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 default_options = {
   8     'notifications' : 'true',
   9     # NOTIFICATIONS must be a string due to issues noted here:
  10     # http://bugs.python.org/issue974019
  11     # ConfigParser.getboolean fails when falling back to the default value
  12     # if the value type is bool.
  13     'youtube-dl' : '/usr/bin/youtube-dl',
  14     'default_icon' : '/usr/share/icons/gnome/48x48/mimetypes/gnome-mime-application-x-shockwave-flash.png',
  15     'youtube-dl_options' : '--no-progress --ignore-errors --continue --max-quality=22 -o "%(stitle)s---%(id)s.%(ext)s"'
  16 }
  17
  18 import ConfigParser # This module has been renamed to configparser in python  3.0
  19 import sys
  20 import os
  21
  22 CONFIG_FILE= os.path.join(os.path.expanduser('~'), '.i_like_pandora.config')
  23 config = ConfigParser.ConfigParser(default_options)
  24 loaded_files = config.read(CONFIG_FILE) # config.read returns an empty array if it fails.
  25 if len(loaded_files) == 0:
  26     print 'Can\'t find a configuration file at', CONFIG_FILE
  27     sys.exit()
  28 try:
  29     USER = config.get('settings', 'username')
  30     DIR = os.path.expanduser(config.get('settings', 'download_folder'))
  31     NOTIFICATIONS = config.getboolean('settings', 'notifications')
  32     YT_DL = config.get('settings', 'youtube-dl')
  33     DEFAULT_ICON = config.get('settings', 'default_icon')
  34     YT_OPT = default_options['youtube-dl_options']
  35 except:
  36     print 'There is a formatting error in the configuration file at', CONFIG_FILE
  37     sys.exit()
  38
  39 from BeautifulSoup import BeautifulSoup
  40 import urllib
  41 import urllib2
  42 import re
  43 import copy
  44 import shlex, subprocess
  45
  46 if NOTIFICATIONS:
  47     import pynotify
  48     import hashlib
  49     import tempfile
  50     import string
  51
  52 def fetch_stations(user):
  53     """ This takes a pandora username and returns the a list of the station tokens that the user is subscribed to. """
  54     stations = []
  55     page = urllib.urlopen('http://www.pandora.com/favorites/profile_tablerows_station.vm?webname=' + USER)
  56     page = BeautifulSoup(page)
  57     table = page.findAll('div', attrs={'class':'station_table_row'})
  58     for row in table:
  59         if row.find('a'):
  60             for attr, value in row.find('a').attrs:
  61                 if attr == 'href':
  62                     stations.append(value[10:])
  63     return stations
  64
  65 def fetch_tracks(stations):
  66     """ Takes a list of station tokens and returns a list of Title + Artist strings.
  67     """
  68     search_strings = []
  69     for station in stations:
  70         page = urllib.urlopen('http://www.pandora.com/favorites/station_tablerows_thumb_up.vm?token=' + station + '&sort_col=thumbsUpDate')
  71         page = BeautifulSoup(page)
  72         titles = []
  73         artists = []
  74         for span in page.findAll('span', attrs={'class':'track_title'}):
  75             for attr, value in span.attrs:
  76                 if attr == 'tracktitle':
  77                     titles.append(value)
  78         for anchor in page.findAll('a'):
  79             artists.append(anchor.string)
  80         if len(titles) == len(artists):
  81             i = 0
  82             for title in titles:
  83                 search_string = title + ' ' + artists[i]
  84                 search_strings.append(search_string)
  85                 i += 1
  86         else:
  87             # This would mean something strange has happened: there
  88             # aren't the same number of titles and artist names on a
  89             # station page.
  90             pass
  91     return search_strings
  92
  93 def search_youtube(search_strings):
  94     """ 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.
  95     """
  96     video_list = []
  97     for search_string in search_strings:
  98         search_url = 'http://youtube.com/results?search_query=' + urllib.quote_plus(search_string)
  99         page = urllib.urlopen(search_url)
 100         page = BeautifulSoup(page)
 101         result = page.find('div', attrs={'class':'video-main-content'})
 102         if result == None:
 103             print 'odd feedback for search, could not find div at ', search_url
 104             continue
 105         for attr, value in result.attrs:
 106             if attr == 'id' and len(value[19:]) == 11:
 107                 video_list.append(value[19:])
 108             elif attr == 'id':
 109                 print 'odd feedback for search', search_url, " : ", value[19:]
 110     return video_list
 111
 112
 113 def check_for_existing(video_list):
 114     """ Checks the download-folder for existing videos with same id and removes from video_list. """
 115     filelist = os.listdir(DIR)
 116     i = 0
 117     for video in copy.deepcopy(video_list):
 118         for files in filelist:
 119             if re.search(video,files):
 120                 del video_list[i]
 121                 i -= 1
 122         i += 1
 123     return video_list
 124
 125 def fetch_videos(video_list):
 126     """ Uses subprocess to trigger a download using youtube-dl of the list created earlier, and triggers notifications if enabled. """
 127     os.chdir(DIR)
 128     args = shlex.split(YT_DL + ' ' + YT_OPT)
 129     if NOTIFICATIONS: regex = re.compile("\[download\] Destination: (.+)")
 130     for video in video_list:
 131         if video:
 132             thread = subprocess.Popen(args + ["http://youtube.com/watch?v=" + video], stdout=subprocess.PIPE)
 133             output = thread.stdout.read()
 134             if NOTIFICATIONS:
 135                 video_file = regex.findall(output)
 136                 if len(video_file) == 0:
 137                     break
 138                 thumbnail = hashlib.md5('file://' + DIR + video_file[0]).hexdigest() + '.png'
 139                 # Two '/'s instead of three because the path is
 140                 # absolute; I'm not sure how this'd work on windows.
 141                 title, sep, vid_id = video_file[0].rpartition('---')
 142                 title = string.replace(title, '_', ' ')
 143                 thumbnail = os.path.join(os.path.expanduser('~/.thumbnails/normal'), thumbnail)
 144                 if not os.path.isfile(thumbnail):
 145                     opener = urllib2.build_opener()
 146                     try:
 147                         page = opener.open('http://img.youtube.com/vi/' + video + '/1.jpg')
 148                         thumb = page.read()
 149                         # The thumbnail really should be saved to
 150                         # ~/.thumbnails/normal (Thumbnail Managing
 151                         # Standard)
 152                         # [http://jens.triq.net/thumbnail-spec/]
 153                         # As others have had problems anyway
 154                         # (http://mail.gnome.org/archives/gnome-list/2010-October/msg00009.html)
 155                         # I decided not to bother at the moment.
 156                         temp = tempfile.NamedTemporaryFile(suffix='.jpg')
 157                         temp.write(thumb)
 158                         temp.flush()
 159                         note = pynotify.Notification(title, 'video downloaded', temp.name)
 160                     except:
 161                         note = pynotify.Notification(title, 'video downloaded', DEFAULT_ICON)
 162                 else:
 163                     # Generally, this will never happen, because the
 164                     # video is a new file.
 165                     note = pynotify.Notification(title, 'video downloaded', thumbnail)
 166                 note.show()
 167
 168 def main():
 169     stations = fetch_stations(USER)
 170     if len(stations) == 0:
 171         print 'Are you sure your pandora profile is public? Can\'t seem to find any stations listed with your account.'
 172     search_strings = fetch_tracks(stations)
 173     videos = search_youtube(search_strings)
 174     videos = check_for_existing(videos)
 175     fetch_videos(videos)
 176
 177 if __name__ ==  "__main__":
 178     main()