]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/soundcloud.py
Update upstream source from tag 'upstream/2019.01.16'
[youtubedl] / youtube_dl / extractor / soundcloud.py
index 02e64e09436a5299c5d4f87f1a3ba871c63af230..81c81c8d58ecf5940adfa200d59c5fa247b30cbb 100644 (file)
@@ -1,8 +1,8 @@
-# encoding: utf-8
+# coding: utf-8
 from __future__ import unicode_literals
 
-import re
 import itertools
+import re
 
 from .common import (
     InfoExtractor,
@@ -11,13 +11,13 @@ from .common import (
 from ..compat import (
     compat_str,
     compat_urlparse,
-    compat_urllib_parse,
+    compat_urllib_parse_urlencode,
 )
 from ..utils import (
-    encode_dict,
     ExtractorError,
     int_or_none,
     unified_strdate,
+    update_url_query,
 )
 
 
@@ -32,8 +32,9 @@ class SoundcloudIE(InfoExtractor):
 
     _VALID_URL = r'''(?x)^(?:https?://)?
                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
+                            (?!stations/track)
                             (?P<uploader>[\w\d-]+)/
-                            (?!(?:tracks|sets(?:/[^/?#]+)?|reposts|likes|spotlight)/?(?:$|[?#]))
+                            (?!(?:tracks|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
                             (?P<title>[\w\d-]+)/?
                             (?P<token>[^?]+?)?(?:[?].*)?$)
                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
@@ -54,6 +55,7 @@ class SoundcloudIE(InfoExtractor):
                 'uploader': 'E.T. ExTerrestrial Music',
                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
                 'duration': 143,
+                'license': 'all-rights-reserved',
             }
         },
         # not streamable song
@@ -67,6 +69,7 @@ class SoundcloudIE(InfoExtractor):
                 'uploader': 'The Royal Concept',
                 'upload_date': '20120521',
                 'duration': 227,
+                'license': 'all-rights-reserved',
             },
             'params': {
                 # rtmp
@@ -85,6 +88,7 @@ class SoundcloudIE(InfoExtractor):
                 'description': 'test chars:  \"\'/\\ä↭',
                 'upload_date': '20131209',
                 'duration': 9,
+                'license': 'all-rights-reserved',
             },
         },
         # private link (alt format)
@@ -99,6 +103,7 @@ class SoundcloudIE(InfoExtractor):
                 'description': 'test chars:  \"\'/\\ä↭',
                 'upload_date': '20131209',
                 'duration': 9,
+                'license': 'all-rights-reserved',
             },
         },
         # downloadable song
@@ -113,12 +118,52 @@ class SoundcloudIE(InfoExtractor):
                 'uploader': 'oddsamples',
                 'upload_date': '20140109',
                 'duration': 17,
+                'license': 'cc-by-sa',
+            },
+        },
+        # private link, downloadable format
+        {
+            'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd',
+            'md5': '64a60b16e617d41d0bef032b7f55441e',
+            'info_dict': {
+                'id': '340344461',
+                'ext': 'wav',
+                'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]',
+                'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366',
+                'uploader': 'Ori Uplift Music',
+                'upload_date': '20170831',
+                'duration': 7449,
+                'license': 'all-rights-reserved',
+            },
+        },
+        # no album art, use avatar pic for thumbnail
+        {
+            'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real',
+            'md5': '59c7872bc44e5d99b7211891664760c2',
+            'info_dict': {
+                'id': '309699954',
+                'ext': 'mp3',
+                'title': 'Sideways (Prod. Mad Real)',
+                'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
+                'uploader': 'garyvee',
+                'upload_date': '20170226',
+                'duration': 207,
+                'thumbnail': r're:https?://.*\.jpg',
+                'license': 'all-rights-reserved',
+            },
+            'params': {
+                'skip_download': True,
             },
         },
     ]
 
-    _CLIENT_ID = '02gUJC0hH2ct1EGOcYXQIzRFU91c72Ea'
-    _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
+    _CLIENT_ID = 'LvWovRaJZlWCHql0bISuum8Bd2KX79mb'
+
+    @staticmethod
+    def _extract_urls(webpage):
+        return [m.group('url') for m in re.finditer(
+            r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
+            webpage)]
 
     def report_resolve(self, video_id):
         """Report information extraction."""
@@ -126,34 +171,35 @@ class SoundcloudIE(InfoExtractor):
 
     @classmethod
     def _resolv_url(cls, url):
-        return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
+        return 'https://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
 
