前言：整理些，以前写的有用的Java编写的小程序。这篇是一个简单的网络爬虫获取Email地址。

package com.lianggzone.demo;

import java.io.IOException;
import java.io.InputStream;
import java.net.HttpURLConnection;
import java.net.URL;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

/**
 * 网络爬虫获取Email地址
 * @author LiangGzone
 */
public class EmailSpider {
	public static void main(String[] args){
	   URL url = null;
	   String regex = "[\\w[.-]]+@[\\w[.-]]+\\.[\\w]+";
	   String tempStr = null;
	   Pattern pattern = Pattern.compile(regex);
	   Matcher matcher = null;
	   try {
		   // 获取URL地址
		   url = new URL("http://zhidao.baidu.com/question/92607106.html"); 
		   // 连接URL地址
		   HttpURLConnection urlConnection = (HttpURLConnection)url.openConnection(); 
		   urlConnection.connect();
		   // 定义输入流
		   InputStream in = urlConnection.getInputStream();   
		   byte[] buf = new byte[4096];            
		   while (in.read(buf) > 0) {         
			   	tempStr = new String(buf);
			   	matcher = pattern.matcher(tempStr);
			   	while(matcher.find()) {
			   		System.out.println(matcher.group());
			   	}
		   }
	   	}catch (IOException e) {
	   		e.printStackTrace();
	   	}
	}
}

转自：CSDN

作者：LiangGzone

链接：http://blog.csdn.net/lianggzone/article/details/8871437