]> jfr.im git - yt-dlp.git/blame - yt_dlp/extractor/rokfin.py
projects / yt-dlp.git / blame
? search:
summary | shortlog | log | commit | commitdiff | tree
blob | blame (incremental) | history | HEAD
[extractor] Deprecate `_sort_formats`
[yt-dlp.git] / yt_dlp / extractor / rokfin.py
CommitLineData
334b1c48 1import itertools
c9b2b368 2import json
e037c405 3import re
4import urllib.parse
334b1c48 5from datetime import datetime
6
c9b2b368 7from .common import InfoExtractor, SearchInfoExtractor
334b1c48 8from ..utils import (
7eaf7f9a 9 ExtractorError,
e037c405 10 determine_ext,
334b1c48 11 float_or_none,
12 format_field,
13 int_or_none,
14 str_or_none,
15 traverse_obj,
e037c405 16 try_get,
17 unescapeHTML,
334b1c48 18 unified_timestamp,
19 url_or_none,
e037c405 20 urlencode_postdata,
334b1c48 21)
22
7eaf7f9a 23_API_BASE_URL = 'https://prod-api-v2.production.rokfin.com/api/v2/public/'
24
25
334b1c48 26class RokfinIE(InfoExtractor):
27 _VALID_URL = r'https?://(?:www\.)?rokfin\.com/(?P<id>(?P<type>post|stream)/\d+)'
e037c405 28 _NETRC_MACHINE = 'rokfin'
29 _AUTH_BASE = 'https://secure.rokfin.com/auth/realms/rokfin-web/protocol/openid-connect'
30 _access_mgmt_tokens = {} # OAuth 2.0: RFC 6749, Sec. 1.4-5
334b1c48 31 _TESTS = [{
32 'url': 'https://www.rokfin.com/post/57548/Mitt-Romneys-Crazy-Solution-To-Climate-Change',
33 'info_dict': {
34 'id': 'post/57548',
35 'ext': 'mp4',
36 'title': 'Mitt Romney\'s Crazy Solution To Climate Change',
37 'thumbnail': r're:https://img\.production\.rokfin\.com/.+',
38 'upload_date': '20211023',
39 'timestamp': 1634998029,
40 'channel': 'Jimmy Dore',
41 'channel_id': 65429,
42 'channel_url': 'https://rokfin.com/TheJimmyDoreShow',
43 'duration': 213.0,
44 'availability': 'public',
45 'live_status': 'not_live',
46 'dislike_count': int,
47 'like_count': int,
48 }
49 }, {
50 'url': 'https://rokfin.com/post/223/Julian-Assange-Arrested-Streaming-In-Real-Time',
51 'info_dict': {
52 'id': 'post/223',
53 'ext': 'mp4',
54 'title': 'Julian Assange Arrested: Streaming In Real Time',
55 'thumbnail': r're:https://img\.production\.rokfin\.com/.+',
56 'upload_date': '20190412',
57 'timestamp': 1555052644,
58 'channel': 'Ron Placone',
59 'channel_id': 10,
60 'channel_url': 'https://rokfin.com/RonPlacone',
61 'availability': 'public',
62 'live_status': 'not_live',
63 'dislike_count': int,
64 'like_count': int,
65 'tags': ['FreeThinkingMedia^', 'RealProgressives^'],
66 }
67 }, {
68 'url': 'https://www.rokfin.com/stream/10543/Its-A-Crazy-Mess-Regional-Director-Blows-Whistle-On-Pfizers-Vaccine-Trial-Data',
69 'info_dict': {
70 'id': 'stream/10543',
71 'ext': 'mp4',
72 'title': '"It\'s A Crazy Mess" Regional Director Blows Whistle On Pfizer\'s Vaccine Trial Data',
73 'thumbnail': r're:https://img\.production\.rokfin\.com/.+',
74 'description': 'md5:324ce2d3e3b62e659506409e458b9d8e',
75 'channel': 'Ryan Cristián',
76 'channel_id': 53856,
77 'channel_url': 'https://rokfin.com/TLAVagabond',
78 'availability': 'public',
79 'is_live': False,
80 'was_live': True,
81 'live_status': 'was_live',
82 'timestamp': 1635874720,
83 'release_timestamp': 1635874720,
84 'release_date': '20211102',
85 'upload_date': '20211102',
86 'dislike_count': int,
87 'like_count': int,
88 'tags': ['FreeThinkingMedia^'],
89 }
90 }]
91
92 def _real_extract(self, url):
93 video_id, video_type = self._match_valid_url(url).group('id', 'type')
e037c405 94 metadata = self._download_json_using_access_token(f'{_API_BASE_URL}{video_id}', video_id)
334b1c48 95
