java尝试HTTP多线程断点续传下载

利用commons-fi

简单的Socket实现的

9月10

大 | 中 | 小

11:06Heck

编程杂谈 From: 本站原创

主要是实现HTTP多线程断点续传下载

，直接看代码吧，废话一点不多说 stupid

。

功能介绍：

1 多线程HTTP下载
2 支持断点续传
3 临时文件下载，成功后改名
4 提供防盗链的破解

package net.heck.tools;

import java.io.BufferedInputStream;
import java.io.BufferedWriter;
import java.io.File;
import java.io.OutputStreamWriter;
import java.io.RandomAccessFile;
import java.net.Socket;
import java.net.URL;
import java.net.URLConnection;
import java.util.HashMap;
import java.util.Map;

/**
 * HTTP的多线程下载工具。
 */
public class HTTPDownloader extends Thread {
  // 要下载的页面
  private String page;

// 保存的路径
  private String savePath;

// 线程数
  private int threadNumber = 2;

// 来源地址
  private String referer;

private String cookie;

int threadPointer = 0;

private Map<Integer, HTTPDownloaderThread> threadPool = new HashMap<Integer, HTTPDownloaderThread>(); // 线程迟

// 最小的块尺寸。如果文件尺寸除以线程数小于这个，则会减少线程数。
  private int MIN_BLOCK = 10 * 1024;

public static void main(String[] args) throws Exception {
    HTTPDownloader d = new HTTPDownloader("http://www.xxxxx.com/a.rar", null, "d://a.rar", 10, null);
    d.down();
  }

public void run() {
    try {
      down();
    } catch (Exception e) {
      e.printStackTrace();
    }
  }

/**
   * 下载操作
   * 
   * @throws Exception
   */
  public void down() throws Exception {
    URL url = new URL(page); // 创建URL
    URLConnection con = url.openConnection(); // 建立连接
    con.setRequestProperty("Referer", referer == null ? page : referer);
    con.setRequestProperty("UserAgent", "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; flashget)");
    int contentLen = con.getContentLength(); // 获得资源长度
    if ((contentLen / MIN_BLOCK + 1) < threadNumber) {
      threadNumber = contentLen / MIN_BLOCK + 1; // 调整下载线程数
    }
    if (threadNumber > 10) {
      threadNumber = 10;
    }
    int begin = 0;
    int step = contentLen / threadNumber + 1;
    int end = 0;
    HTTPDownloaderThread thread;
    for (threadPointer = 0; threadPointer < threadNumber; threadPointer++) {
      end += step;
      if (end > contentLen) {
        end = contentLen;
      }
      thread = new HTTPDownloaderThread(this, threadPointer, begin, end);
      threadPool.put(threadPointer, thread);
      thread.start();
      begin = end;
    }
  }

/**
   * 一个线程完活了。
   * 
   * @param id 完活的线程id
   */
  public synchronized void finished(int id) {
    threadNumber--;
    threadPool.remove(id);
    if (threadNumber <= 0) {
      System.out.println("FINISHED:" + savePath);
      File f1 = new File(savePath + ".tmp");
      File f2 = new File(savePath);
      // 如果目标文件已经存在，则尝试删除它
      // 最多尝试3次，间隔1秒钟。
      int times = 3;
      while (f2.exists() && times > 0) {
        if (f2.delete()) {
          break;
        }
        try {
          Thread.sleep(1000);
        } catch (InterruptedException e) {
          e.printStackTrace();
        }
        times--;
      }
      if (!f2.exists()) {
        if (!f1.renameTo(f2)) {
          System.out.println("改名失败!");
        }
      } else {
        System.out.println("目标文件存在，且无法删除,无法改名");
      }
    } else {
      int size = 0;
      HTTPDownloaderThread o = null;
      // 尝试查找一个可以分担的线程
      for (HTTPDownloaderThread thread : threadPool.values()) {
        if (thread.endPos - thread.curPos > size) {
          size = thread.endPos - thread.curPos;
          o = thread;
        }
      }
      if (size > MIN_BLOCK * 2) {
        if (o.isAlive()) {
          int endPos = o.endPos;
          int beginPos = o.endPos - ((o.endPos - o.curPos) / 2);
          o.endPos = beginPos;
          threadNumber++;
          threadPointer++;
          HTTPDownloaderThread thread = new HTTPDownloaderThread(this, threadPointer, beginPos, endPos);
          threadPool.put(threadPointer, thread);
          System.out.println("A Help Thread for " + o.id + " is started with:" + threadPointer);
          thread.start();
        }
      }
    }
  }

public HTTPDownloader() {
  }

/**
   * 下载
   * 
   * @param page 被下载的页面
   * @param savePath 保存的路径
   */
  public HTTPDownloader(String page, String savePath) {
    this(page, savePath, 10);
  }

/**
   * 下载
   * 
   * @param page 被下载的页面
   * @param savePath 保存的路径
   * @param threadNumber 线程数
   */
  public HTTPDownloader(String page, String savePath, int threadNumber) {
    this(page, page, savePath, 10, null);
  }

