]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/soundcloud.py
New upstream version 2017.09.24
[youtubedl] / 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 ]
140
141 _CLIENT_ID = 'JlZIsxg2hY5WnBgtn3jfS0UYCl0K8DOg'
142 _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
143
144 @staticmethod
145 def _extract_urls(webpage):
146 return [m.group('url') for m in re.finditer(
147 r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
148 webpage)]
149
150 def report_resolve(self, video_id):
151 """Report information extraction."""
152 self.to_screen('%s: Resolving id' % video_id)
153
154 @classmethod
155 def _resolv_url(cls, url):
156 return 'https://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID
157
158 def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
159 track_id = compat_str(info['id'])
160 name = full_title or track_id
161 if quiet:
162 self.report_extraction(name)
163 thumbnail = info.get('artwork_url')
164 if isinstance(thumbnail, compat_str):
165 thumbnail = thumbnail.replace('-large', '-t500x500')
166 ext = 'mp3'
167 result = {
168 'id': track_id,
169 'uploader': info.get('user', {}).get('username'),
170 'upload_date': unified_strdate(info.get('created_at')),
171 'title': info['title'],
172 'description': info.get('description'),
173 'thumbnail': thumbnail,
174 'duration': int_or_none(info.get('duration'), 1000),
175 'webpage_url': info.get('permalink_url'),
176 'license': info.get('license'),
177 }
178 formats = []
179 query = {'client_id': self._CLIENT_ID}
180 if secret_token is not None:
181 query['secret_token'] = secret_token
182 if info.get('downloadable', False):
183 # We can build a direct link to the song
184 format_url = update_url_query(
185 'https://api.soundcloud.com/tracks/%s/download' % track_id, query)
186 formats.append({
187 'format_id': 'download',
188 'ext': info.get('original_format', 'mp3'),
189 'url': format_url,
190 'vcodec': 'none',
191 'preference': 10,
192 })
193
194 # We have to retrieve the url
195 format_dict = self._download_json(
196 'https://api.soundcloud.com/i1/tracks/%s/streams' % track_id,
197 track_id, 'Downloading track url', query=query)
198
199 for key, stream_url in format_dict.items():
200 abr = int_or_none(self._search_regex(
201 r'_(\d+)_url', key, 'audio bitrate', default=None))
202 if key.startswith('http'):
203 stream_formats = [{
204 'format_id': key,
205 'ext': ext,
206 'url': stream_url,
207 }]
208 elif key.startswith('rtmp'):
209 # The url doesn't have an rtmp app, we have to extract the playpath
210 url, path = stream_url.split('mp3:', 1)
211 stream_formats = [{
212 'format_id': key,
213 'url': url,
214 'play_path': 'mp3:' + path,
215 'ext': 'flv',
216 }]
217 elif key.startswith('hls'):
218 stream_formats = self._extract_m3u8_formats(
219 stream_url, track_id, 'mp3', entry_protocol='m3u8_native',
220 m3u8_id=key, fatal=False)
221 else:
222 continue
223
224 for f in stream_formats:
225 f['abr'] = abr
226
227 formats.extend(stream_formats)
228
229 if not formats:
230 # We fallback to the stream_url in the original info, this
231 # cannot be always used, sometimes it can give an HTTP 404 error
232 formats.append({
233 'format_id': 'fallback',
234 'url': update_url_query(info['stream_url'], query),
235 'ext': ext,
236 })
237
238 for f in formats:
239 f['vcodec'] = 'none'
240
241 self._check_formats(formats, track_id)
242 self._sort_formats(formats)
243 result['formats'] = formats
244
245 return result
246
247 def _real_extract(self, url):
248 mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
249 if mobj is None:
250 raise ExtractorError('Invalid URL: %s' % url)
251
252 track_id = mobj.group('track_id')
253
254 if track_id is not None:
255 info_json_url = 'https://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
256 full_title = track_id
257 token = mobj.group('secret_token')
258 if token:
259 info_json_url += '&secret_token=' + token
260 elif mobj.group('player'):
261 query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
262 real_url = query['url'][0]
263 # If the token is in the query of the original url we have to
264 # manually add it
265 if 'secret_token' in query:
266 real_url += '?secret_token=' + query['secret_token'][0]
267 return self.url_result(real_url)
268 else:
269 # extract uploader (which is in the url)
