Python youtube downloader by artist

youtube-downloader.py

#!/usr/bin/python
import sys
import wikipedia
import urllib2
from BeautifulSoup import BeautifulSoup
import os
from threading import Thread
import unicodedata
import commands
import json

bandname = <band name as string>
keywords = ['', 'Lyrics', 'Official']

try:
	wikipage = wikipedia.page(bandname)
except wikipedia.exceptions.WikipediaException:
	print "Page not found"

page = wikipage.html()

start = page.find('Studio albums', page.find('<th scope="row" class="navbox-group" style="background: #EEEEEE;">'))
end = page.find('</div>', start)

parsedpage = wikipage.html()[start:end]
soup = BeautifulSoup(parsedpage)

def conversong(album, bandname, song, song_index, song_title):
	os.system('cd "' + album + '" && ffmpeg -loglevel panic -i "$(find . -name "' + song_title + '"\\*)" -acodec libmp3lame "' + song_title + '.mp3"')
	os.system('cd "' + album + '" && eyeD3 --remove-all -a "' + bandname + '" -A "' + album + '" -t "' + song + '" -n ' + song_index + ' --add-image Cover.jpg:FRONT_COVER "' + song_title + '.mp3"')
	os.system('cd "' + album + '" && rm "' + song_title + '.mp4"')
	os.system('cd "' + album + '" && rm "' + song_title + '.mkv"')
	os.system('cd "' + album + '" && rm "' + song_title + '.f141.m4a"')
	os.system('cd "' + album + '" && rm "' + song_title + '.f140.m4a"')
	return
	
for link_tag in soup.findAll('a'):
	href = link_tag.get('href')
	
	response = urllib2.urlopen('http://en.wikipedia.org' + href)
	page = response.read()
	start = page.find('right; background-color: #eee">Length</th>')
	end = page.find("<b>", start)
	parsedpage = page[start:end]
	soup = BeautifulSoup(parsedpage)
	
	album = link_tag.text.replace('amp;','')
	
	year_find = page.find('bday dtstart published updated">')
	year_release = page[year_find + 32: year_find + 32 + 4]
	
	if year_release == ' dir':
		year_find = page.find(',',page.find('<td class="published">'));
		year_release = page[year_find+2: year_find+6]
	
	if not os.path.exists(year_release + " - " + album):
		os.makedirs(year_release + " - " + album)
	
	start_img = page.find('src="', page.find('class="image"><img alt="', page.find('<td colspan="2" style="text-align:center">')))
	end_img = page.find('"',start_img+5)
	image_url = page[start_img+5:end_img]
	
	image_url = image_url.replace('/thumb','')
	image_url = '/'.join(image_url.split('/')[:-1])
	
	os.system('cd "' + year_release + " - " + album + '" && wget http:' + image_url + ' -O Cover.jpg')
	
	songnum = 0
	
	for link_tag in soup.findAll('td'):
		song = link_tag.text.replace('amp;','')
		
		if song.find("\"") >= 0:
		
			songnum += 1
			song_index = str(songnum)
			
			if songnum < 10:
				song_index = '0' + song_index
			
			song = unicodedata.normalize('NFKD', song.split('"')[1]).encode('ascii','ignore')
			song_title = bandname + " - " + album + " - " + song_index + " - " + song.replace("/","|")
			
			if not os.path.isfile("./" + year_release + " - " + album + "/" + song_title + ".mp3"):
				
				highest_views = 0
				highest_keyword = ""
				
				for place in keywords:
					try:
						json_ = commands.getstatusoutput('youtube-dl -s --print-json "ytsearch:' + song + ' ' + bandname + ' ' + place + '"')
						views = int(json.loads(json_[1])['view_count'])
						
						try:
							file_size = int(json.loads(json_[1])['requested_formats'][0]['filesize']);
						except KeyError:
							print "something went wrong..."
						
						if(views > highest_views and file_size < 50000000):
							highest_views = views
							highest_keyword = place
					except ValueError:
						print "No JSON object could be decoded"
						
				os.system('cd "' + year_release + " - " + album + '" && youtube-dl --max-filesize 50m "ytsearch:' + song + ' ' + bandname + ' ' + highest_keyword + '" --prefer-ffmpeg -o "' + song_title + '.%(ext)s"')
				
				t = Thread(target=conversong, args=(year_release + " - " + album, bandname, song, song_index, song_title,))
				t.start()