黑马程序员——网页爬虫（网页蜘蛛）

------- <a href="http://www.itheima.com"
target="blank">android培训</a>、<a href="http://www.itheima.com"
target="blank">java培训</a>、期待与您交流！ ----------

-通过学习IO流，网络编程，正则表达式等知识编写一个用于获取电子邮箱的程序——网页爬虫

package day25;

import java.io.*;
import java.net.*;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

public class GetMailDemo {

public static void main(String[] args) throws Exception {
getNetMails();

}
//取得网络上的Email
public static void getNetMails() throws Exception{
URL url= new URL("http://10.0.31.236:8080/testWeb/email.html");

URLConnection conn= url.openConnection();

BufferedReader bufIn= new BufferedReader(new InputStreamReader(conn.getInputStream()));

String line= null;

String mailreg= "\\w+@\\w+(\\.\\w+)+";
Pattern p= Pattern.compile(mailreg);

while((line=bufIn.readLine())!=null){
//System.out.println(line);
Matcher m= p.matcher(line);
while(m.find()){
System.out.println(m.group());
}
}
}
//取得本地文件中的Email
public static void getMails() throws Exception{

@SuppressWarnings("resource")
BufferedReader bufr= new BufferedReader(new FileReader("mail.txt"));

String line= null;

String mailreg= "\\w+@\\w+(\\.\\w+)+";
Pattern p= Pattern.compile(mailreg);

while((line=bufr.readLine())!=null){
//System.out.println(line);
Matcher m= p.matcher(line);
while(m.find()){
System.out.println(m.group());
}
}
}

}