setUrl(url);
		System.out.println(filterRemoteUrl.filter());
		
		url="http://www.squarecn.com/fjdsklfjkdslfjldsf";
		filterRemoteUrl.setUrl(url);
		System.out.println(filterRemoteUrl.filter());
		
	}
}