]> git.bitcoin.ninja Git - youtube-dl/blob - youtube_dl/extractor/soundcloud.py
[downloader/http] Randomize HTTP chunk size
[youtube-dl] / youtube_dl / extractor / soundcloud.py
1 # coding: utf-8
2 from __future__ import unicode_literals
3
4 import itertools
5 import re
6
7 from .common import (
8     InfoExtractor,
9     SearchInfoExtractor
10 )
11 from ..compat import (
12     compat_str,
13     compat_urlparse,
14     compat_urllib_parse_urlencode,
15 )
16 from ..utils import (
17     ExtractorError,
18     int_or_none,
19     unified_strdate,
20     update_url_query,
21 )
22
23
24 class SoundcloudIE(InfoExtractor):
25     """Information extractor for soundcloud.com
26        To access the media, the uid of the song and a stream token
27        must be extracted from the page source and the script must make
28        a request to media.soundcloud.com/crossdomain.xml. Then
29        the media can be grabbed by requesting from an url composed
30        of the stream token and uid
31      """
32
33     _VALID_URL = r'''(?x)^(?:https?://)?
34                     (?:(?:(?:www\.|m\.)?soundcloud\.com/
35                             (?!stations/track)
36                             (?P<uploader>[\w\d-]+)/
37                             (?!(?:tracks|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
38                             (?P<title>[\w\d-]+)/?
39                             (?P<token>[^?]+?)?(?:[?].*)?$)
40                        |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
41                           (?:/?\?secret_token=(?P<secret_token>[^&]+))?)
42                        |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
43                     )
44                     '''
45     IE_NAME = 'soundcloud'
46     _TESTS = [
47         {
48             'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
49             'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
50             'info_dict': {
51                 'id': '62986583',
52                 'ext': 'mp3',
53                 'upload_date': '20121011',
54                 'description': 'No Downloads untill we record the finished version this weekend, i was too pumped n i had to post it , earl is prolly gonna b hella p.o\'d',
55                 'uploader': 'E.T. ExTerrestrial Music',
56                 'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
57                 'duration': 143,
58                 'license': 'all-rights-reserved',
59             }
60         },
61         # not streamable song
62         {
63             'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
64             'info_dict': {
65                 'id': '47127627',
66                 'ext': 'mp3',
67                 'title': 'Goldrushed',
68                 'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
69                 'uploader': 'The Royal Concept',
70                 'upload_date': '20120521',
71                 'duration': 227,
72                 'license': 'all-rights-reserved',
73             },
74             'params': {
75                 # rtmp
76                 'skip_download': True,
77             },
78         },
79         # private link
80         {
81             'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
82             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
83             'info_dict': {
84                 'id': '123998367',
85                 'ext': 'mp3',
86                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
87                 'uploader': 'jaimeMF',
88                 'description': 'test chars:  \"\'/\\ä↭',
89                 'upload_date': '20131209',
90                 'duration': 9,
91                 'license': 'all-rights-reserved',
92             },
93         },
94         # private link (alt format)
95         {
96             'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
97             'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
98             'info_dict': {
99                 'id': '123998367',
100                 'ext': 'mp3',
101                 'title': 'Youtube - Dl Test Video \'\' Ä↭',
102                 'uploader': 'jaimeMF',
103                 'description': 'test chars:  \"\'/\\ä↭',
104                 'upload_date': '20131209',
105                 'duration': 9,
106                 'license': 'all-rights-reserved',
107             },
108         },
109         # downloadable song
110         {
111             'url': 'https://soundcloud.com/oddsamples/bus-brakes',
112             'md5': '7624f2351f8a3b2e7cd51522496e7631',
113             'info_dict': {
114                 'id': '128590877',
115                 'ext': 'mp3',
116                 'title': 'Bus Brakes',
117                 'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
118                 'uploader': 'oddsamples',
119                 'upload_date': '20140109',
120                 'duration': 17,
121                 'license': 'cc-by-sa',
122             },
123         },
