Get download speed in python

Question

OK, so I'm creating a little script that downloads video from a page and I can't get the download speed right. I know that download_speed = downloaded_file_size / (current_time - start_time) but i just can't get it right.

What I'm doing wrong?

'''
Python Trilulilu Downloader
Support for Video and Audio
Support for online view
Author: sharkyz of rstforums.com
'''

import re
from multiprocessing.pool import ThreadPool as Pool
import pyprind
import requests
import bs4
import time


url = 'http://www.trilulilu.ro/video-film/pitbull-ay-chico-lengua-afuera-1'


class commands(object):
    def __init__(self, httpadress):
        self.httpadress = httpadress

    def main_function(self):  # Acess, Find, Rewrite, Download
        pool = Pool(2)
        page = requests.get(self.httpadress)
        soup = bs4.BeautifulSoup(page.text, 'lxml')
        locatescript = soup.find(text=re.compile('swfobject.embedSWF'))
        keys = re.findall(r'"([^,]*?)":', locatescript)
        values = re.findall(r'(?<=:)(?:"(.*?)"|\d+)', locatescript)
        vovu = dict(zip(keys, values))

        video_test = ['http://fs{servers}.trilulilu.ro/stream.php?type=video&'
                     'source=site&hash={hashs}&username={userids}&key={keys}'
                     '&format=flv-vp6&sig=&exp='.format(servers=vovu['server'],
                                                         hashs=vovu['hash'],
                                                         userids=vovu['userid'],
                                                         keys=vovu['key']),
                     'http://fs{servers}.trilulilu.ro/stream.php?type=video&'
                     'source=site&hash={hashs}&username={userids}&key={keys}'
                     '&format=mp4-360p&sig=&exp='.format(servers=vovu['server'],
                                                         hashs=vovu['hash'],
                                                         userids=vovu['userid'],
                                                         keys=vovu['key'])]


        # Name the file
        page_title = soup.title.string # Title of trilulilu page
        title_chooser = page_title.split(' - ') # Split the title wherever '-' and create a list with elements




        # Search for the right link to download
        for link in video_test:
            respond = requests.get(link, stream=True)
            file_size = int(respond.headers.get('Content-Length', 0))
            if file_size > 1048576:
                # Check if the link was the mp4 or the flv format and choose name
                if 'mp4' in link:
                    local_name_file = '{} - {}.mp4'.format(title_chooser[0],title_chooser[1])
                elif 'flv' in link:
                    local_name_file = '{} - {}.flv'.format(title_chooser[0],title_chooser[1])
                else:
                    print('Download stopped, not recognizable format!')
                print('Downloading now...\nFile:{}\nSize:{}M'.format(local_name_file, round(file_size / 1000/ 1000, 2)))
                # Progress Bar
                bar = pyprind.ProgBar(file_size / 1024, monitor=True)
                file_downloaded_size = 0
                with open(local_name_file, 'wb') as f:
                    dl = 0
                    for chunk in respond.iter_content(chunk_size=1024):
                        if chunk:
                            dl += len(chunk)
                            start_time = time.mktime(time.localtime())
                            f.write(chunk)
                            end_time = time.mktime(time.localtime())
                            print(dl / (end_time / start_time))
                            f.flush()
                            bar.update()
                    print()
                    print(bar)





start = commands(url).main_function()
start

My results are:

And it keeps adding!

That should be `end_time - start_time`, surely? Also, please read http://stackoverflow.com/help/mcve — jonrsharpe, Sep 13 '14 at 22:53
If you type what you write (`-` instead of `/`) it would work better (but I think overall this is a to simple approach to get a usefull DL speed indicator). — eckes, Sep 13 '14 at 22:55
How i can make it end_time - start_time? Where should i put the end_time? — TaoJS, Sep 13 '14 at 22:56
I corrected 'print(dl / (end_time - start_time))' and now it gives me: ZeroDivisionError: division by zero. — TaoJS, Sep 13 '14 at 23:00
That's what happens when start_time equals end_time. Put `if end_time > start_time:` around the line to avoid the error — khelwood, Sep 13 '14 at 23:02
@khelwood Thank you for your suggestion, now it works but it's updating so sloow, while downloading a 13mb file, end_time was bigger than start_time just once, but the speed was true (2277376.0 which equals to 280Kilobytes). — TaoJS, Sep 13 '14 at 23:12

score 0 · Accepted Answer · answered Sep 13 '14 at 23:13

This seems to be the relevant part of your code:

for chunk in respond.iter_content(chunk_size=1024):
    if chunk:
        dl += len(chunk)
        start_time = time.mktime(time.localtime())
        f.write(chunk)
        end_time = time.mktime(time.localtime())
        print(dl / (end_time / start_time))

If so, you are calculating the time it takes to write a chunk of data to a file, which should be more or less constant. On the other hand the dl variable, which is the downloaded total size, doesn't stop growing. You should take start_time out of the loop and initialize it just at the beginning.

Get download speed in python

1 Answers1