     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
         track_id = compat_str(info['id'])
         name = full_title or track_id
         if quiet:
             self.report_extraction(name)
-
-        thumbnail = info['artwork_url']
-        if thumbnail is not None:
+        thumbnail = info.get('artwork_url') or info.get('user', {}).get('avatar_url')
+        if isinstance(thumbnail, compat_str):
             thumbnail = thumbnail.replace('-large', '-t500x500')
-        ext = 'mp3'
         result = {
             'id': track_id,
-            'uploader': info['user']['username'],
-            'upload_date': unified_strdate(info['created_at']),
+            'uploader': info.get('user', {}).get('username'),
+            'upload_date': unified_strdate(info.get('created_at')),
             'title': info['title'],
-            'description': info['description'],
+            'description': info.get('description'),
             'thumbnail': thumbnail,
             'duration': int_or_none(info.get('duration'), 1000),
             'webpage_url': info.get('permalink_url'),
+            'license': info.get('license'),
         }
         formats = []
+        query = {'client_id': self._CLIENT_ID}
+        if secret_token is not None:
+            query['secret_token'] = secret_token
         if info.get('downloadable', False):
             # We can build a direct link to the song
-            format_url = (
-                'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
-                    track_id, self._CLIENT_ID))
+            format_url = update_url_query(
+                'https://api.soundcloud.com/tracks/%s/download' % track_id, query)
             formats.append({
                 'format_id': 'download',
                 'ext': info.get('original_format', 'mp3'),
@@ -163,46 +209,55 @@ class SoundcloudIE(InfoExtractor):
             })
 
         # We have to retrieve the url
-        streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
-                       'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
         format_dict = self._download_json(
-            streams_url,
-            track_id, 'Downloading track url')
+            'https://api.soundcloud.com/i1/tracks/%s/streams' % track_id,
+            track_id, 'Downloading track url', query=query)
 
         for key, stream_url in format_dict.items():
+            ext, abr = 'mp3', None
+            mobj = re.search(r'_([^_]+)_(\d+)_url', key)
+            if mobj:
+                ext, abr = mobj.groups()
+                abr = int(abr)
             if key.startswith('http'):
-                formats.append({
+                stream_formats = [{
                     'format_id': key,
                     'ext': ext,
                     'url': stream_url,
-                    'vcodec': 'none',
-                })
+                }]
             elif key.startswith('rtmp'):
                 # The url doesn't have an rtmp app, we have to extract the playpath
                 url, path = stream_url.split('mp3:', 1)
-                formats.append({
+                stream_formats = [{
                     'format_id': key,
                     'url': url,
                     'play_path': 'mp3:' + path,
                     'ext': 'flv',
-                    'vcodec': 'none',
-                })
-
-            if not formats:
-                # We fallback to the stream_url in the original info, this
-                # cannot be always used, sometimes it can give an HTTP 404 error
-                formats.append({
-                    'format_id': 'fallback',
-                    'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
-                    'ext': ext,
-                    'vcodec': 'none',
-                })
+                }]
+            elif key.startswith('hls'):
+                stream_formats = self._extract_m3u8_formats(
+                    stream_url, track_id, ext, entry_protocol='m3u8_native',
+                    m3u8_id=key, fatal=False)
+            else:
+                continue
+
+            if abr:
+                for f in stream_formats:
+                    f['abr'] = abr
+
+            formats.extend(stream_formats)
+
+        if not formats:
+            # We fallback to the stream_url in the original info, this
+            # cannot be always used, sometimes it can give an HTTP 404 error
+            formats.append({
+                'format_id': 'fallback',
+                'url': update_url_query(info['stream_url'], query),
+                'ext': 'mp3',
+            })
 
-            for f in formats:
-                if f['format_id'].startswith('http'):
-                    f['protocol'] = 'http'
-                if f['format_id'].startswith('rtmp'):
-                    f['protocol'] = 'rtmp'
+        for f in formats:
+            f['vcodec'] = 'none'
 
         self._check_formats(formats, track_id)
         self._sort_formats(formats)
@@ -216,13 +271,13 @@ class SoundcloudIE(InfoExtractor):
             raise ExtractorError('Invalid URL: %s' % url)
 
         track_id = mobj.group('track_id')
-        token = None
+
         if track_id is not None:
-            info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
+            info_json_url = 'https://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
             full_title = track_id
             token = mobj.group('secret_token')
             if token:
-                info_json_url += "&secret_token=" + token
+                info_json_url += '&secret_token=' + token
         elif mobj.group('player'):
             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
             real_url = query['url'][0]
@@ -243,14 +298,27 @@ class SoundcloudIE(InfoExtractor):
 
             self.report_resolve(full_title)
 
-            url = 'http://soundcloud.com/%s' % resolve_title
+            url = 'https://soundcloud.com/%s' % resolve_title
             info_json_url = self._resolv_url(url)
         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
 
         return self._extract_info_dict(info, full_title, secret_token=token)
 
