diff options
Diffstat (limited to 'hypervideo_dl/extractor/rokfin.py')
| -rw-r--r-- | hypervideo_dl/extractor/rokfin.py | 173 |
1 files changed, 163 insertions, 10 deletions
diff --git a/hypervideo_dl/extractor/rokfin.py b/hypervideo_dl/extractor/rokfin.py index 0fd65db..ade3cd0 100644 --- a/hypervideo_dl/extractor/rokfin.py +++ b/hypervideo_dl/extractor/rokfin.py @@ -1,26 +1,33 @@ -# coding: utf-8 import itertools +import json +import re +import urllib.parse from datetime import datetime -from .common import InfoExtractor +from .common import InfoExtractor, SearchInfoExtractor from ..utils import ( - determine_ext, ExtractorError, + determine_ext, float_or_none, format_field, int_or_none, str_or_none, traverse_obj, + try_get, + unescapeHTML, unified_timestamp, url_or_none, + urlencode_postdata, ) - _API_BASE_URL = 'https://prod-api-v2.production.rokfin.com/api/v2/public/' class RokfinIE(InfoExtractor): _VALID_URL = r'https?://(?:www\.)?rokfin\.com/(?P<id>(?P<type>post|stream)/\d+)' + _NETRC_MACHINE = 'rokfin' + _AUTH_BASE = 'https://secure.rokfin.com/auth/realms/rokfin-web/protocol/openid-connect' + _access_mgmt_tokens = {} # OAuth 2.0: RFC 6749, Sec. 1.4-5 _TESTS = [{ 'url': 'https://www.rokfin.com/post/57548/Mitt-Romneys-Crazy-Solution-To-Climate-Change', 'info_dict': { @@ -84,8 +91,7 @@ class RokfinIE(InfoExtractor): def _real_extract(self, url): video_id, video_type = self._match_valid_url(url).group('id', 'type') - - metadata = self._download_json(f'{_API_BASE_URL}{video_id}', video_id) + metadata = self._download_json_using_access_token(f'{_API_BASE_URL}{video_id}', video_id) scheduled = unified_timestamp(metadata.get('scheduledAt')) live_status = ('was_live' if metadata.get('stoppedAt') @@ -104,9 +110,8 @@ class RokfinIE(InfoExtractor): self.raise_login_required('This video is only available to premium users', True, method='cookies') elif scheduled: self.raise_no_formats( - f'Stream is offline; sheduled for {datetime.fromtimestamp(scheduled).strftime("%Y-%m-%d %H:%M:%S")}', + f'Stream is offline; scheduled for {datetime.fromtimestamp(scheduled).strftime("%Y-%m-%d %H:%M:%S")}', video_id=video_id, expected=True) - self._sort_formats(formats) uploader = traverse_obj(metadata, ('createdBy', 'username'), ('creator', 'username')) timestamp = (scheduled or float_or_none(metadata.get('postedAtMilli'), 1000) @@ -140,7 +145,7 @@ class RokfinIE(InfoExtractor): for page_n in itertools.count(): raw_comments = self._download_json( f'{_API_BASE_URL}comment?postId={video_id[5:]}&page={page_n}&size=50', - video_id, note=f'Downloading viewer comments page {page_n + 1}{format_field(pages_total, template=" of %s")}', + video_id, note=f'Downloading viewer comments page {page_n + 1}{format_field(pages_total, None, " of %s")}', fatal=False) or {} for comment in raw_comments.get('content') or []: @@ -160,6 +165,79 @@ class RokfinIE(InfoExtractor): if not raw_comments.get('content') or is_last or (page_n > pages_total if pages_total else is_last is not False): return + def _perform_login(self, username, password): + # https://openid.net/specs/openid-connect-core-1_0.html#CodeFlowAuth (Sec. 3.1) + login_page = self._download_webpage( + f'{self._AUTH_BASE}/auth?client_id=web&redirect_uri=https%3A%2F%2Frokfin.com%2Ffeed&response_mode=fragment&response_type=code&scope=openid', + None, note='loading login page', errnote='error loading login page') + authentication_point_url = unescapeHTML(self._search_regex( + r'<form\s+[^>]+action\s*=\s*"(https://secure\.rokfin\.com/auth/realms/rokfin-web/login-actions/authenticate\?