96 scheduled = unified_timestamp(metadata.get('scheduledAt'))
97 live_status = ('was_live' if metadata.get('stoppedAt')
98 else 'is_upcoming' if scheduled
99 else 'is_live' if video_type == 'stream'
100 else 'not_live')
101
102 video_url = traverse_obj(metadata, 'url', ('content', 'contentUrl'), expected_type=url_or_none)
103 formats, subtitles = [{'url': video_url}] if video_url else [], {}
104 if determine_ext(video_url) == 'm3u8':
105 formats, subtitles = self._extract_m3u8_formats_and_subtitles(
106 video_url, video_id, fatal=False, live=live_status == 'is_live')
107
108 if not formats:
9461cb58 109 if traverse_obj(metadata, 'premiumPlan', 'premium'):
334b1c48 110 self.raise_login_required('This video is only available to premium users', True, method='cookies')
111 elif scheduled:
112 self.raise_no_formats(
962ffcf8 113 f'Stream is offline; scheduled for {datetime.fromtimestamp(scheduled).strftime("%Y-%m-%d %H:%M:%S")}',
334b1c48 114 video_id=video_id, expected=True)
334b1c48 115
116 uploader = traverse_obj(metadata, ('createdBy', 'username'), ('creator', 'username'))
117 timestamp = (scheduled or float_or_none(metadata.get('postedAtMilli'), 1000)
118 or unified_timestamp(metadata.get('creationDateTime')))
119 return {
120 'id': video_id,
121 'formats': formats,
122 'subtitles': subtitles,
123 'title': str_or_none(traverse_obj(metadata, 'title', ('content', 'contentTitle'))),
124 'duration': float_or_none(traverse_obj(metadata, ('content', 'duration'))),
125 'thumbnail': url_or_none(traverse_obj(metadata, 'thumbnail', ('content', 'thumbnailUrl1'))),
126 'description': str_or_none(traverse_obj(metadata, 'description', ('content', 'contentDescription'))),
127 'like_count': int_or_none(metadata.get('likeCount')),
128 'dislike_count': int_or_none(metadata.get('dislikeCount')),
129 'channel': str_or_none(traverse_obj(metadata, ('createdBy', 'name'), ('creator', 'name'))),
130 'channel_id': traverse_obj(metadata, ('createdBy', 'id'), ('creator', 'id')),
131 'channel_url': url_or_none(f'https://rokfin.com/{uploader}') if uploader else None,
132 'timestamp': timestamp,
133 'release_timestamp': timestamp if live_status != 'not_live' else None,
134 'tags': traverse_obj(metadata, ('tags', ..., 'title'), expected_type=str_or_none),
135 'live_status': live_status,
136 'availability': self._availability(
9461cb58 137 needs_premium=bool(traverse_obj(metadata, 'premiumPlan', 'premium')),
334b1c48 138 is_private=False, needs_subscription=False, needs_auth=False, is_unlisted=False),
139 # 'comment_count': metadata.get('numComments'), # Data provided by website is wrong
140 '__post_extractor': self.extract_comments(video_id) if video_type == 'post' else None,
141 }
142
143 def _get_comments(self, video_id):
144 pages_total = None
145 for page_n in itertools.count():
146 raw_comments = self._download_json(
7eaf7f9a 147 f'{_API_BASE_URL}comment?postId={video_id[5:]}&page={page_n}&size=50',
a70635b8 148 video_id, note=f'Downloading viewer comments page {page_n + 1}{format_field(pages_total, None, " of %s")}',
334b1c48 149 fatal=False) or {}
150
151 for comment in raw_comments.get('content') or []:
152 yield {
153 'text': str_or_none(comment.get('comment')),
154 'author': str_or_none(comment.get('name')),
155 'id': comment.get('commentId'),
156 'author_id': comment.get('userId'),
157 'parent': 'root',
158 'like_count': int_or_none(comment.get('numLikes')),
159 'dislike_count': int_or_none(comment.get('numDislikes')),
160 'timestamp': unified_timestamp(comment.get('postedAt'))
