Some licensing stuff
[jamaendo] / jamaendo / api.py
index 9524040..1e052a8 100644 (file)
-import urllib, threading, os, gzip, time, json, re
-_DUMP_URL = '''http://img.jamendo.com/data/dbdump_artistalbumtrack.xml.gz'''
-_DUMP = os.path.expanduser('''~/.cache/jamaendo/dbdump.xml.gz''')
-
-try:
-    os.makedirs(os.path.dirname(_DUMP))
-except OSError:
-    pass
-
-def has_dump():
-    return os.path.isfile(_DUMP)
-
-def _file_is_old(fil, old_age):
-    return os.path.getmtime(fil) < (time.time() - old_age)
-
-def _dump_is_old():
-    return not has_dump() or _file_is_old(_DUMP, 60*60*24) # 1 day
-
-def refresh_dump(complete_callback, progress_callback=None, force=False):
-    if force or _dump_is_old():
-        downloader = Downloader(complete_callback, progress_callback)
-        downloader.start()
-    else:
-        complete_callback()
-
-class Downloader(threading.Thread):
-    def __init__(self, complete_callback, progress_callback):
-        threading.Thread.__init__(self)
-        self.complete_callback = complete_callback
-        self.progress_callback = progress_callback
-
-    def actual_callback(self, numblocks, blocksize, filesize):
-        if self.progress_callback:
-            try:
-                percent = min((numblocks*blocksize*100)/filesize, 100)
-            except:
-                percent = 100
-            self.progress_callback(percent)
-
-    def run(self):
-        urllib.urlretrieve(_DUMP_URL, _DUMP, self.actual_callback)
-        self.complete_callback()
-
-def fast_iter(context, func):
-    for event, elem in context:
-        func(elem)
-        elem.clear()
-        while elem.getprevious() is not None:
-            del elem.getparent()[0]
-    del context
-
-from lxml import etree
-
-class Obj(object):
-    def __repr__(self):
-        def printable(v):
-            if isinstance(v, basestring):
-                return v.encode('utf-8')
+#!/usr/bin/env python
+#
+# This file is part of Jamaendo.
+# Copyright (c) 2010, Kristoffer Gronlund
+# All rights reserved.
+#
+# Redistribution and use in source and binary forms, with or without
+# modification, are permitted provided that the following conditions are met:
+#     * Redistributions of source code must retain the above copyright
+#       notice, this list of conditions and the following disclaimer.
+#     * Redistributions in binary form must reproduce the above copyright
+#       notice, this list of conditions and the following disclaimer in the
+#       documentation and/or other materials provided with the distribution.
+#     * Neither the name of Jamaendo nor the
+#       names of its contributors may be used to endorse or promote products
+#       derived from this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+# DISCLAIMED. IN NO EVENT SHALL <COPYRIGHT HOLDER> BE LIABLE FOR ANY
+# DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+# (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+# SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
+# An improved, structured jamendo API wrapper for the N900 with cacheing
+# Image / cover downloads.. and more?