[^"]+)"', + login_page, name='Authentication URL')) + + resp_body = self._download_webpage( + authentication_point_url, None, note='logging in', fatal=False, expected_status=404, + data=urlencode_postdata({'username': username, 'password': password, 'rememberMe': 'off', 'credentialId': ''})) + if not self._authentication_active(): + if re.search(r'(?i)(invalid\s+username\s+or\s+password)', resp_body or ''): + raise ExtractorError('invalid username/password', expected=True) + raise ExtractorError('Login failed') + + urlh = self._request_webpage( + f'{self._AUTH_BASE}/auth', None, + note='granting user authorization', errnote='user authorization rejected by Rokfin', + query={ + 'client_id': 'web', + 'prompt': 'none', + 'redirect_uri': 'https://rokfin.com/silent-check-sso.html', + 'response_mode': 'fragment', + 'response_type': 'code', + 'scope': 'openid', + }) + self._access_mgmt_tokens = self._download_json( + f'{self._AUTH_BASE}/token', None, + note='getting access credentials', errnote='error getting access credentials', + data=urlencode_postdata({ + 'code': urllib.parse.parse_qs(urllib.parse.urldefrag(urlh.geturl()).fragment).get('code')[0], + 'client_id': 'web', + 'grant_type': 'authorization_code', + 'redirect_uri': 'https://rokfin.com/silent-check-sso.html' + })) + + def _authentication_active(self): + return not ( + {'KEYCLOAK_IDENTITY', 'KEYCLOAK_IDENTITY_LEGACY', 'KEYCLOAK_SESSION', 'KEYCLOAK_SESSION_LEGACY'} + - set(self._get_cookies(self._AUTH_BASE))) + + def _get_auth_token(self): + return try_get(self._access_mgmt_tokens, lambda x: ' '.join([x['token_type'], x['access_token']])) + + def _download_json_using_access_token(self, url_or_request, video_id, headers={}, query={}): + assert 'authorization' not in headers + headers = headers.copy() + auth_token = self._get_auth_token() + refresh_token = self._access_mgmt_tokens.get('refresh_token') + if auth_token: + headers['authorization'] = auth_token + + json_string, urlh = self._download_webpage_handle( + url_or_request, video_id, headers=headers, query=query, expected_status=401) + if not auth_token or urlh.code != 401 or refresh_token is None: + return self._parse_json(json_string, video_id) + + self._access_mgmt_tokens = self._download_json( + f'{self._AUTH_BASE}/token', video_id, + note='User authorization expired or canceled by Rokfin. Re-authorizing ...', errnote='Failed to re-authorize', + data=urlencode_postdata({ + 'grant_type': 'refresh_token', + 'refresh_token': refresh_token, + 'client_id': 'web' + })) + headers['authorization'] = self._get_auth_token() + if headers['authorization'] is None: + raise ExtractorError('User authorization lost', expected=True) + + return self._download_json(url_or_request, video_id, headers=headers, query=query) + class RokfinPlaylistBaseIE(InfoExtractor): _TYPES = { @@ -183,6 +261,7 @@ class RokfinPlaylistBaseIE(InfoExtractor): class RokfinStackIE(RokfinPlaylistBaseIE): IE_NAME = 'rokfin:stack' + IE_DESC = 'Rokfin Stacks' _VALID_URL = r'https?://(?:www\.)?rokfin\.com/stack/(?P<id>[^/]+)' _TESTS = [{ 'url': 'https://www.rokfin.com/stack/271/Tulsi-Gabbard-Portsmouth-Townhall-FULL--Feb-9-2020', @@ -200,6 +279,7 @@ class RokfinStackIE(RokfinPlaylistBaseIE): class RokfinChannelIE(RokfinPlaylistBaseIE): IE_NAME = 'rokfin:channel' + IE_DESC = 'Rokfin Channels' _VALID_URL = r'https?://(?:www\.)?rokfin\.com/(?!((feed/?)|(discover/?)|(channels/?))$)(?P<id>[^/]+)/?