161 }
162
7eaf7f9a 163 pages_total = int_or_none(raw_comments.get('totalPages')) or None
164 is_last = raw_comments.get('last')
165 if not raw_comments.get('content') or is_last or (page_n > pages_total if pages_total else is_last is not False):
334b1c48 166 return
7eaf7f9a 167
e037c405 168 def _perform_login(self, username, password):
169 # https://openid.net/specs/openid-connect-core-1_0.html#CodeFlowAuth (Sec. 3.1)
170 login_page = self._download_webpage(
171 f'{self._AUTH_BASE}/auth?client_id=web&redirect_uri=https%3A%2F%2Frokfin.com%2Ffeed&response_mode=fragment&response_type=code&scope=openid',
172 None, note='loading login page', errnote='error loading login page')
173 authentication_point_url = unescapeHTML(self._search_regex(
174 r'<form\s+[^>]+action\s*=\s*"(https://secure\.rokfin\.com/auth/realms/rokfin-web/login-actions/authenticate\?[^"]+)"',
175 login_page, name='Authentication URL'))
176
177 resp_body = self._download_webpage(
178 authentication_point_url, None, note='logging in', fatal=False, expected_status=404,
179 data=urlencode_postdata({'username': username, 'password': password, 'rememberMe': 'off', 'credentialId': ''}))
180 if not self._authentication_active():
181 if re.search(r'(?i)(invalid\s+username\s+or\s+password)', resp_body or ''):
182 raise ExtractorError('invalid username/password', expected=True)
183 raise ExtractorError('Login failed')
184
185 urlh = self._request_webpage(
186 f'{self._AUTH_BASE}/auth', None,
187 note='granting user authorization', errnote='user authorization rejected by Rokfin',
188 query={
189 'client_id': 'web',
190 'prompt': 'none',
191 'redirect_uri': 'https://rokfin.com/silent-check-sso.html',
192 'response_mode': 'fragment',
193 'response_type': 'code',
194 'scope': 'openid',
195 })
196 self._access_mgmt_tokens = self._download_json(
197 f'{self._AUTH_BASE}/token', None,
198 note='getting access credentials', errnote='error getting access credentials',
199 data=urlencode_postdata({
200 'code': urllib.parse.parse_qs(urllib.parse.urldefrag(urlh.geturl()).fragment).get('code')[0],
201 'client_id': 'web',
202 'grant_type': 'authorization_code',
203 'redirect_uri': 'https://rokfin.com/silent-check-sso.html'
204 }))
205
206 def _authentication_active(self):
207 return not (
208 {'KEYCLOAK_IDENTITY', 'KEYCLOAK_IDENTITY_LEGACY', 'KEYCLOAK_SESSION', 'KEYCLOAK_SESSION_LEGACY'}
209 - set(self._get_cookies(self._AUTH_BASE)))
210
211 def _get_auth_token(self):
212 return try_get(self._access_mgmt_tokens, lambda x: ' '.join([x['token_type'], x['access_token']]))
213
214 def _download_json_using_access_token(self, url_or_request, video_id, headers={}, query={}):
215 assert 'authorization' not in headers
216 headers = headers.copy()
217 auth_token = self._get_auth_token()
218 refresh_token = self._access_mgmt_tokens.get('refresh_token')
219 if auth_token:
220 headers['authorization'] = auth_token
221
222 json_string, urlh = self._download_webpage_handle(
223 url_or_request, video_id, headers=headers, query=query, expected_status=401)
224 if not auth_token or urlh.code != 401 or refresh_token is None:
225 return self._parse_json(json_string, video_id)
226
227 self._access_mgmt_tokens = self._download_json(
228 f'{self._AUTH_BASE}/token', video_id,
229 note='User authorization expired or canceled by Rokfin. Re-authorizing ...', errnote='Failed to re-authorize',
230 data=urlencode_postdata({
231 'grant_type': 'refresh_token',
232 'refresh_token': refresh_token,
233 'client_id': 'web'
234 }))