+import urllib, threading, os, gzip, time, simplejson, re
+
+_CACHEDIR = None
+_COVERDIR = None
+_GET2 = '''http://api.jamendo.com/get2/'''
+_MP3URL = _GET2+'stream/track/redirect/?id=%d&streamencoding=mp31'
+_OGGURL = _GET2+'stream/track/redirect/?id=%d&streamencoding=ogg2'
+
+
+def set_cache_dir(cachedir):
+    global _CACHEDIR
+    global _COVERDIR
+    _CACHEDIR = cachedir
+    _COVERDIR = os.path.join(_CACHEDIR, 'covers')
+
+    try:
+        os.makedirs(_CACHEDIR)
+    except OSError:
+        pass
+
+    try:
+        os.makedirs(_COVERDIR)
+    except OSError:
+        pass
+
+# These classes can be partially constructed,
+# and if asked for a property they don't know,
+# makes a query internally to get the full story
+
+_ARTIST_FIELDS = ['id', 'name', 'image']
+_ALBUM_FIELDS = ['id', 'name', 'image', 'artist_name', 'artist_id', 'license_url']
+_TRACK_FIELDS = ['id', 'name', 'image', 'artist_name', 'album_name', 'album_id', 'numalbum', 'duration']
+_RADIO_FIELDS = ['id', 'name', 'idstr', 'image']
+
+class LazyQuery(object):
+    def set_from_json(self, json):
+        for key, value in json.iteritems():
+            if key == 'id':
+                assert(self.ID == int(value))
             else:
-                return str(v)
-        return "{%s}" % (", ".join("%s=%s"%(k.encode('utf-8'), printable(v)) \
-                             for k,v in self.__dict__.iteritems() if not k.startswith('_')))
+                if key.endswith('_id'):
+                    value = int(value)
+                setattr(self, key, value)
+
+    def load(self):
+        """Not automatic for now,
+        will have to do artist.load()
+
+        This is filled in further down
+        in the file
+        """
+        raise NotImplemented
+
+    def _needs_load(self):
+        return True
+
+    def _set_from(self, other):
+        raise NotImplemented
+
+    def _needs_load_impl(self, *attrs):
+        for attr in attrs:
+            if getattr(self, attr) is None:
+                return True
+        return False
+
+    def _set_from_impl(self, other, *attrs):
+        for attr in attrs:
+            self._set_if(other, attr)
+
+    def _set_if(self, other, attrname):
+        if getattr(self, attrname) is None and getattr(other, attrname) is not None:
+            setattr(self, attrname, getattr(other, attrname))
+
+    def __repr__(self):
+        try:
+            return u"%s(%s)"%(self.__class__.__name__,
+                              u", ".join(repr(v) for k,v in self.__dict__.iteritems() if not k.startswith('_')))
+        except UnicodeEncodeError:
+            import traceback
+            traceback.print_exc()
+            return u"%s(?)"%(self.__class__.__name__)
+
+class Artist(LazyQuery):
+    def __init__(self, ID, json=None):
+        self.ID = int(ID)
+        self.name = None
+        self.image = None
+        self.albums = None # None means not downloaded
+        if json:
+            self.set_from_json(json)
+
+    def _needs_load(self):
+        return self._needs_load_impl('name', 'albums')
+
+    def _set_from(self, other):
+        return self._set_from_impl(other, 'name', 'image', 'albums')
+
+class Album(LazyQuery):
+    def __init__(self, ID, json=None):
+        self.ID = int(ID)
+        self.name = None
+        self.image = None
+        self.artist_name = None
+        self.artist_id = None
+        self.license_url = None
+        self.tracks = None # None means not downloaded
+        if json:
+            self.set_from_json(json)
+
+    def _needs_load(self):
+        return self._needs_load_impl('name', 'image', 'artist_name', 'artist_id', 'license_url', 'tracks')
+
+    def _set_from(self, other):
+        return self._set_from_impl(other, 'name', 'image', 'artist_name', 'artist_id', 'license_url', 'tracks')
+
+class Track(LazyQuery):
+    def __init__(self, ID, json=None):
+        self.ID = int(ID)
+        self.name = None
+        self.image = None
+        self.artist_name = None
+        self.album_name = None
+        self.album_id = None
+        self.numalbum = None
+        self.duration = None
+        if json:
+            self.set_from_json(json)
+
+    def mp3_url(self):
+       return _MP3URL%(self.ID)
+
+    def ogg_url(self):
+       return _OGGURL%(self.ID)
+
+    def _needs_load(self):
+        return self._needs_load_impl('name', 'artist_name', 'album_name', 'album_id', 'numalbum', 'duration')
+
+    def _set_from(self, other):
+        return self._set_from_impl(other, 'name', 'image', 'artist_name', 'album_name', 'album_id', 'numalbum', 'duration')
+
+class Radio(LazyQuery):
+    def __init__(self, ID, json=None):
+        self.ID = int(ID)
+        self.name = None
+        self.idstr = None
+        self.image = None
+        if json:
+            self.set_from_json(json)
+
+    def _needs_load(self):
+        return self._needs_load_impl('name', 'idstr', 'image')
+
+    def _set_from(self, other):
+        return self._set_from_impl(other, 'name', 'idstr', 'image')
+
+
+_artists = {} # id -> Artist()
+_albums = {} # id -> Album()
+_tracks = {} # id -> Track()
+_radios = {} # id -> Radio()
+
+
+# cache sizes per session (TODO)
+_CACHED_ARTISTS = 100
+_CACHED_ALBUMS = 200
+_CACHED_TRACKS = 500
+_CACHED_RADIOS = 10
+
+# TODO: cache queries?
