]> Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/soundcloud.py
Merge changes from buster branch for the NMU while I was away.
[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 # 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 = 'LvWovRaJZlWCHql0bISuum8Bd2KX79mb'
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 result = {
185 'id': track_id,
186 'uploader': info.get('user', {}).get('username'),
187 'upload_date': unified_strdate(info.get('created_at')),
188 'title': info['title'],
189 'description': info.get('description'),
190 'thumbnail': thumbnail,
191 'duration': int_or_none(info.get('duration'), 1000),
192 'webpage_url': info.get('permalink_url'),
193 'license': info.get('license'),
194 }
195 formats = []
196 query = {'client_id': self._CLIENT_ID}
197 if secret_token is not None:
198 query['secret_token'] = secret_token
199 if info.get('downloadable', False):
200 # We can build a direct link to the song
201 format_url = update_url_query(
202 'https://api.soundcloud.com/tracks/%s/download' % track_id, query)
203 formats.append({
204 'format_id': 'download',
205 'ext': info.get('original_format', 'mp3'),
206 'url': format_url,
207 'vcodec': 'none',
208 'preference': 10,
209 })
210
211 # We have to retrieve the url
212 format_dict = self._download_json(
213 'https://api.soundcloud.com/i1/tracks/%s/streams' % track_id,
214 track_id, 'Downloading track url', query=query)
215
216 for key, stream_url in format_dict.items():
217 ext, abr = 'mp3', None
218 mobj = re.search(r'_([^_]+)_(\d+)_url', key)
219 if mobj:
220 ext, abr = mobj.groups()
221 abr = int(abr)
222 if key.startswith('http'):
223 stream_formats = [{
224 'format_id': key,
225 'ext': ext,
226 'url': stream_url,
227 }]
228 elif key.startswith('rtmp'):
229 # The url doesn't have an rtmp app, we have to extract the playpath
230 url, path = stream_url.split('mp3:', 1)
231 stream_formats = [{
232 'format_id': key,
233 'url': url,
234 'play_path': 'mp3:' + path,
235 'ext': 'flv',
236 }]
237 elif key.startswith('hls'):
238 stream_formats = self._extract_m3u8_formats(
239 stream_url, track_id, ext, entry_protocol='m3u8_native',
240 m3u8_id=key, fatal=False)
241 else:
242 continue
243
244 if abr:
245 for f in stream_formats:
246 f['abr'] = abr
247
248 formats.extend(stream_formats)
249
250 if not formats:
251 # We fallback to the stream_url in the original info, this
252 # cannot be always used, sometimes it can give an HTTP 404 error
253 formats.append({
254 'format_id': 'fallback',
255 'url': update_url_query(info['stream_url'], query),
256 'ext': 'mp3',
257 })
258
259 for f in formats:
260 f['vcodec'] = 'none'
261
262 self._check_formats(formats, track_id)
263 self._sort_formats(formats)
264 result['formats'] = formats
265
266 return result
267
268 def _real_extract(self, url):
269 mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
270 if mobj is None:
271 raise ExtractorError('Invalid URL: %s' % url)
272
273 track_id = mobj.group('track_id')
274
275 if track_id is not None:
276 info_json_url = 'https://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
277 full_title = track_id
278 token = mobj.group('secret_token')
279 if token:
280 info_json_url += '&secret_token=' + token
281 elif mobj.group('player'):
282 query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
283 real_url = query['url'][0]
284 # If the token is in the query of the original url we have to
285 # manually add it
286 if 'secret_token' in query:
287 real_url += '?secret_token=' + query['secret_token'][0]
288 return self.url_result(real_url)
289 else:
290 # extract uploader (which is in the url)
291 uploader = mobj.group('uploader')
292 # extract simple title (uploader + slug of song title)
293 slug_title = mobj.group('title')
294 token = mobj.group('token')
295 full_title = resolve_title = '%s/%s' % (uploader, slug_title)
296 if token:
297 resolve_title += '/%s' % token
298
299 self.report_resolve(full_title)
300
301 url = 'https://soundcloud.com/%s' % resolve_title
302 info_json_url = self._resolv_url(url)
303 info = self._download_json(info_json_url, full_title, 'Downloading info JSON')
304
305 return self._extract_info_dict(info, full_title, secret_token=token)
306
307
308 class SoundcloudPlaylistBaseIE(SoundcloudIE):
309 @staticmethod
310 def _extract_id(e):
311 return compat_str(e['id']) if e.get('id') else None
312
313 def _extract_track_entries(self, tracks):
314 return [
315 self.url_result(
316 track['permalink_url'], SoundcloudIE.ie_key(),
317 video_id=self._extract_id(track))
318 for track in tracks if track.get('permalink_url')]
319
320
321 class SoundcloudSetIE(SoundcloudPlaylistBaseIE):
322 _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/(?P<uploader>[\w\d-]+)/sets/(?P<slug_title>[\w\d-]+)(?:/(?P<token>[^?/]+))?'
