youtube-dl/youtube_dl/extractor/soundcloud.py

600 lines
21 KiB
Python
Raw Normal View History

2016-10-02 13:39:18 +02:00
# coding: utf-8
from __future__ import unicode_literals
import itertools
import re
2013-06-23 20:57:44 +02:00
from .common import (
InfoExtractor,
SearchInfoExtractor
)
from ..compat import (
2013-06-23 20:57:44 +02:00
compat_str,
compat_urlparse,
compat_urllib_parse_urlencode,
)
from ..utils import (
2013-06-23 20:57:44 +02:00
ExtractorError,
int_or_none,
2013-06-23 20:57:44 +02:00
unified_strdate,
update_url_query)
2013-06-23 20:57:44 +02:00
class SoundcloudIE(InfoExtractor):
"""Information extractor for soundcloud.com
To access the media, the uid of the song and a stream token
must be extracted from the page source and the script must make
a request to media.soundcloud.com/crossdomain.xml. Then
the media can be grabbed by requesting from an url composed
of the stream token and uid
"""
_VALID_URL = r'''(?x)^(?:https?://)?
(?:(?:(?:www\.|m\.)?soundcloud\.com/
(?!stations/track)
(?P<uploader>[\w\d-]+)/
(?!(?:tracks|sets(?:/.+?)?|reposts|likes|spotlight)/?(?:$|[?#]))
(?P<title>[\w\d-]+)/?
(?P<token>[^?]+?)?(?:[?].*)?$)
|(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+)
(?:/?\?secret_token=(?P<secret_token>[^&]+))?)
|(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
)
'''
IE_NAME = 'soundcloud'
_TESTS = [
{
'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
'info_dict': {
2014-10-25 22:32:01 +02:00
'id': '62986583',
'ext': 'mp3',
'upload_date': '20121011',
'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',
'uploader': 'E.T. ExTerrestrial Music',
'title': 'Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1',
'duration': 143,
2016-09-18 11:53:05 +02:00
'license': 'all-rights-reserved',
}
},
# not streamable song
{
'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
'info_dict': {
'id': '47127627',
'ext': 'mp3',
'title': 'Goldrushed',
'description': 'From Stockholm Sweden\r\nPovel / Magnus / Filip / David\r\nwww.theroyalconcept.com',
'uploader': 'The Royal Concept',
'upload_date': '20120521',
'duration': 227,
2016-09-18 11:53:05 +02:00
'license': 'all-rights-reserved',
},
'params': {
# rtmp
'skip_download': True,
},
},
# private link
{
'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
'info_dict': {
'id': '123998367',
'ext': 'mp3',
'title': 'Youtube - Dl Test Video \'\' Ä↭',
'uploader': 'jaimeMF',
'description': 'test chars: \"\'/\\ä↭',
'upload_date': '20131209',
'duration': 9,
2016-09-18 11:53:05 +02:00
'license': 'all-rights-reserved',
},
},
# private link (alt format)
{
'url': 'https://api.soundcloud.com/tracks/123998367?secret_token=s-8Pjrp',
'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
'info_dict': {
'id': '123998367',
'ext': 'mp3',
'title': 'Youtube - Dl Test Video \'\' Ä↭',
'uploader': 'jaimeMF',
'description': 'test chars: \"\'/\\ä↭',
'upload_date': '20131209',
'duration': 9,
2016-09-18 11:53:05 +02:00
'license': 'all-rights-reserved',
},
},
# downloadable song
{
2014-07-15 14:18:06 +02:00
'url': 'https://soundcloud.com/oddsamples/bus-brakes',
2014-07-23 01:41:44 +02:00
'md5': '7624f2351f8a3b2e7cd51522496e7631',
'info_dict': {
2014-07-15 14:18:06 +02:00
'id': '128590877',
2014-07-23 01:41:44 +02:00
'ext': 'mp3',
2014-07-15 14:18:06 +02:00
'title': 'Bus Brakes',
2014-10-25 22:32:01 +02:00
'description': 'md5:0053ca6396e8d2fd7b7e1595ef12ab66',
2014-07-15 14:18:06 +02:00
'uploader': 'oddsamples',
'upload_date': '20140109',
'duration': 17,
2016-09-18 11:53:05 +02:00
'license': 'cc-by-sa',
},
},
]
2013-06-23 20:57:44 +02:00
2017-07-29 13:45:57 +02:00
_CLIENT_ID = 'JlZIsxg2hY5WnBgtn3jfS0UYCl0K8DOg'
_IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
@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)]
2013-06-23 20:57:44 +02:00
def report_resolve(self, video_id):
"""Report information extraction."""