 
-class SoundcloudSetIE(SoundcloudIE):
+class SoundcloudPlaylistBaseIE(SoundcloudIE):
+    @staticmethod
+    def _extract_id(e):
+        return compat_str(e['id']) if e.get('id') else None
+
+    def _extract_track_entries(self, tracks):
+        return [
+            self.url_result(
+                track['permalink_url'], SoundcloudIE.ie_key(),
+                video_id=self._extract_id(track))
+            for track in tracks if track.get('permalink_url')]
+
+
+class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
     IE_NAME = 'soundcloud:set'
     _TESTS = [{
@@ -259,7 +327,10 @@ class SoundcloudSetIE(SoundcloudIE):
             'id': '2284613',
             'title': 'The Royal Concept EP',
         },
-        'playlist_mincount': 6,
+        'playlist_mincount': 5,
+    }, {
+        'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -270,7 +341,7 @@ class SoundcloudSetIE(SoundcloudIE):
         # extract simple title (uploader + slug of song title)
         slug_title = mobj.group('slug_title')
         full_title = '%s/sets/%s' % (uploader, slug_title)
-        url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
+        url = 'https://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
 
         token = mobj.group('token')
         if token:
@@ -286,7 +357,7 @@ class SoundcloudSetIE(SoundcloudIE):
             msgs = (compat_str(err['error_message']) for err in info['errors'])
             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
 
-        entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in info['tracks']]
+        entries = self._extract_track_entries(info['tracks'])
 
         return {
             '_type': 'playlist',
@@ -296,7 +367,63 @@ class SoundcloudSetIE(SoundcloudIE):
         }
 
 
-class SoundcloudUserIE(SoundcloudIE):
+class SoundcloudPagedPlaylistBaseIE(SoundcloudPlaylistBaseIE):
+    _API_BASE = 'https://api.soundcloud.com'
+    _API_V2_BASE = 'https://api-v2.soundcloud.com'
+
+    def _extract_playlist(self, base_url, playlist_id, playlist_title):
+        COMMON_QUERY = {
+            'limit': 50,
+            'client_id': self._CLIENT_ID,
+            'linked_partitioning': '1',
+        }
+
+        query = COMMON_QUERY.copy()
+        query['offset'] = 0
+
+        next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
+
+        entries = []
+        for i in itertools.count():
+            response = self._download_json(
+                next_href, playlist_id, 'Downloading track page %s' % (i + 1))
+
+            collection = response['collection']
+            if not collection:
+                break
+
+            def resolve_permalink_url(candidates):
+                for cand in candidates:
+                    if isinstance(cand, dict):
+                        permalink_url = cand.get('permalink_url')
+                        entry_id = self._extract_id(cand)
+                        if permalink_url and permalink_url.startswith('http'):
+                            return permalink_url, entry_id
+
+            for e in collection:
+                permalink_url, entry_id = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
+                if permalink_url:
+                    entries.append(self.url_result(permalink_url, video_id=entry_id))
+
+            next_href = response.get('next_href')
+            if not next_href:
+                break
+
+            parsed_next_href = compat_urlparse.urlparse(response['next_href'])
+            qs = compat_urlparse.parse_qs(parsed_next_href.query)
+            qs.update(COMMON_QUERY)
+            next_href = compat_urlparse.urlunparse(
+                parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
+
+        return {
+            '_type': 'playlist',
+            'id': playlist_id,
+            'title': playlist_title,
+            'entries': entries,
+        }
+
+
+class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
     _VALID_URL = r'''(?x)
                         https?://
                             (?:(?:www|m)\.)?soundcloud\.com/
@@ -313,21 +440,21 @@ class SoundcloudUserIE(SoundcloudIE):
             'id': '114582580',
             'title': 'The Akashic Chronicler (All)',
         },
-        'playlist_mincount': 111,
+        'playlist_mincount': 74,
     }, {
         'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
         'info_dict': {
             'id': '114582580',
             'title': 'The Akashic Chronicler (Tracks)',
         },
-        'playlist_mincount': 50,
+        'playlist_mincount': 37,
     }, {
         'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
         'info_dict': {
             'id': '114582580',
             'title': 'The Akashic Chronicler (Playlists)',
         },
-        'playlist_mincount': 3,
+        'playlist_mincount': 2,
     }, {
         'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
         'info_dict': {
@@ -351,16 +478,13 @@ class SoundcloudUserIE(SoundcloudIE):
         'playlist_mincount': 1,
     }]
 
