java 正则表达式 提取 url_正则表达式提取URL

package com.url;

import java.io.IOException;

import java.io.InputStream;

import java.net.URL;

import java.net.URLConnection;

import java.util.regex.Matcher;

import java.util.regex.Pattern;

public class Urlconnection {

private static Pattern pattern = null;

private static Matcher matcher = null;

private static Pattern pattern1 = null;

private static Matcher matcher1 = null;

private static String regulation = ".*|" ;

private static String regulation1="\\bhref\\s*=\\s*(['\"])(.*?)\\1(\\s+|>)";

//\\bhref\\s*=\\s*(['])(.*?)\\1\\b

public static void main(String[] args) throws IOException {

URL url = new URL("http://www.baidu.com/");

URLConnection conn = url.openConnection() ;

InputStream input = conn.getInputStream();

String charater = conn.getContentEncoding();

if(null == charater){

charater = "GBK";

}

byte[] bytes = new byte[4096];

int leng=-1 ;

StringBuffer sb = new StringBuffer("");

pattern = Pattern.compile(regulation,Pattern.CASE_INSENSITIVE|Pattern.MULTILINE|Pattern.DOTALL);

pattern1 = Pattern.compile(regulation1,Pattern.CASE_INSENSITIVE|Pattern.MULTILINE|Pattern.DOTALL);

while((leng=input.read(bytes))!=-1){

sb.append(new String(bytes,charater)) ;

matcher = pattern.matcher(sb.toString());

while(matcher.find()){

//System.out.println(matcher.group());

String str = matcher.group() ;

//str = str.replaceAll("\"", "'");

matcher1 = pattern1.matcher(str);

while(matcher1.find()){

System.out.println(matcher1.group(2));

}

}

sb.delete(0, sb.length());

}

}

}

posted on

2012-05-01 19:43

画一个圆圈

阅读(406)

评论(0)

编辑

收藏