我分段写入的是一张大图片
简略了对流的关闭的操作
package newFeatures8;
import java.io.BufferedReader;
import java.io.File;
import java.io.FileInputStream;
import java.io.InputStream;
import java.io.InputStreamReader;
import java.io.RandomAccessFile;
import java.net.HttpURLConnection;
import java.net.URL;
public class MuchThreadDown {
private static int threadCount = 3;// 开启3个线程
private static int blockSize = 0;// 每个线程下载的大小
private static int runningTrheadCount = 0;// 当前运行的线程数
private static String path = "http://img.ivsky.com/img/tupian/pre/201707/29/shuiguo_yuebing-001.jpg";
public static void main(String[] args) {
try {
// 1.请求url地址获取服务端资源的大小
URL url = new URL(path);
HttpURLConnection openConnection = (HttpURLConnection) url.openConnection();
openConnection.setRequestMethod("GET");
openConnection.setConnectTimeout(10 * 1000);
int code = openConnection.getResponseCode();
if (code == 200) {
// 获取资源的大小
//如果该资源的大小未知或是其字节比Integer.MAX_VALUE(非常非常接近2G)还大,则返回-1
int filelength = openConnection.getContentLength();
String fileName = getFileName(path);
// 2.在本地创建一个与服务端资源同样大小的一个文件(占位)
RandomAccessFile randomAccessFile = new RandomAccessFile(new File(fileName), "rw");
randomAccessFile.setLength(filelength);// 设置随机访问文件的大小,本地的占位文件
randomAccessFile.close();
// 3.要分配每个线程下载文件的开始位置和结束位置。开始分段
blockSize = filelength / threadCount;// 计算出每个线程理论下载大小
for (int threadId = 0; threadId < threadCount; threadId++) {
int startIndex = threadId * blockSize;// 计算每个线程下载的开始位置
int endIndex = (threadId + 1) * blockSize - 1;// 计算每个线程下载的结束位置
// 如果是最后一个线程,结束位置需要单独计算
if (threadId == threadCount - 1) {
endIndex = filelength - 1;
}
// 4.开启线程去执行下载
new Thread(new DownloadThread(threadId, startIndex, endIndex)).start();
}
}
} catch (Exception e) {
e.printStackTrace();
}
}
public static class DownloadThread implements Runnable{
private int threadId;
private int startIndex;
private int endIndex;
private int lastPostion;
public DownloadThread(int threadId, int startIndex, int endIndex) {
this.threadId = threadId;
this.startIndex = startIndex;
this.endIndex = endIndex;
}
@Override
public void run() {
synchronized (DownloadThread.class) {//静态类的锁是当前类所属的字节码文件对象
runningTrheadCount = runningTrheadCount + 1;// 开启一线程,线程数加1
}
// 分段请求网络连接,分段保存文件到本地
try {
URL url = new URL(path);
HttpURLConnection openConnection = (HttpURLConnection) url.openConnection();
openConnection.setRequestMethod("GET");
openConnection.setConnectTimeout(10 * 1000);
System.out.println("理论上下载: 线程:" + threadId + ",开始位置:" + startIndex + ";结束位置:" + endIndex);
// 读取上次下载结束的位置,本次从这个位置开始直接下载。
File file2 = new File(threadId + ".jpg");
if (file2.exists()) {
BufferedReader bufferedReader = new BufferedReader(
new InputStreamReader(new FileInputStream(file2)));
String lastPostion_str = bufferedReader.readLine();
lastPostion = Integer.parseInt(lastPostion_str);// 读取文件获取上次下载的位置
bufferedReader.close();
} else {
lastPostion = startIndex;
}
// 设置分段下载的头信息。 Range:做分段数据请求用的。
// bytes:0-500:请求服务器资源中0-500之间的字节信息
openConnection.setRequestProperty("Range", "bytes:" + lastPostion + "-" + endIndex);
System.out.println("实际下载: 线程:" + threadId + ",开始位置:" + lastPostion + ";结束位置:" + endIndex);
if (openConnection.getResponseCode() == 200) {// 200:请求全部资源成功,
// 206代表部分资源请求成功
InputStream inputStream = openConnection.getInputStream();
// 请求成功将流写入本地文件中,已经创建的占位那个文件中
RandomAccessFile randomAccessFile = new RandomAccessFile(new File(getFileName(path)), "rw");
randomAccessFile.seek(lastPostion);// 设置随机文件从哪个位置开始写。
// 将流中的数据写入文件
byte[] buffer = new byte[1024];//定义一个1k的缓冲字节数组
int length = -1;
int total = 0;// 记录本次线程下载的总大小
while ((length = inputStream.read(buffer)) != -1) {
//开始写入到本地文件
randomAccessFile.write(buffer, 0, length);
total += length;
// 去保存当前线程下载的位置,保存到文件中
int currentThreadPostion = lastPostion + total;// 计算出当前线程本次下载的位置
// 创建临时的 随机文件保存当前线程下载的位置
File file = new File(threadId + ".jpg");
RandomAccessFile accessfile = new RandomAccessFile(file, "rwd");
accessfile.write(String.valueOf(currentThreadPostion).getBytes());
accessfile.close();
}
// 关闭相关的流信息
inputStream.close();
randomAccessFile.close();
System.out.println("线程:" + threadId + ",下载完毕");
// 当所有线程下载结束,删除存放下载位置的文件。
synchronized (DownloadThread.class) {
runningTrheadCount--;// 标志着一个线程下载结束。
if (runningTrheadCount == 0) {
System.out.println("所有线程下载完成");
for (int threadId = 0; threadId < threadCount; threadId++) {
File file = new File(threadId + ".jpg");
System.out.println(file.getAbsolutePath());
file.delete();
}
System.out.println("合并完成!");
}
}
}
} catch (Exception e) {
e.printStackTrace();
}
}
}
public static String getFileName(String url) {
return url.substring(url.lastIndexOf("/") + 1);
}
}