323 IE_NAME = 'soundcloud:set'
324 _TESTS = [{
325 'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
326 'info_dict': {
327 'id': '2284613',
328 'title': 'The Royal Concept EP',
329 },
330 'playlist_mincount': 5,
331 }, {
332 'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
333 'only_matching': True,
334 }]
335
336 def _real_extract(self, url):
337 mobj = re.match(self._VALID_URL, url)
338
339 # extract uploader (which is in the url)
340 uploader = mobj.group('uploader')
341 # extract simple title (uploader + slug of song title)
342 slug_title = mobj.group('slug_title')
343 full_title = '%s/sets/%s' % (uploader, slug_title)
344 url = 'https://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
345
346 token = mobj.group('token')
347 if token:
348 full_title += '/' + token
349 url += '/' + token
350
351 self.report_resolve(full_title)
352
353 resolv_url = self._resolv_url(url)
354 info = self._download_json(resolv_url, full_title)
355
356 if 'errors' in info:
357 msgs = (compat_str(err['error_message']) for err in info['errors'])
358 raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
359
360 entries = self._extract_track_entries(info['tracks'])
361
362 return {
363 '_type': 'playlist',
364 'entries': entries,
365 'id': '%s' % info['id'],
366 'title': info['title'],
367 }
368
369
370 class SoundcloudPagedPlaylistBaseIE(SoundcloudPlaylistBaseIE):
371 _API_BASE = 'https://api.soundcloud.com'
372 _API_V2_BASE = 'https://api-v2.soundcloud.com'
373
374 def _extract_playlist(self, base_url, playlist_id, playlist_title):
375 COMMON_QUERY = {
376 'limit': 50,
377 'client_id': self._CLIENT_ID,
378 'linked_partitioning': '1',
379 }
380
381 query = COMMON_QUERY.copy()
382 query['offset'] = 0
383
384 next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
385
386 entries = []
387 for i in itertools.count():
388 response = self._download_json(
389 next_href, playlist_id, 'Downloading track page %s' % (i + 1))
390
391 collection = response['collection']
392 if not collection:
393 break
394
395 def resolve_permalink_url(candidates):
396 for cand in candidates:
397 if isinstance(cand, dict):
398 permalink_url = cand.get('permalink_url')
399 entry_id = self._extract_id(cand)
400 if permalink_url and permalink_url.startswith('http'):
401 return permalink_url, entry_id
402
403 for e in collection:
404 permalink_url, entry_id = resolve_permalink_url((e, e.get('track'), e.get('playlist')))
405 if permalink_url:
406 entries.append(self.url_result(permalink_url, video_id=entry_id))
407
408 next_href = response.get('next_href')
409 if not next_href:
410 break
411
412 parsed_next_href = compat_urlparse.urlparse(response['next_href'])
413 qs = compat_urlparse.parse_qs(parsed_next_href.query)
414 qs.update(COMMON_QUERY)
415 next_href = compat_urlparse.urlunparse(
416 parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
417
418 return {
419 '_type': 'playlist',
420 'id': playlist_id,
421 'title': playlist_title,
422 'entries': entries,
423 }
424
425
426 class SoundcloudUserIE(SoundcloudPagedPlaylistBaseIE):
427 _VALID_URL = r'''(?x)
428 https?://
429 (?:(?:www|m)\.)?soundcloud\.com/
430 (?P<user>[^/]+)
431 (?:/
432 (?P<rsrc>tracks|sets|reposts|likes|spotlight)
433 )?
434 /?(?:[?#].*)?$
435 '''
436 IE_NAME = 'soundcloud:user'
437 _TESTS = [{
438 'url': 'https://soundcloud.com/the-akashic-chronicler',
439 'info_dict': {
440 'id': '114582580',
441 'title': 'The Akashic Chronicler (All)',
442 },
443 'playlist_mincount': 74,
444 }, {
445 'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
446 'info_dict': {
447 'id': '114582580',
448 'title': 'The Akashic Chronicler (Tracks)',
449 },
450 'playlist_mincount': 37,
451 }, {
452 'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
453 'info_dict': {
454 'id': '114582580',
455 'title': 'The Akashic Chronicler (Playlists)',
456 },
457 'playlist_mincount': 2,
458 }, {
459 'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
460 'info_dict': {
461 'id': '114582580',
462 'title': 'The Akashic Chronicler (Reposts)',
463 },
464 'playlist_mincount': 7,
465 }, {
466 'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
467 'info_dict': {
468 'id': '114582580',
469 'title': 'The Akashic Chronicler (Likes)',
470 },
471 'playlist_mincount': 321,
472 }, {
473 'url': 'https://soundcloud.com/grynpyret/spotlight',
474 'info_dict': {
475 'id': '7098329',
476 'title': 'Grynpyret (Spotlight)',
477 },
478 'playlist_mincount': 1,
479 }]
480
481 _BASE_URL_MAP = {
482 'all': '%s/profile/soundcloud:users:%%s' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
483 'tracks': '%s/users/%%s/tracks' % SoundcloudPagedPlaylistBaseIE._API_BASE,
