Merge tag 'upstream/2016.02.22'

[youtubedl] / youtube_dl / extractor / audiomack.py

diff --git a/youtube_dl/extractor/audiomack.py b/youtube_dl/extractor/audiomack.py
index 6232d2cd041b7db63b4880f5d983980af65f8a65..3eed91279fd7b6ead45bf3ee01486eab53b91d6a 100644 (file)
--- a/youtube_dl/extractor/audiomack.py
+++ b/youtube_dl/extractor/audiomack.py
@@ -1,69 +1,144 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
 # coding: utf-8
 from __future__ import unicode_literals
 

+import itertools
+import time
+

 from .common import InfoExtractor
 from .soundcloud import SoundcloudIE
 from .common import InfoExtractor
 from .soundcloud import SoundcloudIE

-from ..utils import ExtractorError
-
-import time
+from ..utils import (
+    ExtractorError,
+    url_basename,
+)

 
 
 class AudiomackIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?audiomack\.com/song/(?P<id>[\w/-]+)'
     IE_NAME = 'audiomack'
     _TESTS = [
 
 
 class AudiomackIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?audiomack\.com/song/(?P<id>[\w/-]+)'
     IE_NAME = 'audiomack'
     _TESTS = [

-        #hosted on audiomack
+        # hosted on audiomack

         {
             'url': 'http://www.audiomack.com/song/roosh-williams/extraordinary',
             'info_dict':
             {
         {
             'url': 'http://www.audiomack.com/song/roosh-williams/extraordinary',
             'info_dict':
             {

-                'id' : 'roosh-williams/extraordinary',
+                'id': '310086',

                 'ext': 'mp3',
                 'ext': 'mp3',

-                'title': 'Roosh Williams - Extraordinary'
+                'uploader': 'Roosh Williams',
+                'title': 'Extraordinary'

             }
         },
             }
         },

-        #hosted on soundcloud via audiomack
+        # audiomack wrapper around soundcloud song

         {
         {

+            'add_ie': ['Soundcloud'],

             'url': 'http://www.audiomack.com/song/xclusiveszone/take-kare',
             'url': 'http://www.audiomack.com/song/xclusiveszone/take-kare',

-            'file': '172419696.mp3',
-            'info_dict':
-            {
+            'info_dict': {
+                'id': '172419696',

                 'ext': 'mp3',
                 'ext': 'mp3',

+                'description': 'md5:1fc3272ed7a635cce5be1568c2822997',

                 'title': 'Young Thug ft Lil Wayne - Take Kare',
                 'title': 'Young Thug ft Lil Wayne - Take Kare',

-                "upload_date": "20141016",
-                "description": "New track produced by London On Da Track called “Take Kare\"\n\nhttp://instagram.com/theyoungthugworld\nhttps://www.facebook.com/ThuggerThuggerCashMoney\n",
-                "uploader": "Young Thug World"
+                'uploader': 'Young Thug World',
+                'upload_date': '20141016',

             }
             }

-        }
+        },

     ]
 
     def _real_extract(self, url):
     ]
 
     def _real_extract(self, url):

-        video_id = self._match_id(url)
+        # URLs end with [uploader name]/[uploader title]
+        # this title is whatever the user types in, and is rarely
+        # the proper song title.  Real metadata is in the api response
+        album_url_tag = self._match_id(url)

 
 

+        # Request the extended version of the api for extra fields like artist and title

         api_response = self._download_json(
         api_response = self._download_json(

-            "http://www.audiomack.com/api/music/url/song/%s?_=%d" % (
-                video_id, time.time()),
-            video_id)
-
-        if "url" not in api_response:
-            raise ExtractorError("Unable to deduce api url of song")
-        realurl = api_response["url"]
+            'http://www.audiomack.com/api/music/url/song/%s?extended=1&_=%d' % (
+                album_url_tag, time.time()),
+            album_url_tag)

 
 

-        #Audiomack wraps a lot of soundcloud tracks in their branded wrapper
-        # - if so, pass the work off to the soundcloud extractor
-        if SoundcloudIE.suitable(realurl):
-            return {'_type': 'url', 'url': realurl, 'ie_key': 'Soundcloud'}
+        # 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)

 
 

-        webpage = self._download_webpage(url, video_id)
-        artist = self._html_search_regex(
-            r'<span class="artist">(.*?)</span>', webpage, "artist")
-        songtitle = self._html_search_regex(
-            r'<h1 class="profile-title song-title"><span class="artist">.*?</span>(.*?)</h1>',
-            webpage, "title")
-        title = artist + " - " + songtitle
+        # 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 {
 
         return {

-            'id': video_id,
-            'title': title,
-            'url': realurl,
+            'id': api_response.get('id', album_url_tag),
+            'uploader': api_response.get('artist'),
+            'title': api_response.get('title'),
+            'url': api_response['url'],
+        }
+
+
+class AudiomackAlbumIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?audiomack\.com/album/(?P<id>[\w/-]+)'
+    IE_NAME = 'audiomack:album'
+    _TESTS = [
+        # Standard album playlist
+        {
+            'url': 'http://www.audiomack.com/album/flytunezcom/tha-tour-part-2-mixtape',
+            'playlist_count': 15,
+            'info_dict':
+            {
+                'id': '812251',
+                'title': 'Tha Tour: Part 2 (Official Mixtape)'
+            }
+        },
+        # Album playlist ripped from fakeshoredrive with no metadata
+        {
+            'url': 'http://www.audiomack.com/album/fakeshoredrive/ppp-pistol-p-project',
+            '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,
+            }

         }
         }

+    ]
+
+    def _real_extract(self, url):
+        # URLs end with [uploader name]/[uploader title]
+        # this title is whatever the user types in, and is rarely
+        # the proper song title.  Real metadata is in the api response
+        album_url_tag = self._match_id(url)
+        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
+        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,
+                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)
+            if 'url' not in api_response or 'error' in api_response:
+                raise ExtractorError('Invalid url for track %d of album url %s' % (track_no, url))
+            # URL is good but song id doesn't exist - usually means end of playlist
+            elif not api_response['url']:
+                break
+            else:
+                # Pull out the album metadata and add to result (if it exists)
+                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]
+                song_id = url_basename(api_response['url']).rpartition('.')[0]
+                result['entries'].append({
+                    'id': api_response.get('id', song_id),
+                    'uploader': api_response.get('artist'),
+                    'title': api_response.get('title', song_id),
+                    'url': api_response['url'],
+                })
+        return result