124         # private link, downloadable format
125         {
126             'url': 'https://soundcloud.com/oriuplift/uponly-238-no-talking-wav/s-AyZUd',
127             'md5': '64a60b16e617d41d0bef032b7f55441e',
128             'info_dict': {
129                 'id': '340344461',
130                 'ext': 'wav',
131                 'title': 'Uplifting Only 238 [No Talking] (incl. Alex Feed Guestmix) (Aug 31, 2017) [wav]',
132                 'description': 'md5:fa20ee0fca76a3d6df8c7e57f3715366',
133                 'uploader': 'Ori Uplift Music',
134                 'upload_date': '20170831',
135                 'duration': 7449,
136                 'license': 'all-rights-reserved',
137             },
138         },
139         # no album art, use avatar pic for thumbnail
140         {
141             'url': 'https://soundcloud.com/garyvee/sideways-prod-mad-real',
142             'md5': '59c7872bc44e5d99b7211891664760c2',
143             'info_dict': {
144                 'id': '309699954',
145                 'ext': 'mp3',
146                 'title': 'Sideways (Prod. Mad Real)',
147                 'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
148                 'uploader': 'garyvee',
149                 'upload_date': '20170226',
150                 'duration': 207,
151                 'thumbnail': r're:https?://.*\.jpg',
152                 'license': 'all-rights-reserved',
153             },
154             'params': {
155                 'skip_download': True,
156             },
157         },
158     ]
159
160     _CLIENT_ID = 'DQskPX1pntALRzMp4HSxya3Mc0AO66Ro'
161
162     @staticmethod
163     def _extract_urls(webpage):
164         return [m.group('url') for m in re.finditer(
165             r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
166             webpage)]
167
168     def report_resolve(self, video_id):
169         """Report information extraction."""
170         self.to_screen('%s: Resolving id' % video_id)
171
172     @classmethod
173     def _resolv_url(cls, url):
174         return 'https://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
175
176     def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
177         track_id = compat_str(info['id'])
178         name = full_title or track_id
179         if quiet:
180             self.report_extraction(name)
181         thumbnail = info.get('artwork_url') or info.get('user', {}).get('avatar_url')
182         if isinstance(thumbnail, compat_str):
183             thumbnail = thumbnail.replace('-large', '-t500x500')
184         ext = 'mp3'
185         result = {
186             'id': track_id,
187             'uploader': info.get('user', {}).get('username'),
188             'upload_date': unified_strdate(info.get('created_at')),
189             'title': info['title'],
190             'description': info.get('description'),
191             'thumbnail': thumbnail,
192             'duration': int_or_none(info.get('duration'), 1000),
193             'webpage_url': info.get('permalink_url'),
194             'license': info.get('license'),
195         }
196         formats = []
197         query = {'client_id': self._CLIENT_ID}
198         if secret_token is not None:
199             query['secret_token'] = secret_token
200         if info.get('downloadable', False):
201             # We can build a direct link to the song
202             format_url = update_url_query(
203                 'https://api.soundcloud.com/tracks/%s/download' % track_id, query)
204             formats.append({
205                 'format_id': 'download',
206                 'ext': info.get('original_format', 'mp3'),
207                 'url': format_url,
208                 'vcodec': 'none',
209                 'preference': 10,
210             })
211
212         # We have to retrieve the url
213         format_dict = self._download_json(
214             'https://api.soundcloud.com/i1/tracks/%s/streams' % track_id,
215             track_id, 'Downloading track url', query=query)
216
217         for key, stream_url in format_dict.items():
218             abr = int_or_none(self._search_regex(
219                 r'_(\d+)_url', key, 'audio bitrate', default=None))
220             if key.startswith('http'):
221                 stream_formats = [{
222                     'format_id': key,
223                     'ext': ext,
224                     'url': stream_url,
225                 }]
226             elif key.startswith('rtmp'):
227                 # The url doesn't have an rtmp app, we have to extract the playpath
228                 url, path = stream_url.split('mp3:', 1)
229                 stream_formats = [{
230                     'format_id': key,
231                     'url': url,
232                     'play_path': 'mp3:' + path,
233                     'ext': 'flv',
234                 }]
235             elif key.startswith('hls'):
236                 stream_formats = self._extract_m3u8_formats(
237                     stream_url, track_id, 'mp3', entry_protocol='m3u8_native',