2014-03-24 02:15:31 +01:00
self.to_screen('%s: Resolving id' % video_id)
2013-06-23 20:57:44 +02:00
@classmethod
def _resolv_url(cls, url):
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)
2016-09-20 16:55:57 +02:00
thumbnail = info.get('artwork_url')
if isinstance(thumbnail, compat_str):
thumbnail = thumbnail.replace('-large', '-t500x500')
ext = 'mp3'
result = {
'id': track_id,
2016-09-20 16:55:57 +02:00
'uploader': info.get('user', {}).get('username'),
'upload_date': unified_strdate(info.get('created_at')),
'title': info['title'],
2016-09-20 16:55:57 +02:00
'description': info.get('description'),
'thumbnail': thumbnail,
'duration': int_or_none(info.get('duration'), 1000),
'webpage_url': info.get('permalink_url'),
2016-09-20 16:55:57 +02:00
'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 = update_url_query(
'https://api.soundcloud.com/tracks/{0}/download'.format(track_id), query)
formats.append({
'format_id': 'download',
'ext': info.get('original_format', 'mp3'),
'url': format_url,
'vcodec': 'none',
'preference': 10,
})
# We have to retrieve the url
format_dict = self._download_json(
'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():
abr = int_or_none(self._search_regex(
r'_(\d+)_url', key, 'audio bitrate', default=None))
if key.startswith('http'):
stream_formats = [{
'format_id': key,
'ext': ext,
'url': stream_url,
}]
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)
stream_formats = [{
'format_id': key,
'url': url,
'play_path': 'mp3:' + path,
2015-03-02 17:47:07 +01:00
'ext': 'flv',
}]
2017-01-27 17:08:32 +01:00
elif key.startswith('hls'):
stream_formats = self._extract_m3u8_formats(
2017-01-27 17:08:32 +01:00
stream_url, track_id, 'mp3', entry_protocol='m3u8_native',
m3u8_id=key, fatal=False)
else:
continue
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': ext,
})
for f in formats:
f['vcodec'] = 'none'
self._check_formats(formats, track_id)
self._sort_formats(formats)
result['formats'] = formats
return result
2013-06-23 20:57:44 +02:00
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
2013-06-23 20:57:44 +02:00
if mobj is None:
2014-03-24 02:15:31 +01:00
raise ExtractorError('Invalid URL: %s' % url)
2013-06-23 20:57:44 +02:00
track_id = mobj.group('track_id')
2016-09-18 11:53:05 +02:00
if track_id is not None:
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:
2016-02-14 10:37:17 +01:00
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]
# If the token is in the query of the original url we have to
# manually add it
if 'secret_token' in query:
real_url += '?secret_token=' + query['secret_token'][0]
return self.url_result(real_url)
else:
# extract uploader (which is in the url)
uploader = mobj.group('uploader')
# extract simple title (uploader + slug of song title)
2014-11-23 21:20:46 +01:00
slug_title = mobj.group('title')
token = mobj.group('token')
full_title = resolve_title = '%s/%s' % (uploader, slug_title)
if token:
resolve_title += '/%s' % token
2014-11-23 20:41:03 +01:00
self.report_resolve(full_title)
2014-11-23 20:41:03 +01:00
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')
2013-06-23 20:57:44 +02:00
return self._extract_info_dict(info, full_title, secret_token=token)
2013-06-23 20:57:44 +02:00
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 = [{
'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep',
'info_dict': {
2015-02-01 15:24:38 +01:00
'id': '2284613',
'title': 'The Royal Concept EP',
},
2017-06-27 16:26:46 +02:00
'playlist_mincount': 5,
}, {
'url': 'https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep/token',
'only_matching': True,
}]
2013-06-23 20:57:44 +02:00
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
# extract uploader (which is in the url)
uploader = mobj.group('uploader')
2013-06-23 20:57:44 +02:00
# extract simple title (uploader + slug of song title)
slug_title = mobj.group('slug_title')
2013-06-23 20:57:44 +02:00
full_title = '%s/sets/%s' % (uploader, slug_title)
url = 'https://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
token = mobj.group('token')
if token:
full_title += '/' + token
url += '/' + token
2013-06-23 20:57:44 +02:00
self.report_resolve(full_title)
resolv_url = self._resolv_url(url)
info = self._download_json(resolv_url, full_title)
2013-06-23 20:57:44 +02:00
if 'errors' in info:
msgs = (compat_str(err['error_message']) for err in info['errors'])
raise ExtractorError('unable to download video webpage: %s' % ','.join(msgs))
2013-06-23 20:57:44 +02:00
entries = self._extract_track_entries(info['tracks'])
return {
'_type': 'playlist',
'entries': entries,
2015-02-01 15:24:38 +01:00
'id': '%s' % info['id'],
'title': info['title'],
}
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):
2015-07-31 23:50:03 +02:00
_VALID_URL = r'''(?x)
https?://
(?:(?:www|m)\.)?soundcloud\.com/
(?P<user>[^/]+)
(?:/
(?P<rsrc>tracks|sets|reposts|likes|spotlight)
)?