235 headers['authorization'] = self._get_auth_token()
236 if headers['authorization'] is None:
237 raise ExtractorError('User authorization lost', expected=True)
238
239 return self._download_json(url_or_request, video_id, headers=headers, query=query)
240
7eaf7f9a 241
242class RokfinPlaylistBaseIE(InfoExtractor):
243 _TYPES = {
244 'video': 'post',
245 'audio': 'post',
246 'stream': 'stream',
247 'dead_stream': 'stream',
248 'stack': 'stack',
249 }
250
251 def _get_video_data(self, metadata):
252 for content in metadata.get('content') or []:
253 media_type = self._TYPES.get(content.get('mediaType'))
254 video_id = content.get('id') if media_type == 'post' else content.get('mediaId')
255 if not media_type or not video_id:
256 continue
257
258 yield self.url_result(f'https://rokfin.com/{media_type}/{video_id}', video_id=f'{media_type}/{video_id}',
259 video_title=str_or_none(traverse_obj(content, ('content', 'contentTitle'))))
260
261
262class RokfinStackIE(RokfinPlaylistBaseIE):
263 IE_NAME = 'rokfin:stack'
e037c405 264 IE_DESC = 'Rokfin Stacks'
7eaf7f9a 265 _VALID_URL = r'https?://(?:www\.)?rokfin\.com/stack/(?P<id>[^/]+)'
266 _TESTS = [{
267 'url': 'https://www.rokfin.com/stack/271/Tulsi-Gabbard-Portsmouth-Townhall-FULL--Feb-9-2020',
268 'playlist_count': 8,
269 'info_dict': {
270 'id': '271',
271 },
272 }]
273
274 def _real_extract(self, url):
275 list_id = self._match_id(url)
276 return self.playlist_result(self._get_video_data(
277 self._download_json(f'{_API_BASE_URL}stack/{list_id}', list_id)), list_id)
278
279
280class RokfinChannelIE(RokfinPlaylistBaseIE):
281 IE_NAME = 'rokfin:channel'
e037c405 282 IE_DESC = 'Rokfin Channels'
7eaf7f9a 283 _VALID_URL = r'https?://(?:www\.)?rokfin\.com/(?!((feed/?)|(discover/?)|(channels/?))$)(?P<id>[^/]+)/?$'
284 _TESTS = [{
285 'url': 'https://rokfin.com/TheConvoCouch',
286 'playlist_mincount': 100,
287 'info_dict': {
288 'id': '12071-new',
289 'title': 'TheConvoCouch - New',
290 'description': 'md5:bb622b1bca100209b91cd685f7847f06',
291 },
292 }]
293
294 _TABS = {
295 'new': 'posts',
296 'top': 'top',
297 'videos': 'video',
298 'podcasts': 'audio',
299 'streams': 'stream',
300 'stacks': 'stack',
301 }
302
303 def _real_initialize(self):
304 self._validate_extractor_args()
305
306 def _validate_extractor_args(self):
307 requested_tabs = self._configuration_arg('tab', None)
308 if requested_tabs is not None and (len(requested_tabs) > 1 or requested_tabs[0] not in self._TABS):
309 raise ExtractorError(f'Invalid extractor-arg "tab". Must be one of {", ".join(self._TABS)}', expected=True)
310
311 def _entries(self, channel_id, channel_name, tab):
312 pages_total = None
313 for page_n in itertools.count(0):
314 if tab in ('posts', 'top'):
315 data_url = f'{_API_BASE_URL}user/{channel_name}/{tab}?page={page_n}&size=50'
316 else:
317 data_url = f'{_API_BASE_URL}post/search/{tab}?page={page_n}&size=50&creator={channel_id}'
318 metadata = self._download_json(
319 data_url, channel_name,
a70635b8 320 note=f'Downloading video metadata page {page_n + 1}{format_field(pages_total, None, " of %s")}')
7eaf7f9a 321
322 yield from self._get_video_data(metadata)
323 pages_total = int_or_none(metadata.get('totalPages')) or None
324 is_last = metadata.get('last')
325 if is_last or (page_n > pages_total if pages_total else is_last is not False):
326 return
327
328 def _real_extract(self, url):
329 channel_name = self._match_id(url)
330 channel_info = self._download_json(f'{_API_BASE_URL}user/{channel_name}', channel_name)
