likes_pandora.py

   1 __author__ = ("Dylan Lloyd <dylan@psu.edu>")
   2 __license__ = "BSD"
   3
   4 # SETTINGS
   5
   6 USER = 'alphabethos'
   7 DIR = '/home/dylan/pandora/'
   8 YT_DL = '/usr/bin/youtube-dl' # Path to youtube-dl
   9 NOTIFICATIONS = True # False
  10 DEFAULT_ICON ='/usr/share/icons/gnome/48x48/mimetypes/gnome-mime-application-x-shockwave-flash.png'
  11 YT_OPT = '--no-progress --ignore-errors --continue --max-quality=22 -o "%(stitle)s---%(id)s.%(ext)s"'
  12 # END OF SETTINGS
  13
  14 from BeautifulSoup import BeautifulSoup
  15 import urllib
  16 import urllib2
  17 import pynotify
  18 import tempfile
  19 import string
  20 if NOTIFICATIONS:
  21     import hashlib
  22     import os
  23     import shlex, subprocess
  24     import re
  25
  26 def fetch_stations(user):
  27     """ This takes a pandora username and returns the a list of the station tokens that the user is subscribed to. """
  28     stations = []
  29     page = urllib.urlopen('http://www.pandora.com/favorites/profile_tablerows_station.vm?webname=' + USER)
  30     page = BeautifulSoup(page)
  31     table = page.findAll('div', attrs={'class':'station_table_row'})
  32     for row in table:
  33         if row.find('a'):
  34             for attr, value in row.find('a').attrs:
  35                 if attr == 'href':
  36                     stations.append(value[10:])
  37     return stations
  38
  39 def fetch_tracks(stations):
  40     """ Takes a list of station tokens and returns a list of youtube search urls.
  41         What this should really do is just return the Title + Artist strings.
  42     """
  43     search_urls = []
  44     for station in stations:
  45         page = urllib.urlopen('http://www.pandora.com/favorites/station_tablerows_thumb_up.vm?token=' + station + '&sort_col=thumbsUpDate')
  46         page = BeautifulSoup(page)
  47         titles = []
  48         artists = []
  49         for span in page.findAll('span', attrs={'class':'track_title'}):
  50             for attr, value in span.attrs:
  51                 if attr == 'tracktitle':
  52                     titles.append(value)
  53         for anchor in page.findAll('a'):
  54             artists.append(anchor.string)
  55         if len(titles) == len(artists):
  56             i = 0
  57             for title in titles:
  58                 search_url = 'http://youtube.com/results?search_query=' + urllib.quote_plus(title + ' ' + artists[i])
  59                 search_urls.append(search_url)
  60                 i += 1
  61         else:
  62            pass  ## ERROR
  63     return search_urls
  64
  65 def fetch_search_video_ids(search_urls):
  66     """ This takes a list of youtube search urls and tries to find the first result. It returns a list of youtube video ids.
  67         It really should take a list of ids instead.
  68     """
  69     video_list = []
  70     for url in search_urls:
  71         page = urllib.urlopen(url)
  72         page = BeautifulSoup(page)
  73         result = page.find('div', attrs={'class':'video-main-content'})
  74         for attr, value in result.attrs:
  75             if attr == 'id' and len(value[19:]) == 11:
  76                 video_list.append(value[19:])
  77             elif attr == 'id':
  78                 print 'odd feedback for url', url, " : ", value[19:]
  79     return video_list
  80
  81
  82 def check_for_existing():
  83     """ Checks the download-folder for existing videos with same id and removes from videolist. """
  84     videolist = get_video_ids()
  85     filelist = os.listdir(DIR)
  86     for video in copy.deepcopy(videolist):
  87         for files in filelist:
  88             if re.search(video,files):
  89                 del videolist[video]
  90     return videolist
  91
  92 def fetch_videos(videolist):
  93     """ Uses subprocess to trigger a download using youtube-dl of the list created earlier, and triggers notifications if enabled. """
  94     os.chdir(DIR)
  95     args = shlex.split(YT_DL + ' ' + YT_OPT)
  96     if NOTIFICATIONS: regex = re.compile("\[download\] Destination: (.+)")
  97     for item in videolist:
  98         if item:
  99             thread = subprocess.Popen(args + [item], stdout=subprocess.PIPE)
 100             output = thread.stdout.read()
 101             if NOTIFICATIONS:
 102                 video_file = regex.findall(output)
 103                 if len(video_file) == 0:
 104                     break
 105                 thumbnail = hashlib.md5('file://' + DIR + video_file[0]).hexdigest() + '.png'
 106                 # Two '/'s instead of three because the path is
 107                 # absolute; I'm not sure how this'd work on windows.
 108                 title, sep, vid_id = video_file[0].rpartition('---')
 109                 title = string.replace(title, '_', ' ')
 110                 thumbnail = os.path.join(os.path.expanduser('~/.thumbnails/normal'), thumbnail)
 111                 if not os.path.isfile(thumbnail):
 112                     opener = urllib2.build_opener()
 113                     try:
 114                         page = opener.open('http://img.youtube.com/vi/' + item + '/1.jpg')
 115                         thumb = page.read()
 116                         # The thumbnail really should be saved to
 117                         # ~/.thumbnails/normal (Thumbnail Managing
 118                         # Standard)
 119                         # [http://jens.triq.net/thumbnail-spec/]
 120                         # As others have had problems anyway
 121                         # (http://mail.gnome.org/archives/gnome-list/2010-October/msg00009.html)
 122                         # I decided not to bother at the moment.
 123                         temp = tempfile.NamedTemporaryFile(suffix='.jpg')
 124                         temp.write(thumb)
 125                         temp.flush()
 126                         note = pynotify.Notification(title, 'video downloaded', temp.name)
 127                     except:
 128                         note = pynotify.Notification(title, 'video downloaded', DEFAULT_ICON)
 129                 else:
 130                     # Generally, this will never happen, because the
 131                     # video is a new file.
 132                     note = pynotify.Notification(title, 'video downloaded', thumbnail)
 133                 note.show()
 134
 135
 136 def main():
 137     stations = fetch_stations(USER)
 138     search_urls = fetch_tracks(stations)
 139     video_list = fetch_search_video_ids(search_urls)
 140     fetch_videos(video_list)
 141
 142 if __name__ ==  "__main__":
 143     main()