實(shí)現(xiàn)簡(jiǎn)單的多線程下載，需要關(guān)注如下幾點(diǎn)：
1.文件的大小：可以從reponse header中提取，如“Content-Length:911”表示大小是911字節(jié)
2.任務(wù)拆分：指定各個(gè)線程下載的文件的哪一塊，可以通過request header中添加“Range: bytes=300-400”（表示下載300~400byte的內(nèi)容），注意可以請(qǐng)求的文件的range是[0, size-1]字節(jié)的。
3.下載文件的聚合：各個(gè)線程將自己下載的文件塊保存為臨時(shí)文件，所有線程都完成后，再將這些臨時(shí)文件按順序聚合寫入到最終的一個(gè)文件中。

實(shí)現(xiàn)代碼：
代碼如下:
#!/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