[youtube] Fix extraction.

[youtube-dl] / youtube_dl / extractor / audiomack.py

diff --git a/youtube_dl/extractor/audiomack.py b/youtube_dl/extractor/audiomack.py
index 73cde78d319e71a13c3b53ad096614001c21e689..cc7771354c10b39778f0b2dcad55052db42066b3 100644 (file)
--- a/youtube_dl/extractor/audiomack.py
+++ b/youtube_dl/extractor/audiomack.py
@@ -1,11 +1,16 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import itertools
+import time
+
 from .common import InfoExtractor
 from .soundcloud import SoundcloudIE
-from ..utils import ExtractorError
-
-import time
+from ..compat import compat_str
+from ..utils import (
+    ExtractorError,
+    url_basename,
+)
 
 
 class AudiomackIE(InfoExtractor):
@@ -19,36 +24,25 @@ class AudiomackIE(InfoExtractor):
             {
                 'id': '310086',
                 'ext': 'mp3',
-                'artist': 'Roosh Williams',
+                'uploader': 'Roosh Williams',
                 'title': 'Extraordinary'
             }
         },
         # audiomack wrapper around soundcloud song
         {
             'add_ie': ['Soundcloud'],
-            'url': 'http://www.audiomack.com/song/xclusiveszone/take-kare',
+            'url': 'http://www.audiomack.com/song/hip-hop-daily/black-mamba-freestyle',
             'info_dict': {
-                'id': '172419696',
+                'id': '258901379',
                 'ext': 'mp3',
-                'description': 'md5:1fc3272ed7a635cce5be1568c2822997',
-                'title': 'Young Thug ft Lil Wayne - Take Kare',
-                'uploader': 'Young Thug World',
-                'upload_date': '20141016',
+                'description': 'mamba day freestyle for the legend Kobe Bryant ',
+                'title': 'Black Mamba Freestyle [Prod. By Danny Wolf]',
+                'uploader': 'ILOVEMAKONNEN',
+                'upload_date': '20160414',
             }
         },
     ]
 
-    @staticmethod
-    def create_song_dictionary(api_response, album_url_tag, track_no=0):
-        # All keys are the same in audiomack api and InfoExtractor format
-        entry = {key: api_response[key] for key in ['title', 'artist', 'id', 'url'] if key in api_response}
-        # Fudge values in the face of missing metadata
-        if 'id' not in entry:
-            entry['id'] = track_no
-        if 'title' not in entry:
-            entry['title'] = album_url_tag
-        return entry
-
     def _real_extract(self, url):
         # URLs end with [uploader name]/[uploader title]
         # this title is whatever the user types in, and is rarely
@@ -63,14 +57,19 @@ class AudiomackIE(InfoExtractor):
 
         # API is inconsistent with errors
         if 'url' not in api_response or not api_response['url'] or 'error' in api_response:
-            raise ExtractorError('Invalid url %s', url)
+            raise ExtractorError('Invalid url %s' % url)
 
         # Audiomack wraps a lot of soundcloud tracks in their branded wrapper
         # if so, pass the work off to the soundcloud extractor
         if SoundcloudIE.suitable(api_response['url']):
-            return {'_type': 'url', 'url': api_response['url'], 'ie_key': 'Soundcloud'}
+            return self.url_result(api_response['url'], SoundcloudIE.ie_key())
 
-        return self.create_song_dictionary(api_response, album_url_tag)
+        return {
+            'id': compat_str(api_response.get('id', album_url_tag)),
+            'uploader': api_response.get('artist'),
+            'title': api_response.get('title'),
+            'url': api_response['url'],
+        }
 
 
 class AudiomackAlbumIE(InfoExtractor):
@@ -90,7 +89,22 @@ class AudiomackAlbumIE(InfoExtractor):
         # Album playlist ripped from fakeshoredrive with no metadata
         {
             'url': 'http://www.audiomack.com/album/fakeshoredrive/ppp-pistol-p-project',
-            'playlist_count': 10
+            'info_dict': {
+                'title': 'PPP (Pistol P Project)',
+                'id': '837572',
+            },
+            'playlist': [{
+                'info_dict': {
+                    'title': 'PPP (Pistol P Project) - 9. Heaven or Hell (CHIMACA) ft Zuse (prod by DJ FU)',
+                    'id': '837577',
+                    'ext': 'mp3',
+                    'uploader': 'Lil Herb a.k.a. G Herbo',
+                }
+            }],
+            'params': {
+                'playliststart': 9,
+                'playlistend': 9,
+            }
         }
     ]
 
@@ -102,11 +116,12 @@ class AudiomackAlbumIE(InfoExtractor):
         result = {'_type': 'playlist', 'entries': []}
         # There is no one endpoint for album metadata - instead it is included/repeated in each song's metadata
         # Therefore we don't know how many songs the album has and must infi-loop until failure
-        track_no = 0
-        while True:
+        for track_no in itertools.count():
             # Get song's metadata
-            api_response = self._download_json('http://www.audiomack.com/api/music/url/album/%s/%d?extended=1&_=%d'
-                                               % (album_url_tag, track_no, time.time()), album_url_tag)
+            api_response = self._download_json(
+                'http://www.audiomack.com/api/music/url/album/%s/%d?extended=1&_=%d'
+                % (album_url_tag, track_no, time.time()), album_url_tag,
+                note='Querying song information (%d)' % (track_no + 1))
 
             # Total failure, only occurs when url is totally wrong
             # Won't happen in middle of valid playlist (next case)
@@ -120,6 +135,11 @@ class AudiomackAlbumIE(InfoExtractor):
                 for resultkey, apikey in [('id', 'album_id'), ('title', 'album_title')]:
                     if apikey in api_response and resultkey not in result:
                         result[resultkey] = api_response[apikey]
-                result['entries'].append(AudiomackIE.create_song_dictionary(api_response, album_url_tag, track_no))
-            track_no += 1
+                song_id = url_basename(api_response['url']).rpartition('.')[0]
+                result['entries'].append({
+                    'id': compat_str(api_response.get('id', song_id)),
+                    'uploader': api_response.get('artist'),
+                    'title': api_response.get('title', song_id),
+                    'url': api_response['url'],
+                })
         return result