+
+class Query(object):
+    rate_limit = 1.1 # seconds between queries
+    last_query = time.time() - 1.5
+
+    @classmethod
+    def _ratelimit(cls):
+        now = time.time()
+        if now - cls.last_query < cls.rate_limit:
+            time.sleep(cls.rate_limit - (now - cls.last_query))
+        cls.last_query = now
+
+    def __init__(self):
+        pass
 
-class DB(object):
+    def _geturl(self, url):
+        print "*** %s" % (url)
+        Query._ratelimit()
+        f = urllib.urlopen(url)
+        ret = simplejson.load(f)
+        f.close()
+        return ret
+
+    def __str__(self):
+        return "#%s" % (self.__class__.__name__)
+
+    def execute(self):
+        raise NotImplemented
+
+class CoverCache(object):
+    """
+    cache and fetch covers
+    TODO: background thread that
+    fetches and returns covers,
+    asynchronously, LIFO
+    """
     def __init__(self):
-        self.fil = None
+        self._covers = {} # (albumid, size) -> file
+        coverdir = _COVERDIR if _COVERDIR else '/tmp'
+        if os.path.isdir(coverdir):
+            covermatch = re.compile(r'(\d+)\-(\d+)\.jpg')
+            for fil in os.listdir(coverdir):
+                fl = os.path.join(coverdir, fil)
+                m = covermatch.match(fil)
+                if m and os.path.isfile(fl):
+                    self._covers[(int(m.group(1)), int(m.group(2)))] = fl
+
+    def fetch_cover(self, albumid, size):
+        coverdir = _COVERDIR if _COVERDIR else '/tmp'
+        to = os.path.join(coverdir, '%d-%d.jpg'%(albumid, size))
+        if not os.path.isfile(to):
+            url = _GET2+'image/album/redirect/?id=%d&imagesize=%d'%(albumid, size)
+            urllib.urlretrieve(url, to)
+            self._covers[(albumid, size)] = to
+        return to
+
+    def get_cover(self, albumid, size):
+        cover = self._covers.get((albumid, size), None)
+        if not cover:
+            cover = self.fetch_cover(albumid, size)
+        return cover
+
+    def get_async(self, albumid, size, cb):
+        cover = self._covers.get((albumid, size), None)
+        if cover:
+            cb(cover)
+        else:
+            # TODO
+            cover = self.fetch_cover(albumid, size)
+            cb(cover)
+
+_cover_cache = CoverCache()
+
+def get_album_cover(albumid, size=200):
+    return _cover_cache.get_cover(albumid, size)
+
+def get_album_cover_async(cb, albumid, size=200):
+    _cover_cache.get_async(albumid, size, cb)
+
+class CustomQuery(Query):
+    def __init__(self, url):
+        Query.__init__(self)
+        self.url = url
+
+    def execute(self):
+        return self._geturl(self.url)
+
+    def __str__(self):
+        return self.url
+
+class GetQuery(Query):
+    queries = {
+        'artist' : {
+            'url' : _GET2+'+'.join(_ARTIST_FIELDS)+'/artist/json/?',
+            'params' : 'artist_id=%d',
+            'constructor' : Artist
+            },
+        'album' : {
+            'url' : _GET2+'+'.join(_ALBUM_FIELDS)+'/album/json/?',
+            'params' : 'album_id=%d',
+            'constructor' : Album
+            },
+        'albums' : {
+            'url' : _GET2+'+'.join(_ALBUM_FIELDS)+'/album/json/?',
+            'params' : 'artist_id=%d',
+            'constructor' : [Album]
+            },
+        'track' : {
+            'url' : _GET2+'+'.join(_TRACK_FIELDS)+'/track/json/track_album+album_artist?',
+            'params' : 'id=%d',
+            'constructor' : Track
+            },
+        'tracks' : {
+            'url' : _GET2+'+'.join(_TRACK_FIELDS)+'/track/json/track_album+album_artist?',
+            'params' : 'order=numalbum_asc&album_id=%d',
+            'constructor' : [Track]
+            },
+        'radio' : {
+            'url' : _GET2+'+'.join(_TRACK_FIELDS)+'/track/json/radio_track_inradioplaylist+track_album+album_artist/?',
+            'params' : 'order=random_asc&radio_id=%d',
+            'constructor' : [Track]
