近期研究多线程下载，写了个demo。整理下来，或许会对别人有帮助。
多线程下载的话一般开启两到三个线程吧。假设线程太多的话时间会浪费在线程的切换上，倒是浪费了大把的时间。线程多了也不是一件好事。
原理的话看我的还有一篇博文，其实是将代码移植到了安卓系统上。java实现的多线程下载demo
public class MainActivity extends Activity {
protected static final int DOWNLOAD_ERROR = 1;
private static final int THREAD_ERROR = 2;
public static final int DWONLOAD_FINISH = 3;
private EditText et_path;
private EditText et_count;
/**
* 存放进度条的布局
*/
private LinearLayout ll_container;
/**
* 进度条的集合
*/
private List<ProgressBar> pbs;
/**
* android下的消息处理器。在主线程创建。才干够更新ui
*/
private Handler handler = new Handler(){
public void handleMessage(Message msg) {
switch (msg.what) {
case DOWNLOAD_ERROR:
Toast.makeText(getApplicationContext(), "下载失败", 0).show();
break;
case THREAD_ERROR:
Toast.makeText(getApplicationContext(), "下载失败,请重试", 0).show();
break;
case DWONLOAD_FINISH:
Toast.makeText(getApplicationContext(), "完成下载", 0).show();
break;
}
};
};
/**
* 线程的数量
*/
private int threadCount = 3;
/**
* 每一个下载区块的大小
*/
private long blocksize;
/**
* 正在执行的线程的数量
*/
private int runningThreadCount;
@Override
protected void onCreate(Bundle savedInstanceState) {
super.onCreate(savedInstanceState);
setContentView(R.layout.activity_main);
et_path = (EditText) findViewById(R.id.et_addr);
et_count = (EditText) findViewById(R.id.et_num);
ll_container = (LinearLayout) findViewById(R.id.ll_pb);
}
public void download(View view){
//下载文件的路径
final String path = et_path.getText().toString().trim();
if(TextUtils.isEmpty(path)){
Toast.makeText(this, "对不起下载路径不能为空", 0).show();
return;
}
String count = et_count.getText().toString().trim();
if(TextUtils.isEmpty(path)){
Toast.makeText(this, "对不起,线程数量不能为空", 0).show();
return;
}
threadCount = Integer.parseInt(count);
//清空掉旧的进度条
ll_container.removeAllViews();
//在界面里面加入count个进度条
pbs = new ArrayList<ProgressBar>();
for(int j=0;j<threadCount;j++){
ProgressBar pb = (ProgressBar) View.inflate(this, R.layout.pb, null);
ll_container.addView(pb);
pbs.add(pb);
}
Toast.makeText(this, "開始下载", 0).show();
new Thread(){
public void run() {
try {
URL url = new URL(path);
HttpURLConnection conn = (HttpURLConnection) url.openConnection();
conn.setRequestMethod("GET");
conn.setConnectTimeout(5000);
int code = conn.getResponseCode();
if (code == 200) {
long size = conn.getContentLength();// 得到服务端返回的文件的大小
System.out.println("server文件的大小：" + size);
blocksize = size / threadCount;
// 1.首先在本地创建一个大小跟server一模一样的空白文件。
File file = new File(Environment.getExternalStorageDirectory(),getFileName(path));
RandomAccessFile raf = new RandomAccessFile(file, "rw");
raf.setLength(size);
// 2.开启若干个子线程分别去下载相应的资源。
runningThreadCount = threadCount;
for (int i = 1; i <= threadCount; i++) {
long startIndex = (i - 1) * blocksize;
long endIndex = i * blocksize - 1;
if (i == threadCount) {
// 最后一个线程
endIndex = size - 1;
}
System.out.println("开启线程：" + i + "下载的位置：" + startIndex + "~"
+ endIndex);
int threadSize = (int) (endIndex - startIndex);
pbs.get(i-1).setMax(threadSize);
new DownloadThread(path, i, startIndex, endIndex).start();
}
}
conn.disconnect();
} catch (Exception e) {
e.printStackTrace();
Message msg = Message.obtain();
msg.what = DOWNLOAD_ERROR;
handler.sendMessage(msg);
}
};
}.start();
}
private class DownloadThread extends Thread {
private int threadId;
private long startIndex;
private long endIndex;
private String path;
public DownloadThread(String path, int threadId, long startIndex,
long endIndex) {
this.path = path;
this.threadId = threadId;
this.startIndex = startIndex;
this.endIndex = endIndex;
}
@Override
public void run() {
try {
// 当前线程下载的总大小
int total = 0;
File positionFile = new File(Environment.getExternalStorageDirectory(),getFileName(path)+threadId + ".txt");
URL url = new URL(path);
HttpURLConnection conn = (HttpURLConnection) url
.openConnection();
conn.setRequestMethod("GET");
// 接着从上一次的位置继续下载数据
if (positionFile.exists() && positionFile.length() > 0) {// 推断是否有记录
FileInputStream fis = new FileInputStream(positionFile);
BufferedReader br = new BufferedReader(
new InputStreamReader(fis));
// 获取当前线程上次下载的总大小是多少
String lasttotalstr = br.readLine();
int lastTotal = Integer.valueOf(lasttotalstr);
System.out.println("上次线程" + threadId + "下载的总大小："
+ lastTotal);
startIndex += lastTotal;
total += lastTotal;// 加上上次下载的总大小。
fis.close();
//存数据库。
//_id path threadid total
}
conn.setRequestProperty("Range", "bytes=" + startIndex + "-"
+ endIndex);
conn.setConnectTimeout(5000);
int code = conn.getResponseCode();
System.out.println("code=" + code);
InputStream is = conn.getInputStream();
File file = new File(Environment.getExternalStorageDirectory(),getFileName(path));
RandomAccessFile raf = new RandomAccessFile(file, "rw");
// 指定文件開始写的位置。
raf.seek(startIndex);
System.out.println("第" + threadId + "个线程：写文件的開始位置："
+ String.valueOf(startIndex));
int len = 0;
byte[] buffer = new byte[1024];
while ((len = is.read(buffer)) != -1) {
RandomAccessFile rf = new RandomAccessFile(positionFile,
"rwd");
raf.write(buffer, 0, len);
total += len;
rf.write(String.valueOf(total).getBytes());
rf.close();
pbs.get(threadId-1).setProgress(total);
}
is.close();
raf.close();
} catch (Exception e) {
e.printStackTrace();
Message msg = Message.obtain();
msg.what = THREAD_ERROR;
handler.sendMessage(msg);
} finally {
// 仅仅有全部的线程都完成下载后 才干够删除记录文件。
synchronized (MainActivity.class) {
System.out.println("线程" + threadId + "完成下载了");
runningThreadCount--;
if (runningThreadCount < 1) {
System.out.println("全部的线程都工作完成了。
删除暂时记录的文件");
for (int i = 1; i <= threadCount; i++) {
File f = new File(Environment.getExternalStorageDirectory(),getFileName(path)+ i + ".txt");
System.out.println(f.delete());
}
Message msg = Message.obtain();
msg.what = DWONLOAD_FINISH;
handler.sendMessage(msg);
}
}
}
}
}
private String getFileName(String path){
int start = path.lastIndexOf("/")+1;
return path.substring(start);
}
}
安卓系统下须要加上訪问网络的权限和訪问本地内存卡的权限。例如以下图：
源代码已经上传到了CSDN: http://download.csdn.net/detail/rootusers/8508137
时间： 2024-10-05 22:13:45