2011-08-05 29 views
2

我使用urllib2來恢復下載器,大致基於this方法。我可以結束程序並重新啓動程序,並從停止的地方開始下載,下載最終以相同大小下載的文件,就像它一次下載完成一樣。當網絡重新連接時,Python urllib2恢復下載不起作用

但是,在禁用和重新啓用網絡時,我已經對它進行了測試,並且沒有正確下載。文件大小最終比文件應該長,文件無法正常工作。有沒有我錯過了,或者這可能是一個urllib2錯誤?

import urllib2 
    opener = urllib2.build_opener(); 

    self.count = 0 # Counts downloaded size. 
    self.downloading = True 
    while (not(self.success) and self.downloading): 
     try: 
      self.Err = "" 
      self._netfile = self.opener.open(self.url) 
      self.filesize = float(self._netfile.info()['Content-Length']) 

      if (os.path.exists(self.localfile) and os.path.isfile(self.localfile)): 
       self.count = os.path.getsize(self.localfile) 
      print self.count,"of",self.filesize,"downloaded." 
      if self.count >= self.filesize: 
       #already downloaded 
       self.downloading = False 
       self.success = True 
       self._netfile.close() 
       return 

      if (os.path.exists(self.localfile) and os.path.isfile(self.localfile)): 
       #File already exists, start where it left off: 
       #This seems to corrupt the file sometimes? 
       self._netfile.close() 
       req = urllib2.Request(self.url) 
       print "file downloading at byte: ",self.count 
       req.add_header("Range","bytes=%s-" % (self.count)) 
       self._netfile = self.opener.open(req) 
      if (self.downloading): #Don't do it if cancelled, downloading=false. 
       next = self._netfile.read(1024) 
       self._outfile = open(self.localfile,"ab") #to append binary 
       self._outfile.write(next) 
       self.readsize = desc(self.filesize) # get size mb/kb 
       self.count += 1024 
       while (len(next)>0 and self.downloading): 
        next = self._netfile.read(1024) 
        self._outfile.write(next) 
        self.count += len(next) 
       self.success = True 
     except IOError, e: 
      print e 
      self.Err=("Download error, retrying in a few seconds: "+str(e)) 
      try: 
       self._netfile.close() 
      except Exception: 
       pass 
      time.sleep(8) #Then repeat 
+0

已經有(幾乎)插入urllib替換它可以恢復:http://urlgrabber.baseurl.org/ –

+0

你有沒有嘗試過禁用/啓用網絡?它會自動重新下載嗎? – NoBugs

+0

我認爲它被內部的一些Linux軟件包管理使用,所以它應該被很好的測試 - 我很久以前就成功地使用它了。它甚至有一個重試次數設置等。 –

回答

1

我與self._netfile.close()在IO錯誤處理程序,這似乎有固定的加self._outfile.close()。我想這個錯誤是由於打開再次追加而未關閉而引起的。