238                     m3u8_id=key, fatal=False)
239             else:
240                 continue
241
242             for f in stream_formats:
243                 f['abr'] = abr
244
245             formats.extend(stream_formats)
246
247         if not formats:
248             # We fallback to the stream_url in the original info, this
249             # cannot be always used, sometimes it can give an HTTP 404 error
250             formats.append({
251                 'format_id': 'fallback',
252                 'url': update_url_query(info['stream_url'], query),
253                 'ext': ext,
254             })
255
256         for f in formats:
257             f['vcodec'] = 'none'
258
259         self._check_formats(formats, track_id)
260         self._sort_formats(formats)
261         result['formats'] = formats
262
263         return result
264
265     def _real_extract(self, url):
266         mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
267         if mobj is None:
268             raise ExtractorError('Invalid URL: %s' % url)
269
270         track_id = mobj.group('track_id')
271
272         if track_id is not None:
273             info_json_url = 'https://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
274             full_title = track_id
275             token = mobj.group('secret_token')
276             if token:
277                 info_json_url += '&secret_token=' + token
278         elif mobj.group('player'):
279             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
280             real_url = query['url'][0]
281             # If the token is in the query of the original url we have to
282             # manually add it
283             if 'secret_token' in query:
284                 real_url += '?secret_token=' + query['secret_token'][0]
285             return self.url_result(real_url)
286         else:
287             # extract uploader (which is in the url)
288             uploader = mobj.group('uploader')
289             # extract simple title (uploader + slug of song title)
290             slug_title = mobj.group('title')
291             token = mobj.group('token')
292             full_title = resolve_title = '%s/%s' % (uploader, slug_title)
293             if token:
294                 resolve_title += '/%s' % token
295
296             self.report_resolve(full_title)
297
298             url = 'https://soundcloud.com/%s' % resolve_title
299             info_json_url = self._resolv_url(url)
300         info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
301
302         return self._extract_info_dict(info, full_title, secret_token=token)
303
304
305 class SoundcloudPlaylistBaseIE(SoundcloudIE):
306     @staticmethod
307     def _extract_id(e):
308         return compat_str(e['id']) if e.get('id') else None
309
310     def _extract_track_entries(self, tracks):
311         return [
312             self.url_result(
313                 track['permalink_url'], SoundcloudIE.ie_key(),
314                 video_id=self._extract_id(track))
315             for track in tracks if track.get('permalink_url')]
316
317
318 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
319     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
320     IE_NAME = 'soundcloud:set'
321     _TESTS = [{
322         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
323         'info_dict': {
324             'id': '2284613',
325             'title': 'The Royal Concept EP',
326         },
327         'playlist_mincount': 5,
328     }, {
329         'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
330         'only_matching': True,
331     }]
332
333     def _real_extract(self, url):
334         mobj = re.match(self._VALID_URL, url)
335
336         # extract uploader (which is in the url)
337         uploader = mobj.group('uploader')
338         # extract simple title (uploader + slug of song title)
339         slug_title = mobj.group('slug_title')
340         full_title = '%s/sets/%s' % (uploader, slug_title)
341         url = 'https://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
342
343         token = mobj.group('token')
344         if token:
345             full_title += '/' + token
346             url += '/' + token
347
348         self.report_resolve(full_title)
349
350         resolv_url = self._resolv_url(url)
351         info = self._download_json(resolv_url, full_title)
352
353         if 'errors' in info:
354             msgs = (compat_str(err['error_message']) for err in info['errors'])
355             raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
356
357         entries = self._extract_track_entries(info['tracks'])
358
359         return {
360             '_type': 'playlist',
361             'entries': entries,
362             'id': '%s' % info['id'],
363             'title': info['title'],
364         }
365
366
367 class SoundcloudPagedPlaylistBaseIE(SoundcloudPlaylistBaseIE):
368     _API_BASE = 'https://api.soundcloud.com'
