一个简单的Java采集文字和图片类,实现从远程网站上获取文本或者图片.

import java.io.BufferedReader;
import java.io.ByteArrayOutputStream;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.net.URL;
import java.net.URLConnection;

public class CollectData
{
//获取文本,返回字符串 start-开始字符串 end-结束字符串
//结果不包含开始串,不包含结束串
public String getText(String textUrl,
String start,String end,String charset)
{
	if(charset == null || charset.length() < 1)
	charset = "GBK";
	try
	{
	URL url = new URL(textUrl);
	URLConnection con = url.openConnection();
	InputStream is = con.getInputStream();
	BufferedReader br = new BufferedReader(
	new InputStreamReader(is,charset));
	String s;
	StringBuffer sb = new StringBuffer();
	while((s = br.readLine())!=null)
	{
	sb.append(s);
	}
	br.close();
	String htmlContent = sb.toString();
	if(htmlContent.indexOf(start) != -1 &&
	htmlContent.lastIndexOf(end) != -1)
	{
	String content = htmlContent.substring(
	htmlContent.indexOf(start)
	+start.length(),htmlContent.lastIndexOf(end));
	return content;
	}
	}catch(Exception e)
	{
	e.printStackTrace();
	}
	return "";
}

//获取图片,返回字节数组
public byte [] getImage(String imgUrl)
{
	try
	{
	URL url = new URL(imgUrl);
	URLConnection con = url.openConnection();
	InputStream is = con.getInputStream();
	ByteArrayOutputStream baos = new ByteArrayOutputStream();
	byte [] b = new byte[1024];
	int length = 0;
	while((length = is.read(b)) > 0)
	{
	baos.write(b,0,length);
	}
	return baos.toByteArray();
	}catch(Exception e)
	{
	e.printStackTrace();
	return null;
	}
}

public static void main(String [] args)
{
	CollectData c = new CollectData();
	String result = c.getText("http://www.baidu.com",
	"","","GBK");
	System.out.println(result);
	c.getImage("http://www.google.cn/logos/newyear09.gif");
}
}