likes_pandora.py

   1 #!/usr/bin/env python
   2 # -*- coding: utf-8 -*-
   3
   4 __author__ = ("Dylan Lloyd <dylan@psu.edu>")
   5 __license__ = "BSD"
   6
   7 # SETTINGS
   8
   9 USER = 'alphabethos' # pandora account name http://pandora.com/people/<USER>
  10 DIR = '/home/dylan/pandora/' # where to download the videos - will not be automatically created
  11 YT_DL = '/usr/bin/youtube-dl' # Path to youtube-dl
  12 NOTIFICATIONS = True # False
  13 DEFAULT_ICON ='/usr/share/icons/gnome/48x48/mimetypes/gnome-mime-application-x-shockwave-flash.png' # for notifications
  14 YT_OPT = '--no-progress --ignore-errors --continue --max-quality=22 -o "%(stitle)s---%(id)s.%(ext)s"'
  15 # END OF SETTINGS
  16
  17 from BeautifulSoup import BeautifulSoup
  18 import urllib
  19 import urllib2
  20 import os
  21 import re
  22 import copy
  23 import shlex, subprocess
  24
  25 if NOTIFICATIONS:
  26     import pynotify
  27     import hashlib
  28     import tempfile
  29     import string
  30
  31 def fetch_stations(user):
  32     """ This takes a pandora username and returns the a list of the station tokens that the user is subscribed to. """
  33     stations = []
  34     page = urllib.urlopen('http://www.pandora.com/favorites/profile_tablerows_station.vm?webname=' + USER)
  35     page = BeautifulSoup(page)
  36     table = page.findAll('div', attrs={'class':'station_table_row'})
  37     for row in table:
  38         if row.find('a'):
  39             for attr, value in row.find('a').attrs:
  40                 if attr == 'href':
  41                     stations.append(value[10:])
  42     return stations
  43
  44 def fetch_tracks(stations):
  45     """ Takes a list of station tokens and returns a list of Title + Artist strings.
  46     """
  47     search_strings = []
  48     for station in stations:
  49         page = urllib.urlopen('http://www.pandora.com/favorites/station_tablerows_thumb_up.vm?token=' + station + '&sort_col=thumbsUpDate')
  50         page = BeautifulSoup(page)
  51         titles = []
  52         artists = []
  53         for span in page.findAll('span', attrs={'class':'track_title'}):
  54             for attr, value in span.attrs:
  55                 if attr == 'tracktitle':
  56                     titles.append(value)
  57         for anchor in page.findAll('a'):
  58             artists.append(anchor.string)
  59         if len(titles) == len(artists):
  60             i = 0
  61             for title in titles:
  62                 search_string = title + ' ' + artists[i]
  63                 search_strings.append(search_string)
  64                 i += 1
  65         else:
  66            pass  ## ERROR
  67     return search_strings
  68
  69 def search_youtube(search_strings):
  70     """ This takes a list of search strings and tries to find the first result. It returns a list of the youtube video ids of those results.
  71     """
  72     video_list = []
  73     for search_string in search_strings:
  74         search_url = 'http://youtube.com/results?search_query=' + urllib.quote_plus(search_string)
  75         page = urllib.urlopen(search_url)
  76         page = BeautifulSoup(page)
  77         result = page.find('div', attrs={'class':'video-main-content'})
  78         if result == None:
  79             print 'odd feedback for search, could not find div at ', search_url
  80             continue
  81         for attr, value in result.attrs:
  82             if attr == 'id' and len(value[19:]) == 11:
  83                 video_list.append(value[19:])
  84             elif attr == 'id':
  85                 print 'odd feedback for search', search_url, " : ", value[19:]
  86     return video_list
  87
  88
  89 def check_for_existing(video_list):
  90     """ Checks the download-folder for existing videos with same id and removes from video_list. """
  91     filelist = os.listdir(DIR)
  92     i = 0
  93     for video in copy.deepcopy(video_list):
  94         for files in filelist:
  95             if re.search(video,files):
  96                 del video_list[i]
  97                 i -= 1
  98         i += 1
  99     return video_list
 100
 101 def fetch_videos(video_list):
 102     """ Uses subprocess to trigger a download using youtube-dl of the list created earlier, and triggers notifications if enabled. """
 103     os.chdir(DIR)
 104     args = shlex.split(YT_DL + ' ' + YT_OPT)
 105     if NOTIFICATIONS: regex = re.compile("\[download\] Destination: (.+)")
 106     for video in video_list:
 107         if video:
 108             thread = subprocess.Popen(args + ["http://youtube.com/watch?v=" + video], stdout=subprocess.PIPE)
 109             output = thread.stdout.read()
 110             if NOTIFICATIONS:
 111                 video_file = regex.findall(output)
 112                 if len(video_file) == 0:
 113                     break
 114                 thumbnail = hashlib.md5('file://' + DIR + video_file[0]).hexdigest() + '.png'
 115                 # Two '/'s instead of three because the path is
 116                 # absolute; I'm not sure how this'd work on windows.
 117                 title, sep, vid_id = video_file[0].rpartition('---')
 118                 title = string.replace(title, '_', ' ')
 119                 thumbnail = os.path.join(os.path.expanduser('~/.thumbnails/normal'), thumbnail)
 120                 if not os.path.isfile(thumbnail):
 121                     opener = urllib2.build_opener()
 122                     try:
 123                         page = opener.open('http://img.youtube.com/vi/' + video + '/1.jpg')
 124                         thumb = page.read()
 125                         # The thumbnail really should be saved to
 126                         # ~/.thumbnails/normal (Thumbnail Managing
 127                         # Standard)
 128                         # [http://jens.triq.net/thumbnail-spec/]
 129                         # As others have had problems anyway
 130                         # (http://mail.gnome.org/archives/gnome-list/2010-October/msg00009.html)
 131                         # I decided not to bother at the moment.
 132                         temp = tempfile.NamedTemporaryFile(suffix='.jpg')
 133                         temp.write(thumb)
 134                         temp.flush()
 135                         note = pynotify.Notification(title, 'video downloaded', temp.name)
 136                     except:
 137                         note = pynotify.Notification(title, 'video downloaded', DEFAULT_ICON)
 138                 else:
 139                     # Generally, this will never happen, because the
 140                     # video is a new file.
 141                     note = pynotify.Notification(title, 'video downloaded', thumbnail)
 142                 note.show()
 143
 144
 145 def main():
 146     stations = fetch_stations(USER)
 147     if len(stations) == 0:
 148         print 'are you sure your pandora profile is public?'
 149     search_strings = fetch_tracks(stations)
 150     videos = search_youtube(search_strings)
 151     videos = check_for_existing(videos)
 152     fetch_videos(videos)
 153
 154 if __name__ ==  "__main__":
 155     main()