369     _API_V2_BASE = 'https://api-v2.soundcloud.com'
370
371     def _extract_playlist(self, base_url, playlist_id, playlist_title):
372         COMMON_QUERY = {
373             'limit': 50,
374             'client_id': self._CLIENT_ID,
375             'linked_partitioning': '1',
376         }
377
378         query = COMMON_QUERY.copy()
379         query['offset'] = 0
380
381         next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
382
383         entries = []
384         for i in itertools.count():
385             response = self._download_json(
386                 next_href, playlist_id, 'Downloading track page %s' % (i + 1))
387
388             collection = response['collection']
389             if not collection:
390                 break
391
392             def resolve_permalink_url(candidates):
393                 for cand in candidates:
394                     if isinstance(cand, dict):
395                         permalink_url = cand.get('permalink_url')
396                         entry_id = self._extract_id(cand)
397                         if permalink_url and permalink_url.startswith('http'):
398                             return permalink_url, entry_id
399
400             for e in collection:
401                 permalink_url, entry_id = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
402                 if permalink_url:
403                     entries.append(self.url_result(permalink_url, video_id=entry_id))
404
405             next_href = response.get('next_href')
406             if not next_href:
407                 break
408
409             parsed_next_href = compat_urlparse.urlparse(response['next_href'])
410             qs = compat_urlparse.parse_qs(parsed_next_href.query)
411             qs.update(COMMON_QUERY)
412             next_href = compat_urlparse.urlunparse(
413                 parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
414
415         return {
416             '_type': 'playlist',
417             'id': playlist_id,
418             'title': playlist_title,
419             'entries': entries,
420         }
421
422
423 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
424     _VALID_URL = r'''(?x)
425                         https?://
426                             (?:(?:www|m)\.)?soundcloud\.com/
427                             (?P<user>[^/]+)
428                             (?:/
429                                 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
430                             )?
431                             /?(?:[?#].*)?$
432                     '''
433     IE_NAME = 'soundcloud:user'
434     _TESTS = [{
435         'url': 'https://soundcloud.com/the-akashic-chronicler',
436         'info_dict': {
437             'id': '114582580',
438             'title': 'The Akashic Chronicler (All)',
439         },
440         'playlist_mincount': 74,
441     }, {
442         'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
443         'info_dict': {
444             'id': '114582580',
445             'title': 'The Akashic Chronicler (Tracks)',
446         },
447         'playlist_mincount': 37,
448     }, {
449         'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
450         'info_dict': {
451             'id': '114582580',
452             'title': 'The Akashic Chronicler (Playlists)',
453         },
454         'playlist_mincount': 2,
455     }, {
456         'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
457         'info_dict': {
458             'id': '114582580',
459             'title': 'The Akashic Chronicler (Reposts)',
460         },
461         'playlist_mincount': 7,
462     }, {
463         'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
464         'info_dict': {
465             'id': '114582580',
466             'title': 'The Akashic Chronicler (Likes)',
467         },
468         'playlist_mincount': 321,
469     }, {
470         'url': 'https://soundcloud.com/grynpyret/spotlight',
471         'info_dict': {
472             'id': '7098329',
473             'title': 'Grynpyret (Spotlight)',
474         },
475         'playlist_mincount': 1,
476     }]
477
478     _BASE_URL_MAP = {
479         'all': '%s/profile/soundcloud:users:%%s' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
480         'tracks': '%s/users/%%s/tracks' % SoundcloudPagedPlaylistBaseIE._API_BASE,
481         'sets': '%s/users/%%s/playlists' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
482         'reposts': '%s/profile/soundcloud:users:%%s/reposts' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
483         'likes': '%s/users/%%s/likes' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
484         'spotlight': '%s/users/%%s/spotlight' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
485     }
486
487     _TITLE_MAP = {
488         'all': 'All',
489         'tracks': 'Tracks',
490         'sets': 'Playlists',
491         'reposts': 'Reposts',
492         'likes': 'Likes',
493         'spotlight': 'Spotlight',
494     }
495