+            },
+        'favorite_albums' : {
+            'url' : _GET2+'+'.join(_ALBUM_FIELDS)+'/album/json/album_user_starred/?',
+            'params' : 'user_idstr=%s',
+            'constructor' : [Album]
+            },
+    #http://api.jamendo.com/get2/id+name+url+image+artist_name/album/jsonpretty/album_user_starred/?user_idstr=sylvinus&n=all
+    #q = SearchQuery('album', user_idstr=user)
+
+        }
+#http://api.jamendo.com/get2/id+name+image+artist_name+album_name+album_id+numalbum+duration/track/json/radio_track_inradioplaylist+track_album+album_artist/?order=numradio_asc&radio_id=283
+
+    def __init__(self, what, ID):
+        Query.__init__(self)
+        self.ID = ID
+        info = GetQuery.queries[what]
+        self.url = info['url']
+        self.params = info['params']
+        self.constructor = info['constructor']
+
+    def construct(self, data):
+        constructor = self.constructor
+        if isinstance(constructor, list):
+            constructor = constructor[0]
+        if isinstance(data, list):
+            return [constructor(int(x['id']), json=x) for x in data]
+        else:
+            return constructor(int(data['id']), json=data)
+
+    def execute(self):
+        js = self._geturl(self.url + self.params % (self.ID))
+        if not js:
+            return None
+        return self.construct(js)
+
+    def __str__(self):
+        return self.url + self.params % (self.ID)
+
+class SearchQuery(GetQuery):
+    def __init__(self, what, query=None, order=None, user=None, count=10):
+        GetQuery.__init__(self, what, None)
+        self.query = query
+        self.order = order
+        self.count = count
+        self.user = user
 
-    def connect(self):
-        self.fil = gzip.open(_DUMP)
+    def execute(self):
+        params = {}
+        if self.query:
+            params['searchquery'] = self.query
+        if self.order:
+            params['order'] = self.order
+        if self.count:
+            params['n'] = self.count
+        if self.user:
+            params['user_idstr'] = self.user
+        js = self._geturl(self.url +  urllib.urlencode(params))
+        if not js:
+            return None
+        return self.construct(js)
 
-    def close(self):
-        self.fil.close()
+    def __str__(self):
+        params = {'searchquery':self.query, 'order':self.order, 'n':self.count}
+        return self.url +  urllib.urlencode(params)
 
-    def make_obj(self, element):
-        if element.text is not None and element.text != "":
-            return element.text
+class JamendoAPIException(Exception):
+    def __init__(self, url):
+        Exception.__init__(self, url)
+
+def _update_cache(cache, new_items):
+    if not isinstance(new_items, list):
+        new_items = [new_items]
+    for item in new_items:
+        old = cache.get(item.ID)
+        if old:
+            old._set_from(item)
         else:
-            ret = Obj()
-            for child in element:
-                setattr(ret, child.tag, self.make_obj(child))
-            return ret
-
-    def artist_walker(self):
-        for event, element in etree.iterparse(self.fil, tag="artist"):
-            yield self.make_obj(element)
-            element.clear()
-            while element.getprevious() is not None:
-                del element.getparent()[0]
-        raise StopIteration
-
-    def search_artists(self, substr):
-        substr = substr.lower()
-        #return [dir(artist) for artist in self.artist_walker() if artist.name.find(substr) > -1]
-        artist = self.artist_walker().next()
-        print artist
-        print artist.name
-        return [artist.id]
+            cache[item.ID] = item
+
+def get_artist(artist_id):
+    """Returns: Artist"""
+    a = _artists.get(artist_id, None)
+    if not a:
+        q = GetQuery('artist', artist_id)
+        a = q.execute()
+        if not a:
+            raise JamendoAPIException(str(q))