/?(?:[?#].*)?$
'''
IE_NAME = 'soundcloud:user'
_TESTS = [{
'url': 'https://soundcloud.com/the-akashic-chronicler',
'info_dict': {
'id': '114582580',
'title': 'The Akashic Chronicler (All)',
},
'playlist_mincount': 74,
}, {
'url': 'https://soundcloud.com/the-akashic-chronicler/tracks',
'info_dict': {
'id': '114582580',
'title': 'The Akashic Chronicler (Tracks)',
},
'playlist_mincount': 37,
2015-06-24 16:49:22 +02:00
}, {
'url': 'https://soundcloud.com/the-akashic-chronicler/sets',
'info_dict': {
'id': '114582580',
'title': 'The Akashic Chronicler (Playlists)',
},
'playlist_mincount': 2,
}, {
'url': 'https://soundcloud.com/the-akashic-chronicler/reposts',
'info_dict': {
'id': '114582580',
'title': 'The Akashic Chronicler (Reposts)',
},
2015-08-22 02:30:00 +02:00
'playlist_mincount': 7,
}, {
'url': 'https://soundcloud.com/the-akashic-chronicler/likes',
'info_dict': {
'id': '114582580',
'title': 'The Akashic Chronicler (Likes)',
},
2015-08-22 02:30:00 +02:00
'playlist_mincount': 321,
}, {
'url': 'https://soundcloud.com/grynpyret/spotlight',
'info_dict': {
'id': '7098329',
2017-06-27 16:26:46 +02:00
'title': 'Grynpyret (Spotlight)',
},
'playlist_mincount': 1,
}]
_BASE_URL_MAP = {
'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 = {
'all': 'All',
'tracks': 'Tracks',
'sets': 'Playlists',
'reposts': 'Reposts',
'likes': 'Likes',
'spotlight': 'Spotlight',
}
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
uploader = mobj.group('user')
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'
return self._extract_playlist(
self._BASE_URL_MAP[resource] % user['id'], compat_str(user['id']),
'%s (%s)' % (user['username'], self._TITLE_MAP[resource]))
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 _real_extract(self, url):
track_name = self._match_id(url)
webpage = self._download_webpage(url, track_name)
track_id = self._search_regex(
r'soundcloud:track-stations:(\d+)', webpage, 'track id')
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(SoundcloudPlaylistBaseIE):
2014-09-18 18:57:04 +02:00
_VALID_URL = r'https?://api\.soundcloud\.com/playlists/(?P<id>[0-9]+)(?:/?\?secret_token=(?P<token>[^&]+?))?$'
IE_NAME = 'soundcloud:playlist'
2014-09-18 18:57:04 +02:00
_TESTS = [{
'url': 'https://api.soundcloud.com/playlists/4110309',
2014-09-18 18:57:04 +02:00
'info_dict': {
'id': '4110309',
'title': 'TILT Brass - Bowery Poetry Club, August \'03 [Non-Site SCR 02]',
'description': 're:.*?TILT Brass - Bowery Poetry Club',
},
'playlist_count': 6,
}]
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
playlist_id = mobj.group('id')
base_url = '%s//api.soundcloud.com/playlists/%s.json?' % (self.http_scheme(), playlist_id)
data_dict = {
'client_id': self._CLIENT_ID,
}
token = mobj.group('token')
if token:
data_dict['secret_token'] = token
data = compat_urllib_parse_urlencode(data_dict)
data = self._download_json(
base_url + data, playlist_id, 'Downloading playlist')
entries = self._extract_track_entries(data['tracks'])
return {
'_type': 'playlist',
'id': playlist_id,
'title': data.get('title'),
'description': data.get('description'),
'entries': entries,
}
class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
IE_NAME = 'soundcloud:search'
IE_DESC = 'Soundcloud search'
_MAX_RESULTS = float('inf')
_TESTS = [{
'url': 'scsearch15:post-avant jazzcore',
'info_dict': {
'title': 'post-avant jazzcore',
},
'playlist_count': 15,
}]
_SEARCH_KEY = 'scsearch'
_MAX_RESULTS_PER_PAGE = 200
_DEFAULT_RESULTS_PER_PAGE = 50
_API_V2_BASE = 'https://api-v2.soundcloud.com'
def _get_collection(self, endpoint, collection_id, **query):
2015-11-21 19:49:58 +01:00
limit = min(
query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
self._MAX_RESULTS_PER_PAGE)
query['limit'] = limit
query['client_id'] = self._CLIENT_ID
query['linked_partitioning'] = '1'
query['offset'] = 0
data = compat_urllib_parse_urlencode(query)
next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
collected_results = 0
for i in itertools.count(1):
2015-11-21 15:04:35 +01:00
response = self._download_json(
next_url, collection_id, 'Downloading page {0}'.format(i),
2015-11-21 15:04:35 +01:00
'Unable to download API page')
collection = response.get('collection', [])
if not collection:
break
collection = list(filter(bool, collection))
collected_results += len(collection)
for item in collection:
yield self.url_result(item['uri'], SoundcloudIE.ie_key())
if not collection or collected_results >= limit:
break
2015-11-21 15:04:35 +01:00
next_url = response.get('next_href')
if not next_url:
break
def _get_n_results(self, query, n):
tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
return self.playlist_result(tracks, playlist_title=query)