此处荒无人烟

#!/usr/bin/python
#coding:utf-8
import re
import urllib
import urlparse
import httplib
import time
import sys
import os
from threading import Thread

songtype = {'7d99bb4c7bd4602c342e2bb826ee8777':'.wma',
           '25e4f07f5123910814d9b8f3958385ba':'.Wma',
            '51bbd020689d1ce1c845a484995c0cce':'.WMA',
            'b3a7a4e64bcd8aabe4cabe0e55b57af5':'.mp3',
        'd82029f73bcaf052be8930f6f4247184':'.MP3',
         '5fd91d90d9618feca4740ac1f2e7948f':'.Mp3'}

outputdir = '/home/pxiaohai/Music/'

class MultiThreadDown(Thread,urllib.FancyURLopener):

    def __init__(self,threadname,url,filename,ranges):
        Thread.__init__(self,name=threadname)
        urllib.FancyURLopener.__init__(self)
        self.name = threadname
        self.url = url
        self.filename = filename
        self.ranges = ranges
        self.downloaded = 0

    def run(self):

        try:
            self.downloaded = os.path.getsize(self.filename)
            print self.filename
        except OSError:
            self.downloaded = 0

        self.startpoint = self.ranges[0] + self.downloaded

        if self.startpoint >= self.ranges[1]:
            print 'part %s has been downloaded over.' % self.filename
            return
        self.oneTimeSize = 16384

        print 'task %s will download from %d to %d' %
                     (self.name,self.startpoint,self.ranges[1])
        self.addheader("Range","bytes=%d-%d" %
                          (self.ranges[0],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)
            data = self.urlhandle.read(self.oneTimeSize)

class DownSong():
    def __init__(self,ms):
        self.ms = ms

    def down(self):
        global outputdir
        global songtype

        for song in ms:
            suburl = 'http://www.songtaste.com/song/%s/' % (song[0])
            songname = song[1]
            m = re.compile('playmedia1\((.+?)\)')
            f = urllib.urlopen(suburl)
            data = f.read()
            s = []
            s = m.findall(data)[-1].replace('\'','').replace('"','').replace(' ','').split(',')
            print s
            print s[1]
            songname = songname + songtype.get(s[5])
            songurl = s[6] + s[2] + songtype.get(s[5])

            try:
                url1=urllib.urlopen(songurl)
            except:
                show = u'\r %s can not be downloaded' % (songname)
                print show
                continue

            blocks = 4
            filesize = self.GetUrlFileSize(songurl)
            ranges = self.SpliteBlocks(filesize,blocks)

            print ranges

            threadname = ["thread_%d" % i for i in range(0,blocks)]
            filename =  ["tmpfiles_%d" % i for i in range(0,blocks)]

            tasks = []
            for i in range(0,blocks):
                task = MultiThreadDown(threadname[i],songurl,filename[i],ranges[i])
                task.setDaemon(True)
                task.start()
                tasks.append(task)
            time.sleep(1)
            songname = songname.decode('gb18030','replace').strip()
            songname = ''.join(songname.split())
            show = u'\r %s is starting download' % (songname)
            print show

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

            show = u'\r %s  completed' % (songname)
            print show
            songname = outputdir + songname
            filehandle = open(songname,'wb+')
            for i in filename:
                f = open(i,'rb')
                filehandle.write(f.read())
                f.close()

                try:
                    os.remove(i)
                    print ""
                except:
                    pass

            filehandle.close()

    def GetUrlFileSize(self,url):
        urlHandler = urllib.urlopen(url)
        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(self,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(self,tasks):
        for task in tasks:
            if task.isAlive():
                return True
        return False

if __name__ == '__main__':
#    if len(sys.argv) < 1:
#        print u'please input the url'
#    else:
#        if not re.match("^https?://[^ ]+",sys.argv[0]):
#            print u'please input the url like
                       "http://www.songtaste.com/music/catsong/cat2/"'

    m = re.compile(r'(.+?)')
    #
    url = 'http://www.songtaste.com/music/catsong/cat2/'
    f = urllib.urlopen(url)
    data = f.read()

    ms = []
    ms = m.findall(data)
    ds = DownSong(ms)
    ds.down()

看一篇博文http://hi.baidu.com/volansw/blog/item/38b13ebf3da4080f18d81f71.html 自己也做一个多线程下载的例子

然后下载例子中给的那本pdf 的书 完全没问题

然后自己琢磨着做了一个 下载songtaste 上音乐的的多线程的例子

但是发现下载下来的 文件不对

具体的出错是这样的 比如 总共文件大小为 4

正常的结果应该是tmpfile_0 为 1 tmpfile_1 为1 tmpfile_2 为 1 tmpfile_3 为1

但现在的结果是

正常的结果应该是tmpfile_0 为 1 tmpfile_1 为2 tmpfile_2 为 3 tmpfile_3 为4

为什么会这样呢

于是我找来上述连接中给的例子 把url 换成 sogntaset 上的音乐url

http://224.cachefile28.rayfile.com/59f9/zh-cn/download/c4f83b63ace9a140cacf6c7273417f0a/preview.mp3

试了一下 发现用那个的例子 也会出现和我一样的错误

纠结了一个下午 和一个晚上 实在想不出是什么原因