+        _update_cache(_artists, a)
+        if isinstance(a, list):
+            a = a[0]
+    return a
+
+def get_albums(artist_id):
+    """Returns: [Album]"""
+    q = GetQuery('albums', artist_id)
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_artists, a)
+    return a
+
+def get_album(album_id):
+    """Returns: Album"""
+    a = _albums.get(album_id, None)
+    if not a:
+        q = GetQuery('album', album_id)
+        a = q.execute()
+        if not a:
+            raise JamendoAPIException(str(q))
+        _update_cache(_albums, a)
+        if isinstance(a, list):
+            a = a[0]
+    return a
+
+def get_tracks(album_id):
+    """Returns: [Track]"""
+    q = GetQuery('tracks', album_id)
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_tracks, a)
+    return a
+
+def get_track(track_id):
+    """Returns: Track"""
+    a = _tracks.get(track_id, None)
+    if not a:
+        q = GetQuery('track', track_id)
+        a = q.execute()
+        if not a:
+            raise JamendoAPIException(str(q))
+        _update_cache(_tracks, a)
+        if isinstance(a, list):
+            a = a[0]
+    return a
+
+def get_radio_tracks(radio_id):
+    """Returns: [Track]"""
+    q = GetQuery('radio', radio_id)
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_tracks, a)
+    return a
+
+def search_artists(query):
+    """Returns: [Artist]"""
+    q = SearchQuery('artist', query, 'searchweight_desc')
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_artists, a)
+    return a
+
+def search_albums(query):
+    """Returns: [Album]"""
+    q = SearchQuery('album', query, 'searchweight_desc')
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_albums, a)
+    return a
+
+def search_tracks(query):
+    """Returns: [Track]"""
+    q = SearchQuery('track', query=query, order='searchweight_desc')
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_tracks, a)
+    return a
+
+def albums_of_the_week():
+    """Returns: [Album]"""
+    q = SearchQuery('album', order='ratingweek_desc')
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_albums, a)
+    return a
+
+def new_releases():
+    """Returns: [Track] (playlist)"""
+    q = SearchQuery('track', order='releasedate_desc')
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_tracks, a)
+    return a
+
+def tracks_of_the_week():
+    """Returns: [Track] (playlist)"""
+    q = SearchQuery('track', order='ratingweek_desc')
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_tracks, a)
+    return a
+
+def get_radio(radio_id):
+    """Returns: Radio"""
+    q = CustomQuery(_GET2+"id+name+idstr+image/radio/json?id=%d"%(radio_id))
+    js = q.execute()
+    if not js:
+        raise JamendoAPIException(str(q))
+    if isinstance(js, list):
+        ks = js[0]
+    return Radio(radio_id, json=js)
+
+def starred_radios():
+    """Returns: [Radio]"""
+    q = CustomQuery(_GET2+"id+name+idstr+image/radio/json?order=starred_desc")
+    js = q.execute()
+    if not js:
+        raise JamendoAPIException(str(q))
+    return [Radio(int(radio['id']), json=radio) for radio in js]
+
+def favorite_albums(user):
+    """Returns: [Album]"""
+    q = SearchQuery('favorite_albums', user=user, count=20)
+    a = q.execute()
+    if not a:
+        raise JamendoAPIException(str(q))
+    _update_cache(_albums, a)
+    return a
+
+### Set loader functions for classes
+
+def _artist_loader(self):
+    if self._needs_load():
+        artist = get_artist(self.ID)
+        self._set_from(artist)
+Artist.load = _artist_loader
+
+def _album_loader(self):
+    if self._needs_load():
+        album = get_album(self.ID)
+        album.tracks = get_tracks(self.ID)
+        self._set_from(album)
+Album.load = _album_loader
+
+def _track_loader(self):
+    track = get_track(self.ID)
+    self._set_from(track)
+Track.load = _track_loader
+
+def _radio_loader(self):
+    radio = get_radio(self.ID)
+    self._set_from(radio)
+Radio.load = _radio_loader