$' _TESTS = [{ 'url': 'https://rokfin.com/TheConvoCouch', @@ -237,7 +317,7 @@ class RokfinChannelIE(RokfinPlaylistBaseIE): data_url = f'{_API_BASE_URL}post/search/{tab}?page={page_n}&size=50&creator={channel_id}' metadata = self._download_json( data_url, channel_name, - note=f'Downloading video metadata page {page_n + 1}{format_field(pages_total, template=" of %s")}') + note=f'Downloading video metadata page {page_n + 1}{format_field(pages_total, None, " of %s")}') yield from self._get_video_data(metadata) pages_total = int_or_none(metadata.get('totalPages')) or None @@ -254,3 +334,76 @@ class RokfinChannelIE(RokfinPlaylistBaseIE): return self.playlist_result( self._entries(channel_id, channel_name, self._TABS[tab]), f'{channel_id}-{tab}', f'{channel_name} - {tab.title()}', str_or_none(channel_info.get('description'))) + + +class RokfinSearchIE(SearchInfoExtractor): + IE_NAME = 'rokfin:search' + IE_DESC = 'Rokfin Search' + _SEARCH_KEY = 'rkfnsearch' + _TYPES = { + 'video': (('id', 'raw'), 'post'), + 'audio': (('id', 'raw'), 'post'), + 'stream': (('content_id', 'raw'), 'stream'), + 'dead_stream': (('content_id', 'raw'), 'stream'), + 'stack': (('content_id', 'raw'), 'stack'), + } + _TESTS = [{ + 'url': 'rkfnsearch5:"zelenko"', + 'playlist_count': 5, + 'info_dict': { + 'id': '"zelenko"', + 'title': '"zelenko"', + } + }] + _db_url = None + _db_access_key = None + + def _real_initialize(self): + self._db_url, self._db_access_key = self.cache.load(self.ie_key(), 'auth', default=(None, None)) + if not self._db_url: + self._get_db_access_credentials() + + def _search_results(self, query): + total_pages = None + for page_number in itertools.count(1): + search_results = self._run_search_query( + query, data={'query': query, 'page': {'size': 100, 'current': page_number}}, + note=f'Downloading page {page_number}{format_field(total_pages, None, " of ~%s")}') + total_pages = traverse_obj(search_results, ('meta', 'page', 'total_pages'), expected_type=int_or_none) + + for result in search_results.get('results') or []: + video_id_key, video_type = self._TYPES.get(traverse_obj(result, ('content_type', 'raw')), (None, None)) + video_id = traverse_obj(result, video_id_key, expected_type=int_or_none) + if video_id and video_type: + yield self.url_result(url=f'https://rokfin.com/{video_type}/{video_id}') + if not search_results.get('results'): + return + + def _run_search_query(self, video_id, data, **kwargs): + data = json.dumps(data).encode() + for attempt in range(2): + search_results = self._download_json( + self._db_url, video_id, data=data, fatal=(attempt == 1), + headers={'authorization': self._db_access_key}, **kwargs) + if search_results: + return search_results + self.write_debug('Updating access credentials') + self._get_db_access_credentials(video_id) + + def _get_db_access_credentials(self, video_id=None): + auth_data = {'SEARCH_KEY': None, 'ENDPOINT_BASE': None} + notfound_err_page = self._download_webpage( + 'https://rokfin.com/discover', video_id, expected_status=404, note='Downloading home page') + for js_file_path in re.findall(r'<script\b[^>]*\ssrc\s*=\s*"(/static/js/[^">]+)"', notfound_err_page): + js_content = self._download_webpage( + f'https://rokfin.com{js_file_path}', video_id, note='Downloading JavaScript file', fatal=False) + auth_data.update(re.findall( + rf'REACT_APP_({"|".join(auth_data.keys())})\s*:\s*"([^"]+)"', js_content or '')) + if not all(auth_data.values()): + continue + + self._db_url = url_or_none(f'{auth_data["ENDPOINT_BASE"]}/api/as/v1/engines/rokfin-search/search.json') + self._db_access_key = f'Bearer {auth_data["SEARCH_KEY"]}' + self.cache.store(self.ie_key(), 'auth', (self._db_url, self._db_access_key)) + return + raise ExtractorError('Unable to extract access credentials') |