git.maemo.org Git - jamaendo/blob - jamaendo/api.py

   1 import urllib, threading, os, gzip, time, simplejson, re
   2 _DUMP_URL = '''http://img.jamendo.com/data/dbdump_artistalbumtrack.xml.gz'''
   3 _DUMP = os.path.expanduser('''~/.cache/jamaendo/dbdump.xml.gz''')
   4
   5 try:
   6     os.makedirs(os.path.dirname(_DUMP))
   7 except OSError:
   8     pass
   9
  10 def has_dump():
  11     return os.path.isfile(_DUMP)
  12
  13 def _file_is_old(fil, old_age):
  14     return os.path.getmtime(fil) < (time.time() - old_age)
  15
  16 def _dump_is_old():
  17     return not has_dump() or _file_is_old(_DUMP, 60*60*24) # 1 day
  18
  19 def refresh_dump(complete_callback, progress_callback=None, force=False):
  20     if force or _dump_is_old():
  21         downloader = Downloader(complete_callback, progress_callback)
  22         downloader.start()
  23     else:
  24         complete_callback()
  25
  26 class Downloader(threading.Thread):
  27     def __init__(self, complete_callback, progress_callback):
  28         threading.Thread.__init__(self)
  29         self.complete_callback = complete_callback
  30         self.progress_callback = progress_callback
  31
  32     def actual_callback(self, numblocks, blocksize, filesize):
  33         if self.progress_callback:
  34             try:
  35                 percent = min((numblocks*blocksize*100)/filesize, 100)
  36             except:
  37                 percent = 100
  38             self.progress_callback(percent)
  39
  40     def run(self):
  41         urllib.urlretrieve(_DUMP_URL, _DUMP, self.actual_callback)
  42         self.complete_callback()
  43
  44 def fast_iter(context, func):
  45     for event, elem in context:
  46         func(elem)
  47         elem.clear()
  48         while elem.getprevious() is not None:
  49             del elem.getparent()[0]
  50     del context
  51
  52 from lxml import etree
  53
  54 class Obj(object):
  55     def __repr__(self):
  56         def printable(v):
  57             if isinstance(v, basestring):
  58                 return v.encode('utf-8')
  59             else:
  60                 return str(v)
  61         return "{%s}" % (", ".join("%s=%s"%(k.encode('utf-8'), printable(v)) \
  62                              for k,v in self.__dict__.iteritems() if not k.startswith('_')))
  63
  64 class LocalDB(object):
  65     def __init__(self):
  66         self.fil = None
  67
  68     def connect(self):
  69         self.fil = gzip.open(_DUMP)
  70
  71     def close(self):
  72         self.fil.close()
  73
  74     def make_album_brief(self, element):
  75         ret = {}
  76         for info in element:
  77             if info.tag == 'id':
  78                 ret['id'] = int(info.text)
  79             elif info.tag == 'name':
  80                 ret['name'] = info.text
  81         return ret
  82
  83     def make_artist_obj(self, element):
  84         ret = {}
  85         for child in element:
  86             if child.tag == 'id':
  87                 ret['id'] = int(child.text)
  88             elif child.tag in ('name', 'image'):
  89                 ret[child.tag] = child.text
  90             elif child.tag == 'Albums':
  91                 ret['albums'] = [self.make_album_brief(a) for a in child]
  92         return ret
  93
  94     def make_track_obj(self, element):
  95         ret = {}
  96         for info in element:
  97             if info.tag == 'id':
  98                 _id = int(info.text)
  99                 ret['id'] = _id
 100                 ret['mp3'] = Query.track_mp3(_id)
 101                 ret['ogg'] = Query.track_ogg(_id)
 102             elif info.tag in ('name', 'numalbum'):
 103                 ret[info.tag] = info.text
 104         return ret
 105
 106     def make_album_obj(self, element):
 107         ret = {}
 108         artist = element.getparent().getparent()
 109         if artist is not None:
 110             for child in artist:
 111                 if child.tag == 'name':
 112                     ret['artist'] = child.text
 113                 elif child.tag == 'id':
 114                     ret['artist_id'] = int(child.text)
 115         for child in element:
 116             if child.tag == 'id':
 117                 ret['id'] = int(child.text)
 118             elif child.tag in ('name', 'image'):
 119                 if child.text:
 120                     ret[child.tag] = child.text
 121                 else:
 122                     ret[child.tag] = ""
 123             elif child.tag == 'Tracks':
 124                 ret['tracks'] = [self.make_track_obj(t) for t in child]
 125         return ret
 126
 127     def artist_walker(self, name_match):
 128         for event, element in etree.iterparse(self.fil, tag="artist"):
 129             name = element.xpath('./name')[0].text.lower()