496     def _real_extract(self, url):
497         mobj = re.match(self._VALID_URL, url)
498         uploader = mobj.group('user')
499
500         url = 'https://soundcloud.com/%s/' % uploader
501         resolv_url = self._resolv_url(url)
502         user = self._download_json(
503             resolv_url, uploader, 'Downloading user info')
504
505         resource = mobj.group('rsrc') or 'all'
506
507         return self._extract_playlist(
508             self._BASE_URL_MAP[resource] % user['id'], compat_str(user['id']),
509             '%s (%s)' % (user['username'], self._TITLE_MAP[resource]))
510
511
512 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
513     _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
514     IE_NAME = 'soundcloud:trackstation'
515     _TESTS = [{
516         'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
517         'info_dict': {
518             'id': '286017854',
519             'title': 'Track station: your-text',
520         },
521         'playlist_mincount': 47,
522     }]
523
524     def _real_extract(self, url):
525         track_name = self._match_id(url)
526
527         webpage = self._download_webpage(url, track_name)
528
529         track_id = self._search_regex(
530             r'soundcloud:track-stations:(\d+)', webpage, 'track id')
531
532         return self._extract_playlist(
533             '%s/stations/soundcloud:track-stations:%s/tracks'
534             % (self._API_V2_BASE, track_id),
535             track_id, 'Track station: %s' % track_name)
536
537
538 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
539     _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
540     IE_NAME = 'soundcloud:playlist'
541     _TESTS = [{
542         'url': 'https://api.soundcloud.com/playlists/4110309',
543         'info_dict': {
544             'id': '4110309',
545             'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
546             'description': 're:.*?TILT Brass - Bowery Poetry Club',
547         },
548         'playlist_count': 6,
549     }]
550
551     def _real_extract(self, url):
552         mobj = re.match(self._VALID_URL, url)
553         playlist_id = mobj.group('id')
554         base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
555
556         data_dict = {
557             'client_id': self._CLIENT_ID,
558         }
559         token = mobj.group('token')
560
561         if token:
562             data_dict['secret_token'] = token
563
564         data = compat_urllib_parse_urlencode(data_dict)
565         data = self._download_json(
566             base_url + data, playlist_id, 'Downloading playlist')
567
568         entries = self._extract_track_entries(data['tracks'])
569
570         return {
571             '_type': 'playlist',
572             'id': playlist_id,
573             'title': data.get('title'),
574             'description': data.get('description'),
575             'entries': entries,
576         }
577
578
579 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
580     IE_NAME = 'soundcloud:search'
581     IE_DESC = 'Soundcloud search'
582     _MAX_RESULTS = float('inf')
583     _TESTS = [{
584         'url': 'scsearch15:post-avant jazzcore',
585         'info_dict': {
586             'title': 'post-avant jazzcore',
587         },
588         'playlist_count': 15,
589     }]
590
591     _SEARCH_KEY = 'scsearch'
592     _MAX_RESULTS_PER_PAGE = 200
593     _DEFAULT_RESULTS_PER_PAGE = 50
594     _API_V2_BASE = 'https://api-v2.soundcloud.com'
595
596     def _get_collection(self, endpoint, collection_id, **query):
597         limit = min(
598             query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
599             self._MAX_RESULTS_PER_PAGE)
600         query['limit'] = limit
601         query['client_id'] = self._CLIENT_ID
602         query['linked_partitioning'] = '1'
603         query['offset'] = 0
604         data = compat_urllib_parse_urlencode(query)
605         next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
606
607         collected_results = 0
608
609         for i in itertools.count(1):
610             response = self._download_json(
611                 next_url, collection_id, 'Downloading page {0}'.format(i),
612                 'Unable to download API page')
613
614             collection = response.get('collection', [])
615             if not collection:
616                 break
617
618             collection = list(filter(bool, collection))
619             collected_results += len(collection)
620
621             for item in collection:
622                 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
623
624             if not collection or collected_results >= limit:
625                 break
626
627             next_url = response.get('next_href')
628             if not next_url:
629                 break
630
631     def _get_n_results(self, query, n):
632         tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
633         return self.playlist_result(tracks, playlist_title=query)