实现简单的多线程下载，需要关注如下几点：
1.文件的大小：可以从reponse header中提取，如“Content-Length:911”表示大小是911字节
2.任务拆分：指定各个线程下载的文件的哪一块，可以通过request header中添加“Range: bytes=300-400”（表示下载300~400byte的内容），注意可以请求的文件的range是[0, size-1]字节的。
3.下载文件的聚合：各个线程将自己下载的文件块保存为临时文件，所有线程都完成后，再将这些临时文件按顺序聚合写入到最终的一个文件中。

实现代码：

复制代码 代码如下:
#!/usr/bin/python
# -*- coding: utf-8 -*-
# filename: paxel.py
# FROM: http://jb51.net/code/view/58/full/
# Jay modified it a little and save for further potential usage.

\’\’\’It is a multi-thread downloading tool

    It was developed following axel.
        Author: volans
        E-mail: volansw [at] gmail.com
\’\’\’

import sys
import os
import time
import urllib
from threading import Thread

# in case you want to use http_proxy
local_proxies = {\’http\’: \’http://131.139.58.200:8080\’}

 
class AxelPython(Thread, urllib.FancyURLopener):
    \’\’\’Multi-thread downloading class.

        run() is a vitural method of Thread.
    \’\’\’
    def __init__(self, threadname, url, filename, ranges=0, proxies={}):
        Thread.__init__(self, name=threadname)
        urllib.FancyURLopener.__init__(self, proxies)
        self.name = threadname
        self.url = url
        self.filename = filename
        self.ranges = ranges
        self.downloaded = 0

    def run(self):
        \’\’\’vertual function in Thread\’\’\’
        try:
            self.downloaded = os.path.getsize(self.filename)
        except OSError:
            #print \’never downloaded\’
            self.downloaded = 0

        # rebuild start poind
        self.startpoint = self.ranges[0] + self.downloaded

        # This part is completed
        if self.startpoint >= self.ranges[1]:
            print \’Part %s has been downloaded over.\’ % self.filename
            return

        self.oneTimeSize = 16384  # 16kByte/time
        print \’task %s will download from %d to %d\’ % (self.name, self.startpoint, self.ranges[1])

        self.addheader(\”Range\”, \”bytes=%d-%d\” % (self.startpoint, self.ranges[1]))
        self.urlhandle = self.open(self.url)

        data = self.urlhandle.read(self.oneTimeSize)
        while data:
            filehandle = open(self.filename, \’ab+\’)
            filehandle.write(data)
            filehandle.close()

            self.downloaded += len(data)
            #print \”%s\” % (self.name)
            #progress = u\’\\r…\’

            data = self.urlhandle.read(self.oneTimeSize)

 
def GetUrlFileSize(url, proxies={}):
    urlHandler = urllib.urlopen(url, proxies=proxies)
    headers = urlHandler.info().headers
    length = 0
    for header in headers:
        if header.find(\’Length\’) != -1:
            length = header.split(\’:\’)[-1].strip()
            length = int(length)
    return length

 
def SpliteBlocks(totalsize, blocknumber):
    blocksize = totalsize / blocknumber
    ranges = []
    for i in range(0, blocknumber – 1):
        ranges.append((i * blocksize, i * blocksize + blocksize – 1))
    ranges.append((blocksize * (blocknumber – 1), totalsize – 1))

    return ranges

 
def islive(tasks):
    for task in tasks:
        if task.isAlive():
            return True
    return False

 
def paxel(url, output, blocks=6, proxies=local_proxies):
    \’\’\’ paxel
    \’\’\’
    size = GetUrlFileSize(url, proxies)
    ranges = SpliteBlocks(size, blocks)

    threadname = [\”thread_%d\” % i for i in range(0, blocks)]
    filename = [\”tmpfile_%d\” % i for i in range(0, blocks)]

    tasks = []
    for i in range(0, blocks):
        task = AxelPython(threadname[i], url, filename[i], ranges[i])
        task.setDaemon(True)
        task.start()
        tasks.append(task)

    time.sleep(2)
    while islive(tasks):
        downloaded = sum([task.downloaded for task in tasks])
        process = downloaded / float(size) * 100
        show = u\’\\rFilesize:%d Downloaded:%d Completed:%.2f%%\’ % (size, downloaded, process)
        sys.stdout.write(show)
        sys.stdout.flush()
        time.sleep(0.5)

    filehandle = open(output, \’wb+\’)
    for i in filename:
        f = open(i, \’rb\’)
        filehandle.write(f.read())
        f.close()
        try:
            os.remove(i)
            pass
        except:
            pass

    filehandle.close()

if __name__ == \’__main__\’:
    url = \’http://dldir1.qq.com/qqfile/QQforMac/QQ_V3.1.1.dmg\’
    output = \’download.file\’
    paxel(url, output, blocks=4, proxies={})