270 uploader = mobj.group('uploader')
271 # extract simple title (uploader + slug of song title)
272 slug_title = mobj.group('title')
273 token = mobj.group('token')
274 full_title = resolve_title = '%s/%s' % (uploader, slug_title)
275 if token:
276 resolve_title += '/%s' % token
277
278 self.report_resolve(full_title)
279
280 url = 'https://soundcloud.com/%s' % resolve_title
281 info_json_url = self._resolv_url(url)
282 info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
283
284 return self._extract_info_dict(info, full_title, secret_token=token)
285
286
287 class SoundcloudPlaylistBaseIE(SoundcloudIE):
288 @staticmethod
289 def _extract_id(e):
290 return compat_str(e['id']) if e.get('id') else None
291
292 def _extract_track_entries(self, tracks):
293 return [
294 self.url_result(
295 track['permalink_url'], SoundcloudIE.ie_key(),
296 video_id=self._extract_id(track))
297 for track in tracks if track.get('permalink_url')]
298
299
300 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
301 _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
302 IE_NAME = 'soundcloud:set'
303 _TESTS = [{
304 'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
305 'info_dict': {
306 'id': '2284613',
307 'title': 'The Royal Concept EP',
308 },
309 'playlist_mincount': 5,
310 }, {
311 'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
312 'only_matching': True,
313 }]
314
315 def _real_extract(self, url):
316 mobj = re.match(self._VALID_URL, url)
317
318 # extract uploader (which is in the url)
319 uploader = mobj.group('uploader')
320 # extract simple title (uploader + slug of song title)
321 slug_title = mobj.group('slug_title')
322 full_title = '%s/sets/%s' % (uploader, slug_title)
323 url = 'https://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
324
325 token = mobj.group('token')
326 if token:
327 full_title += '/' + token
328 url += '/' + token
329
330 self.report_resolve(full_title)
331
332 resolv_url = self._resolv_url(url)
333 info = self._download_json(resolv_url, full_title)
334
335 if 'errors' in info:
336 msgs = (compat_str(err['error_message']) for err in info['errors'])
337 raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
338
339 entries = self._extract_track_entries(info['tracks'])
340
341 return {
342 '_type': 'playlist',
343 'entries': entries,
344 'id': '%s' % info['id'],
345 'title': info['title'],
346 }
347
348
349 class SoundcloudPagedPlaylistBaseIE(SoundcloudPlaylistBaseIE):
350 _API_BASE = 'https://api.soundcloud.com'
351 _API_V2_BASE = 'https://api-v2.soundcloud.com'
352
353 def _extract_playlist(self, base_url, playlist_id, playlist_title):
354 COMMON_QUERY = {
355 'limit': 50,
356 'client_id': self._CLIENT_ID,
357 'linked_partitioning': '1',
358 }
359
360 query = COMMON_QUERY.copy()
361 query['offset'] = 0
362
363 next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
364
365 entries = []
366 for i in itertools.count():
367 response = self._download_json(
368 next_href, playlist_id, 'Downloading track page %s' % (i + 1))
369
370 collection = response['collection']
371 if not collection:
372 break
373
374 def resolve_permalink_url(candidates):
375 for cand in candidates:
376 if isinstance(cand, dict):
377 permalink_url = cand.get('permalink_url')
378 entry_id = self._extract_id(cand)
379 if permalink_url and permalink_url.startswith('http'):
380 return permalink_url, entry_id
381
382 for e in collection:
383 permalink_url, entry_id = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
384 if permalink_url:
385 entries.append(self.url_result(permalink_url, video_id=entry_id))
386
387 next_href = response.get('next_href')
388 if not next_href:
389 break
390
391 parsed_next_href = compat_urlparse.urlparse(response['next_href'])
392 qs = compat_urlparse.parse_qs(parsed_next_href.query)
393 qs.update(COMMON_QUERY)
394 next_href = compat_urlparse.urlunparse(
395 parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
396
397 return {
398 '_type': 'playlist',
399 'id': playlist_id,
400 'title': playlist_title,
401 'entries': entries,
402 }
403
404
405 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
406 _VALID_URL = r'''(?x)
407 https?://
408 (?:(?:www|m)\.)?soundcloud\.com/
409 (?P<user>[^/]+)
410 (?:/
411 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
412 )?