 130             if name and name.find(name_match) > -1:
 131                 yield self.make_artist_obj(element)
 132             element.clear()
 133             while element.getprevious() is not None:
 134                 del element.getparent()[0]
 135         raise StopIteration
 136
 137     def album_walker(self, name_match):
 138         for event, element in etree.iterparse(self.fil, tag="album"):
 139             name = element.xpath('./name')[0].text
 140             if name and name.lower().find(name_match) > -1:
 141                 yield self.make_album_obj(element)
 142             element.clear()
 143             while element.getprevious() is not None:
 144                 del element.getparent()[0]
 145         raise StopIteration
 146
 147     def artistid_walker(self, artistids):
 148         for event, element in etree.iterparse(self.fil, tag="artist"):
 149             _id = element.xpath('./id')[0].text
 150             if _id and int(_id) in artistids:
 151                 yield self.make_artist_obj(element)
 152             element.clear()
 153             while element.getprevious() is not None:
 154                 del element.getparent()[0]
 155         raise StopIteration
 156
 157     def albumid_walker(self, albumids):
 158         for event, element in etree.iterparse(self.fil, tag="album"):
 159             _id = element.xpath('./id')[0].text
 160             if _id and (int(_id) in albumids):
 161                 yield self.make_album_obj(element)
 162             element.clear()
 163             while element.getprevious() is not None:
 164                 del element.getparent()[0]
 165         raise StopIteration
 166
 167     def search_artists(self, substr):
 168         substr = substr.lower()
 169         return (artist for artist in self.artist_walker(substr))
 170
 171     def search_albums(self, substr):
 172         substr = substr.lower()
 173         return (album for album in self.album_walker(substr))
 174
 175     def get_artists(self, artistids):
 176         return (artist for artist in self.artistid_walker(artistids))
 177
 178     def get_albums(self, albumids):
 179         return (album for album in self.albumid_walker(albumids))
 180
 181 _GET2 = '''http://api.jamendo.com/get2/'''
 182
 183 class Query(object):
 184     last_query = time.time()
 185     caching = True
 186     cache_time = 60*60*24
 187     rate_limit = 1.0 # max queries per second
 188
 189     def __init__(self, order,
 190                  select=['id', 'name', 'image', 'artist_name'],
 191                  request='album',
 192                  track=['track_album', 'album_artist'],
 193                  count=5):
 194         if request == 'track':
 195             self.url = "%s%s/%s/json/%s?n=%s&order=%s" % (_GET2, '+'.join(select), request, '+'.join(track), count, order)
 196         else:
 197             self.url = "%s%s/%s/json/?n=%s&order=%s" % (_GET2, '+'.join(select), request, count, order)
 198
 199
 200     def _ratelimit(self):
 201         now = time.time()
 202         if now - self.last_query < self.rate_limit:
 203             time.sleep(self.rate_limit - (now - self.last_query))
 204         self.last_query = now
 205
 206     def __call__(self):
 207         """ratelimited query"""
 208         self._ratelimit()
 209         f = urllib.urlopen(self.url)
 210         ret = simplejson.load(f)
 211         f.close()
 212         return ret
 213
 214     @staticmethod
 215     def album_cover(albumid, size=200):
 216         to = '~/.cache/jamaendo/cover-%d-%d.jpg'%(albumid, size)
 217         if not os.path.isfile(to):
 218             url = _GET2+'image/album/redirect/?id=%d&imagesize=%d'%(albumid, size)
 219             urllib.urlretrieve(url, to)
 220         return to
 221
 222     @staticmethod
 223     def track_ogg(trackid):
 224        return _GET2+ 'stream/track/redirect/?id=%d&streamencoding=ogg2'%(trackid)
 225
 226     @staticmethod
 227     def track_mp3(trackid):
 228        return _GET2+ 'stream/track/redirect/?id=%d&streamencoding=mp31'%(trackid)
 229
 230 class Queries(object):
 231     albums_this_week = Query(order='ratingweek_desc')
 232     albums_all_time = Query(order='ratingtotal_desc')
 233     albums_this_month = Query(order='ratingmonth_desc')
 234     albums_today = Query(order='ratingday_desc')
 235     playlists_all_time = Query(select=['id','name', 'user_idstr'], request='playlist', order='ratingtotal_desc')
 236     tracks_this_month = Query(select=['id', 'name',
 237                                       'stream',
 238                                       'album_name', 'artist_name',
 239                                       'album_id', 'artist_id'],
 240                               request='track',
 241                               order='ratingmonth_desc')