本帖最后由 fair_jm 于 2012-9-20 19:52 编辑
去学校的网站下点东西 发现没有批量吐血 顿时热血澎湃啊.....不说废话了
这是咱学校网站的一个下载页:http://kczy.zjut.edu.cn/jsjwl/downloadcenter.asp?Page=2
如果chrome说有恶意程序你自己判断 因为...我们学校的神马精品课程管理系统等等都会跳
废话不多说 源码贴上 bug是一定有的 我测试了两次发现可以用之后就不去管了(下午还得上课 晚上也是...额)
所以注释什么的也没写.......以后再补上 如果有人要注释的话
原理超简单 下载速度很慢 没用NIO的关系吧？...
[mw_shl_code=java,true]package org.cc.network;
/**
* 用于简单下载一个网页(未加密)的特定后缀的文件
* @author  Fair_jm
* @version 1.0
*/
import java.io.File;
import java.io.FileOutputStream;
import java.io.FileWriter;
import java.io.IOException;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.Reader;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.Date;
import java.util.concurrent.Executor;
import java.util.concurrent.ExecutorService;
import java.util.concurrent.Executors;
import java.util.concurrent.Semaphore;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
public class FileDownloader {
private int limit=5;
private Semaphore licence=null;
private ExecutorService exec=null;
private String suffix="ppt";
private String base_page="http://kczy.zjut.edu.cn/jsjwl/";
private String task_page;
private StringBuffer page_buffer;
public int getLimit() {
return limit;
}
public void setLimit(int limit) throws Exception {
if (this.limit!=licence.availablePermits())
throw new Exception("程序正在运行,无法修改");
this.limit = limit;
this.licence=new Semaphore(limit);
}
public FileDownloader(String task_page){
this.page_buffer=new StringBuffer();
this.task_page=task_page;
this.licence=new Semaphore(limit);
this.exec=Executors.newCachedThreadPool();
}
public FileDownloader(String task_page,String base_page){
this(task_page);
this.base_page=base_page;
}
public FileDownloader(String task_page,String base_page,ExecutorService exec){
this(task_page,base_page);
this.exec=exec;
}
public FileDownloader(String task_page,String base_page,ExecutorService exec,int limit){
this(task_page,base_page,exec);
licence=new Semaphore(limit);
this.base_page=base_page;
}
private void init() throws IOException{
URL url=new URL(task_page);
Reader reader=new InputStreamReader(url.openStream());
char[] buf = new char[1024];
int len=0;
while(!((len=reader.read(buf))<0)){
page_buffer.append(buf,0,len);
}
reader.close();
}
public void download() throws IOException{
this.init();
Pattern pattern = Pattern.compile("<a href=\"(.*)"+"\\."+suffix+"\"");
Matcher matcher = pattern.matcher(page_buffer);
while(matcher.find()){
String file_page = base_page + matcher.group(1)+"."+suffix;
//System.out.println(file_page);
exec.submit(new DownloadHelp(file_page));
}
}
private class DownloadHelp implements Runnable{
private String task;
private URL url;
public DownloadHelp(String url) throws IOException{
this.task=url;
this.url=new URL(url);
}
@Override
public void run() {
FileOutputStream fw=null;
InputStream reader=null;
try {
FileDownloader.this.licence.acquire();
} catch (InterruptedException e) {
Thread.currentThread().interrupt();
e.printStackTrace();
}
try{
String[] names=task.split("/");
String fileName=names[names.length-1];
//System.out.println(fileName);
File file=new File(FileDownloader.class.getResource(".").getPath()+fileName);
reader=url.openStream();
byte[] buf = new byte[1024];
Date date_begin=new Date();
int len=0;
int temp=0;
while(file.exists()){
String temps=Integer.toString(temp);
String name=FileDownloader.class.getResource(".").getPath()+temps+fileName;
file=new File(name);
temp++;
//System.out.println(file.getAbsolutePath());
}
//System.out.println(file.getAbsolutePath());
System.out.println(file.getName()+"启动");
file.createNewFile();
fw=new FileOutputStream(file);
while(!((len=reader.read(buf))<0)){
fw.write(buf, 0, len);
}
Date date_end=new Date();
System.out.println(file.getName()+"完成"+"\n用时:"+(date_end.getTime()-date_begin.getTime()));
FileDownloader.this.licence.release();
}catch(IOException e){
e.printStackTrace();
}finally{
try {
fw.close();
reader.close();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
}
}
}
[/mw_shl_code]
一下测试用的main方法
[mw_shl_code=java,true]package org.cc.network;
import java.io.IOException;
public class Test {
public static void main(String[] args) {
FileDownloader fd=new FileDownloader("http://kczy.zjut.edu.cn/jsjwl/downloadcenter.asp?Page=2");
try {
fd.download();
} catch (IOException e) {
// TODO Auto-generated catch block
e.printStackTrace();
}
}
}
[/mw_shl_code]
以下结果:
[mw_shl_code=text,true]020101025111274266.ppt启动
02010102511822860.ppt启动
0201010251110374374.ppt启动
0201010251110374374.ppt完成
用时:11982ms
020101025111274266.ppt完成
用时:13411ms
02010102511822860.ppt完成
用时:15095ms
[/mw_shl_code]