331 channel_id = channel_info['id']
332 tab = self._configuration_arg('tab', default=['new'])[0]
333
334 return self.playlist_result(
335 self._entries(channel_id, channel_name, self._TABS[tab]),
336 f'{channel_id}-{tab}', f'{channel_name} - {tab.title()}', str_or_none(channel_info.get('description')))
c9b2b368 337
338
339class RokfinSearchIE(SearchInfoExtractor):
340 IE_NAME = 'rokfin:search'
341 IE_DESC = 'Rokfin Search'
342 _SEARCH_KEY = 'rkfnsearch'
343 _TYPES = {
344 'video': (('id', 'raw'), 'post'),
345 'audio': (('id', 'raw'), 'post'),
346 'stream': (('content_id', 'raw'), 'stream'),
347 'dead_stream': (('content_id', 'raw'), 'stream'),
348 'stack': (('content_id', 'raw'), 'stack'),
349 }
350 _TESTS = [{
351 'url': 'rkfnsearch5:"zelenko"',
352 'playlist_count': 5,
353 'info_dict': {
354 'id': '"zelenko"',
355 'title': '"zelenko"',
356 }
357 }]
358 _db_url = None
359 _db_access_key = None
360
361 def _real_initialize(self):
9809740b 362 self._db_url, self._db_access_key = self.cache.load(self.ie_key(), 'auth', default=(None, None))
c9b2b368 363 if not self._db_url:
364 self._get_db_access_credentials()
365
366 def _search_results(self, query):
367 total_pages = None
368 for page_number in itertools.count(1):
369 search_results = self._run_search_query(
370 query, data={'query': query, 'page': {'size': 100, 'current': page_number}},
a70635b8 371 note=f'Downloading page {page_number}{format_field(total_pages, None, " of ~%s")}')
c9b2b368 372 total_pages = traverse_obj(search_results, ('meta', 'page', 'total_pages'), expected_type=int_or_none)
373
374 for result in search_results.get('results') or []:
375 video_id_key, video_type = self._TYPES.get(traverse_obj(result, ('content_type', 'raw')), (None, None))
376 video_id = traverse_obj(result, video_id_key, expected_type=int_or_none)
377 if video_id and video_type:
378 yield self.url_result(url=f'https://rokfin.com/{video_type}/{video_id}')
379 if not search_results.get('results'):
380 return
381
382 def _run_search_query(self, video_id, data, **kwargs):
383 data = json.dumps(data).encode()
384 for attempt in range(2):
385 search_results = self._download_json(
386 self._db_url, video_id, data=data, fatal=(attempt == 1),
387 headers={'authorization': self._db_access_key}, **kwargs)
388 if search_results:
389 return search_results
390 self.write_debug('Updating access credentials')
391 self._get_db_access_credentials(video_id)
392
393 def _get_db_access_credentials(self, video_id=None):
394 auth_data = {'SEARCH_KEY': None, 'ENDPOINT_BASE': None}
395 notfound_err_page = self._download_webpage(
396 'https://rokfin.com/discover', video_id, expected_status=404, note='Downloading home page')
397 for js_file_path in re.findall(r'<script\b[^>]*\ssrc\s*=\s*"(/static/js/[^">]+)"', notfound_err_page):
398 js_content = self._download_webpage(
399 f'https://rokfin.com{js_file_path}', video_id, note='Downloading JavaScript file', fatal=False)
400 auth_data.update(re.findall(
401 rf'REACT_APP_({"|".join(auth_data.keys())})\s*:\s*"([^"]+)"', js_content or ''))
402 if not all(auth_data.values()):
403 continue
404
405 self._db_url = url_or_none(f'{auth_data["ENDPOINT_BASE"]}/api/as/v1/engines/rokfin-search/search.json')
406 self._db_access_key = f'Bearer {auth_data["SEARCH_KEY"]}'
9809740b 407 self.cache.store(self.ie_key(), 'auth', (self._db_url, self._db_access_key))
c9b2b368 408 return
409 raise ExtractorError('Unable to extract access credentials')
fork of https://github.com/yt-dlp/yt-dlp/