/**
   * 下载
   * 
   * @param page 被下载的页面
   * @param savePath 保存的路径
   * @param threadNumber 线程数
   * @param referer 来源
   */
  public HTTPDownloader(String page, String referer, String savePath, int threadNumber, String cookie) {
    this.page = page;
    this.savePath = savePath;
    this.threadNumber = threadNumber;
    this.referer = referer;
  }

public String getPage() {
    return page;
  }

public void setPage(String page) {
    this.page = page;
  }

public String getSavePath() {
    return savePath;
  }

public void setSavePath(String savePath) {
    this.savePath = savePath;
  }

public int getThreadNumber() {
    return threadNumber;
  }

public void setThreadNumber(int threadNumber) {
    this.threadNumber = threadNumber;
  }

public String getReferer() {
    return referer;
  }

public void setReferer(String referer) {
    this.referer = referer;
  }

public String getCookie() {
    return cookie;
  }

public void setCookie(String cookie) {
    this.cookie = cookie;
  }
}

/**
 * 下载线程
 * 
 */
class HTTPDownloaderThread extends Thread {
  HTTPDownloader manager;

int startPos;

int endPos;

int id;

int curPos;

int BUFFER_SIZE = 40960;

int readByte = 0;

HTTPDownloaderThread(HTTPDownloader manager, int id, int startPos, int endPos) {
    this.id = id;
    this.manager = manager;
    this.startPos = startPos;
    this.endPos = endPos;
  }

public void run() {
    System.out.println("线程" + id + "启动," + startPos + "-" + endPos);
    // 创建一个buff
    BufferedInputStream bis = null;
    RandomAccessFile fos = null;
    // 缓冲区大小
    byte[] buf = new byte[BUFFER_SIZE];
    boolean timeout = false;
    Socket socket = null;
    try {
      curPos = startPos;
      File file = new File(manager.getSavePath() + ".tmp");
      // 创建RandomAccessFile
      fos = new RandomAccessFile(file, "rw");
      // 从startPos开始
      fos.seek(startPos);
      int index = manager.getPage().indexOf("/", 8);
      String host = manager.getPage().substring(7, index);
      // System.out.println(host);
      socket = new Socket(host, 80);
      socket.setSoTimeout(30000);
      // 写入数据
      BufferedWriter wr = new BufferedWriter(new OutputStreamWriter(socket.getOutputStream(), "UTF-8"));
      StringBuilder b = new StringBuilder();
      b.append("GET " + manager.getPage().substring(index) + " HTTP/1.1\r\n");
      b.append("Host: " + host + "\r\n");
      b.append("Referer: " + (manager.getReferer() == null ? manager.getPage() : manager.getReferer()) + "\r\n");
      b.append("UserAgent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; flashget; \r\n");
      b.append("Range: bytes=" + startPos + "-" + endPos + "\r\n");
      b.append("\r\n");
      // System.out.println(b.toString());
      wr.write(b.toString());
      wr.flush();
      // 下面一段向根据文件写入数据,curPos为当前写入的未知,这里会判断是否小于endPos,
      // 如果超过endPos就代表该线程已经执行完毕
      bis = new BufferedInputStream(socket.getInputStream());
      // 读取直到换行
      int ch;
      boolean foundBR = false;
      while (true) {
        ch = bis.read();
        if (ch == 0xD) {
          ch = bis.read();
          if (ch == 0xA) {
            if (foundBR) {
              break;
            }
            foundBR = true;
          } else {
            foundBR = false;
          }
        } else {
          foundBR = false;
        }
      }
      int len = -1;
      while (curPos < endPos) {
        // System.out.println(id + "=" + (endPos - curPos));
        len = bis.read(buf, 0, BUFFER_SIZE);
        if (len == -1) {
          break;
        }
        fos.write(buf, 0, len);
        // System.out.println(id + "=Write OK!");
        curPos = curPos + len;
        if (curPos > endPos) {
          // 获取正确读取的字节数
          readByte += len - (curPos - endPos) + 1;
        } else {
          readByte += len;
        }
      }
      System.out.println("线程" + id + "已经下载完毕:" + readByte);
    } catch (Exception ex) {
      timeout = true;
    } finally {
      if (bis != null) {
        try {
          bis.close();
        } catch (Exception e) {
          System.out.println("关闭文件失败(1)!");
        }
      }
      if (fos != null) {
        try {
          fos.close();
        } catch (Exception e) {
          System.out.println("关闭文件失败(2)!");
        }
      }
      if (socket != null) {
        try {
          socket.close();
        } catch (Exception e) {
          System.out.println("关闭链接失败!");
        }
      }
    }
    if (timeout) {
      System.out.println(id + " timeout, restart...");
      new HTTPDownloaderThread(manager, id, curPos, endPos).start();
    } else {
      manager.finished(id);
    }
  }
}

来源：Heck's Blog
地址：https://www.heckjj.com/http-breakpoint-thread-download/
转载时须以链接形式注明作者和原始出处及本声明，否则将追究法律责任，谢谢配合！

Tags: http多线程断点续传 , 断点下载

阅读(2382) | 评论(0) | 引用(0)

< 2026 > < 4 >
日	一	二	三	四	五	六
			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30

Heck's Blog

java尝试HTTP多线程断点续传下载

关于博主

分类

日历

统计

最新日志

其他

选择模板

链接

最新评论

赞助商链接

归档