413 /?(?:[?#].*)?$
414 '''
415 IE_NAME = 'soundcloud:user'
416 _TESTS = [{
417 'url': 'https://soundcloud.com/the-akashic-chronicler',
418 'info_dict': {
419 'id': '114582580',
420 'title': 'The Akashic Chronicler (All)',
421 },
422 'playlist_mincount': 74,
423 }, {
424 'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
425 'info_dict': {
426 'id': '114582580',
427 'title': 'The Akashic Chronicler (Tracks)',
428 },
429 'playlist_mincount': 37,
430 }, {
431 'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
432 'info_dict': {
433 'id': '114582580',
434 'title': 'The Akashic Chronicler (Playlists)',
435 },
436 'playlist_mincount': 2,
437 }, {
438 'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
439 'info_dict': {
440 'id': '114582580',
441 'title': 'The Akashic Chronicler (Reposts)',
442 },
443 'playlist_mincount': 7,
444 }, {
445 'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
446 'info_dict': {
447 'id': '114582580',
448 'title': 'The Akashic Chronicler (Likes)',
449 },
450 'playlist_mincount': 321,
451 }, {
452 'url': 'https://soundcloud.com/grynpyret/spotlight',
453 'info_dict': {
454 'id': '7098329',
455 'title': 'Grynpyret (Spotlight)',
456 },
457 'playlist_mincount': 1,
458 }]
459
460 _BASE_URL_MAP = {
461 'all': '%s/profile/soundcloud:users:%%s' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
462 'tracks': '%s/users/%%s/tracks' % SoundcloudPagedPlaylistBaseIE._API_BASE,
463 'sets': '%s/users/%%s/playlists' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
464 'reposts': '%s/profile/soundcloud:users:%%s/reposts' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
465 'likes': '%s/users/%%s/likes' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
466 'spotlight': '%s/users/%%s/spotlight' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
467 }
468
469 _TITLE_MAP = {
470 'all': 'All',
471 'tracks': 'Tracks',
472 'sets': 'Playlists',
473 'reposts': 'Reposts',
474 'likes': 'Likes',
475 'spotlight': 'Spotlight',
476 }
477
478 def _real_extract(self, url):
479 mobj = re.match(self._VALID_URL, url)
480 uploader = mobj.group('user')
481
482 url = 'https://soundcloud.com/%s/' % uploader
483 resolv_url = self._resolv_url(url)
484 user = self._download_json(
485 resolv_url, uploader, 'Downloading user info')
486
487 resource = mobj.group('rsrc') or 'all'
488
489 return self._extract_playlist(
490 self._BASE_URL_MAP[resource] % user['id'], compat_str(user['id']),
491 '%s (%s)' % (user['username'], self._TITLE_MAP[resource]))
492
493
494 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
495 _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
496 IE_NAME = 'soundcloud:trackstation'
497 _TESTS = [{
498 'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
499 'info_dict': {
500 'id': '286017854',
501 'title': 'Track station: your-text',
502 },
503 'playlist_mincount': 47,
504 }]
505
506 def _real_extract(self, url):
507 track_name = self._match_id(url)
508
509 webpage = self._download_webpage(url, track_name)
510
511 track_id = self._search_regex(
512 r'soundcloud:track-stations:(\d+)', webpage, 'track id')
513
514 return self._extract_playlist(
515 '%s/stations/soundcloud:track-stations:%s/tracks'
516 % (self._API_V2_BASE, track_id),
517 track_id, 'Track station: %s' % track_name)
518
519
520 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
521 _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
522 IE_NAME = 'soundcloud:playlist'
523 _TESTS = [{
524 'url': 'https://api.soundcloud.com/playlists/4110309',
525 'info_dict': {
526 'id': '4110309',
527 'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
528 'description': 're:.*?TILT Brass - Bowery Poetry Club',
529 },
530 'playlist_count': 6,
531 }]
532
533 def _real_extract(self, url):
534 mobj = re.match(self._VALID_URL, url)
535 playlist_id = mobj.group('id')
536 base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
537
538 data_dict = {
539 'client_id': self._CLIENT_ID,
540 }
541 token = mobj.group('token')
542
543 if token:
544 data_dict['secret_token'] = token
545
546 data = compat_urllib_parse_urlencode(data_dict)
547 data = self._download_json(
548 base_url + data, playlist_id, 'Downloading playlist')
549
550 entries = self._extract_track_entries(data['tracks'])
551
552 return {
553 '_type': 'playlist',
554 'id': playlist_id,
555 'title': data.get('title'),
556 'description': data.get('description'),
557 'entries': entries,
558 }
559
560
561 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
562 IE_NAME = 'soundcloud:search'
563 IE_DESC = 'Soundcloud search'
564 _MAX_RESULTS = float('inf')
565 _TESTS = [{
566 'url': 'scsearch15:post-avant jazzcore',
567 'info_dict': {
568 'title': 'post-avant jazzcore',
569 },
570 'playlist_count': 15,
571 }]
572
573 _SEARCH_KEY = 'scsearch'
574 _MAX_RESULTS_PER_PAGE = 200
575 _DEFAULT_RESULTS_PER_PAGE = 50
576 _API_V2_BASE = 'https://api-v2.soundcloud.com'
577
578 def _get_collection(self, endpoint, collection_id, **query):
579 limit = min(
580 query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
581 self._MAX_RESULTS_PER_PAGE)
582 query['limit'] = limit
583 query['client_id'] = self._CLIENT_ID
584 query['linked_partitioning'] = '1'
585 query['offset'] = 0
586 data = compat_urllib_parse_urlencode(query)
587 next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
588
589 collected_results = 0
590
591 for i in itertools.count(1):
592 response = self._download_json(
593 next_url, collection_id, 'Downloading page {0}'.format(i),
594 'Unable to download API page')
595
596 collection = response.get('collection', [])
597 if not collection:
598 break
599
600 collection = list(filter(bool, collection))
601 collected_results += len(collection)
602
603 for item in collection:
604 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
605
606 if not collection or collected_results >= limit:
607 break
608
609 next_url = response.get('next_href')
610 if not next_url:
611 break
612
613 def _get_n_results(self, query, n):
614 tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
615 return self.playlist_result(tracks, playlist_title=query)