-    _API_BASE = 'https://api.soundcloud.com'
-    _API_V2_BASE = 'https://api-v2.soundcloud.com'
-
     _BASE_URL_MAP = {
-        'all': '%s/profile/soundcloud:users:%%s' % _API_V2_BASE,
-        'tracks': '%s/users/%%s/tracks' % _API_BASE,
-        'sets': '%s/users/%%s/playlists' % _API_V2_BASE,
-        'reposts': '%s/profile/soundcloud:users:%%s/reposts' % _API_V2_BASE,
-        'likes': '%s/users/%%s/likes' % _API_V2_BASE,
-        'spotlight': '%s/users/%%s/spotlight' % _API_V2_BASE,
+        'all': '%s/profile/soundcloud:users:%%s' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
+        'tracks': '%s/users/%%s/tracks' % SoundcloudPagedPlaylistBaseIE._API_BASE,
+        'sets': '%s/users/%%s/playlists' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
+        'reposts': '%s/profile/soundcloud:users:%%s/reposts' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
+        'likes': '%s/users/%%s/likes' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
+        'spotlight': '%s/users/%%s/spotlight' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
     }
 
     _TITLE_MAP = {
@@ -376,69 +500,49 @@ class SoundcloudUserIE(SoundcloudIE):
         mobj = re.match(self._VALID_URL, url)
         uploader = mobj.group('user')
 
-        url = 'http://soundcloud.com/%s/' % uploader
+        url = 'https://soundcloud.com/%s/' % uploader
         resolv_url = self._resolv_url(url)
         user = self._download_json(
             resolv_url, uploader, 'Downloading user info')
 
         resource = mobj.group('rsrc') or 'all'
-        base_url = self._BASE_URL_MAP[resource] % user['id']
 
-        next_href = None
+        return self._extract_playlist(
+            self._BASE_URL_MAP[resource] % user['id'], compat_str(user['id']),
+            '%s (%s)' % (user['username'], self._TITLE_MAP[resource]))
 
-        entries = []
-        for i in itertools.count():
-            if not next_href:
-                data = compat_urllib_parse.urlencode({
-                    'offset': i * 50,
-                    'limit': 50,
-                    'client_id': self._CLIENT_ID,
-                    'linked_partitioning': '1',
-                    'representation': 'speedy',
-                })
-                next_href = base_url + '?' + data
 
-            response = self._download_json(
-                next_href, uploader, 'Downloading track page %s' % (i + 1))
-
-            collection = response['collection']
-
-            if not collection:
-                self.to_screen('%s: End page received' % uploader)
-                break
+class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
+    _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
+    IE_NAME = 'soundcloud:trackstation'
+    _TESTS = [{
+        'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
+        'info_dict': {
+            'id': '286017854',
+            'title': 'Track station: your-text',
+        },
+        'playlist_mincount': 47,
+    }]
 
-            def resolve_permalink_url(candidates):
-                for cand in candidates:
-                    if isinstance(cand, dict):
-                        permalink_url = cand.get('permalink_url')
-                        if permalink_url and permalink_url.startswith('http'):
-                            return permalink_url
+    def _real_extract(self, url):
+        track_name = self._match_id(url)
 
-            for e in collection:
-                permalink_url = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
-                if permalink_url:
-                    entries.append(self.url_result(permalink_url))
+        webpage = self._download_webpage(url, track_name)
 
-            if 'next_href' in response:
-                next_href = response['next_href']
-                if not next_href:
-                    break
-            else:
-                next_href = None
+        track_id = self._search_regex(
+            r'soundcloud:track-stations:(\d+)', webpage, 'track id')
 
-        return {
-            '_type': 'playlist',
-            'id': compat_str(user['id']),
-            'title': '%s (%s)' % (user['username'], self._TITLE_MAP[resource]),
-            'entries': entries,
-        }
+        return self._extract_playlist(
+            '%s/stations/soundcloud:track-stations:%s/tracks'
+            % (self._API_V2_BASE, track_id),
+            track_id, 'Track station: %s' % track_name)
 
 
-class SoundcloudPlaylistIE(SoundcloudIE):
+class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
     IE_NAME = 'soundcloud:playlist'
     _TESTS = [{
-        'url': 'http://api.soundcloud.com/playlists/4110309',
+        'url': 'https://api.soundcloud.com/playlists/4110309',
         'info_dict': {
             'id': '4110309',
             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
@@ -460,11 +564,11 @@ class SoundcloudPlaylistIE(SoundcloudIE):
         if token:
             data_dict['secret_token'] = token
 
-        data = compat_urllib_parse.urlencode(data_dict)
+        data = compat_urllib_parse_urlencode(data_dict)
         data = self._download_json(
             base_url + data, playlist_id, 'Downloading playlist')
 
-        entries = [self.url_result(track['permalink_url'], 'Soundcloud') for track in data['tracks']]
+        entries = self._extract_track_entries(data['tracks'])
 
         return {
             '_type': 'playlist',
@@ -500,7 +604,7 @@ class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
         query['client_id'] = self._CLIENT_ID
         query['linked_partitioning'] = '1'
         query['offset'] = 0
-        data = compat_urllib_parse.urlencode(encode_dict(query))
+        data = compat_urllib_parse_urlencode(query)
         next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
 
         collected_results = 0