484 'sets': '%s/users/%%s/playlists' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
485 'reposts': '%s/profile/soundcloud:users:%%s/reposts' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
486 'likes': '%s/users/%%s/likes' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
487 'spotlight': '%s/users/%%s/spotlight' % SoundcloudPagedPlaylistBaseIE._API_V2_BASE,
488 }
489
490 _TITLE_MAP = {
491 'all': 'All',
492 'tracks': 'Tracks',
493 'sets': 'Playlists',
494 'reposts': 'Reposts',
495 'likes': 'Likes',
496 'spotlight': 'Spotlight',
497 }
498
499 def _real_extract(self, url):
500 mobj = re.match(self._VALID_URL, url)
501 uploader = mobj.group('user')
502
503 url = 'https://soundcloud.com/%s/' % uploader
504 resolv_url = self._resolv_url(url)
505 user = self._download_json(
506 resolv_url, uploader, 'Downloading user info')
507
508 resource = mobj.group('rsrc') or 'all'
509
510 return self._extract_playlist(
511 self._BASE_URL_MAP[resource] % user['id'], compat_str(user['id']),
512 '%s (%s)' % (user['username'], self._TITLE_MAP[resource]))
513
514
515 class SoundcloudTrackStationIE(SoundcloudPagedPlaylistBaseIE):
516 _VALID_URL = r'https?://(?:(?:www|m)\.)?soundcloud\.com/stations/track/[^/]+/(?P<id>[^/?#&]+)'
517 IE_NAME = 'soundcloud:trackstation'
518 _TESTS = [{
519 'url': 'https://soundcloud.com/stations/track/officialsundial/your-text',
520 'info_dict': {
521 'id': '286017854',
522 'title': 'Track station: your-text',
523 },
524 'playlist_mincount': 47,
525 }]
526
527 def _real_extract(self, url):
528 track_name = self._match_id(url)
529
530 webpage = self._download_webpage(url, track_name)
531
532 track_id = self._search_regex(
533 r'soundcloud:track-stations:(\d+)', webpage, 'track id')
534
535 return self._extract_playlist(
536 '%s/stations/soundcloud:track-stations:%s/tracks'
537 % (self._API_V2_BASE, track_id),
538 track_id, 'Track station: %s' % track_name)
539
540
541 class SoundcloudPlaylistIE(SoundcloudPlaylistBaseIE):
542 _VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
543 IE_NAME = 'soundcloud:playlist'
544 _TESTS = [{
545 'url': 'https://api.soundcloud.com/playlists/4110309',
546 'info_dict': {
547 'id': '4110309',
548 'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
549 'description': 're:.*?TILT Brass - Bowery Poetry Club',
550 },
551 'playlist_count': 6,
552 }]
553
554 def _real_extract(self, url):
555 mobj = re.match(self._VALID_URL, url)
556 playlist_id = mobj.group('id')
557 base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
558
559 data_dict = {
560 'client_id': self._CLIENT_ID,
561 }
562 token = mobj.group('token')
563
564 if token:
565 data_dict['secret_token'] = token
566
567 data = compat_urllib_parse_urlencode(data_dict)
568 data = self._download_json(
569 base_url + data, playlist_id, 'Downloading playlist')
570
571 entries = self._extract_track_entries(data['tracks'])
572
573 return {
574 '_type': 'playlist',
575 'id': playlist_id,
576 'title': data.get('title'),
577 'description': data.get('description'),
578 'entries': entries,
579 }
580
581
582 class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
583 IE_NAME = 'soundcloud:search'
584 IE_DESC = 'Soundcloud search'
585 _MAX_RESULTS = float('inf')
586 _TESTS = [{
587 'url': 'scsearch15:post-avant jazzcore',
588 'info_dict': {
589 'title': 'post-avant jazzcore',
590 },
591 'playlist_count': 15,
592 }]
593
594 _SEARCH_KEY = 'scsearch'
595 _MAX_RESULTS_PER_PAGE = 200
596 _DEFAULT_RESULTS_PER_PAGE = 50
597 _API_V2_BASE = 'https://api-v2.soundcloud.com'
598
599 def _get_collection(self, endpoint, collection_id, **query):
600 limit = min(
601 query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
602 self._MAX_RESULTS_PER_PAGE)
603 query['limit'] = limit
604 query['client_id'] = self._CLIENT_ID
605 query['linked_partitioning'] = '1'
606 query['offset'] = 0
607 data = compat_urllib_parse_urlencode(query)
608 next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
609
610 collected_results = 0
611
612 for i in itertools.count(1):
613 response = self._download_json(
614 next_url, collection_id, 'Downloading page {0}'.format(i),
615 'Unable to download API page')
616
617 collection = response.get('collection', [])
618 if not collection:
619 break
620
621 collection = list(filter(bool, collection))
622 collected_results += len(collection)
623
624 for item in collection:
625 yield self.url_result(item['uri'], SoundcloudIE.ie_key())
626
627 if not collection or collected_results >= limit:
628 break
629
630 next_url = response.get('next_href')
631 if not next_url:
632 break
633
634 def _get_n_results(self, query, n):
635 tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
636 return self.playlist_result(tracks, playlist_title=query)