Scrape econtalk episodes, and speed them up!

econtalk.py

import os
import re
import shlex
import requests
import subprocess

from multiprocessing import Pool
from bs4 import BeautifulSoup

reg = re.compile("http:\/\/files.*.mp3")

def get_mp3_info(url):
    req = requests.get(url.strip())
    data = req.text
    soup = BeautifulSoup(data, "html.parser")
    href = soup.find(href=reg).get('href')
    title = soup.title.string
    title = title.split('|')[0].strip()
    return href, title

def download_file(url):
    local_filename = url.split('/')[-1]
    r = requests.get(url, stream=True)
    with open(local_filename, 'wb') as f:
        for chunk in r.iter_content(chunk_size=1024): 
            if chunk:
                f.write(chunk)
    return local_filename

def speedup(filename, title):
    with open(os.devnull, 'w') as devnull:
        command_line = "ffmpeg -i {} -filter:a \"atempo=1.25\" -vn \"{}.mp3\"".format(filename, title)
        args = shlex.split(command_line)
        subprocess.call(args, stdout=devnull, stderr=devnull)

def process(url):
    print("Finding file for {}".format(url))
    loc, title = get_mp3_info(url)
    filename = download_file(loc)
    print("Shortening {}".format(title))
    speedup(filename, title)
    os.remove(filename)

def main():
    pool = Pool(10)
    with open('econtalk.txt') as f:
        episodes = f.readlines()
    
    pool.map(process, episodes)
    pool.close() 
    pool.join() 

if __name__ == '__main__':
    main()