From 67db51cdfe08bcaf9cf2500851ba2d0f04163757 Mon Sep 17 00:00:00 2001 From: Nick Lai Date: Fri, 10 Apr 2020 21:34:20 +0800 Subject: [PATCH] [MicrosoftStream] Add extractor for channels and refactor single video extractor --- youtube_dl/extractor/extractors.py | 5 +- youtube_dl/extractor/microsoftstream.py | 247 +++++++++++++++++------- 2 files changed, 176 insertions(+), 76 deletions(-) diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index 847a58f16..6086437fa 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -619,7 +619,10 @@ from .metacritic import MetacriticIE from .mgoon import MgoonIE from .mgtv import MGTVIE from .miaopai import MiaoPaiIE -from .microsoftstream import MicrosoftStreamIE +from .microsoftstream import ( + MicrosoftStreamIE, + MicrosoftStreamChannelIE +) from .microsoftvirtualacademy import ( MicrosoftVirtualAcademyIE, MicrosoftVirtualAcademyCourseIE, diff --git a/youtube_dl/extractor/microsoftstream.py b/youtube_dl/extractor/microsoftstream.py index 57841c5ef..99515479a 100644 --- a/youtube_dl/extractor/microsoftstream.py +++ b/youtube_dl/extractor/microsoftstream.py @@ -1,11 +1,13 @@ # coding: utf-8 from __future__ import unicode_literals from .common import InfoExtractor -from ..utils import ExtractorError +from ..utils import (ExtractorError, update_url_query, merge_dicts) class MicrosoftStreamBaseIE(InfoExtractor): _LOGIN_URL = 'https://web.microsoftstream.com/?noSignUpCheck=1' # expect redirection + _ACCESS_TOKEN = None # A JWT token + _API_GATEWAY = None def is_logged_in(self, webpage): """ @@ -37,43 +39,14 @@ class MicrosoftStreamBaseIE(InfoExtractor): self._API_GATEWAY = self._html_search_regex(r"\"ApiGatewayUri\":\"(?P.+?)\"", webpage, 'APIGateway') return self._API_GATEWAY - -class MicrosoftStreamIE(MicrosoftStreamBaseIE): """ - Extractor for single Microsoft Stream video - """ - IE_NAME = 'microsoftstream' - _VALID_URL = r'https?://(?:(?:web|www)\.)?microsoftstream\.com/video/(?P[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12})' # https://regex101.com/r/K1mlgK/1/ - _NETRC_MACHINE = 'microsoftstream' - _ACCESS_TOKEN = None # A JWT token - _API_GATEWAY = None - _TEXTTRACKS_RESPONSE = None - _VIDEO_ID = None - - _TEST = [{ - 'url': 'https://web.microsoftstream.com/video/c883c6a5-9895-4900-9a35-62f4b5d506c9', - 'info_dict': { - 'id': 'c883c6a5-9895-4900-9a35-62f4b5d506c9', - 'ext': 'mp4', - 'title': 'Webinar for Researchers: Use of GitLab', - 'thumbnail': r're:^https?://.*$' - }, - 'skip': 'Requires Microsoft 365 account credentials', - }, { - 'url': 'https://web.microsoftstream.com/video/c883c6a5-9895-4900-9a35-62f4b5d506c9', - 'only_matching': True, - }, { - 'url': 'https://www.microsoftstream.com/video/1541f3f9-7fed-4901-ae70-0f7cb775679f', - 'only_matching': True, - }] - - """ - Getters - - The following getters include helpful message to prompt developers for potential errors. + Common getters """ @property def api_gateway(self): + """ + Return the start of an API endoint, like "https://aaea-1.api.microsoftstream.com/api/" + """ if self._API_GATEWAY is None: raise ExtractorError('API gateway is None. Did you forget to call "_extract_api_gateway"?') return self._API_GATEWAY @@ -82,34 +55,15 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): def access_token(self): if self._ACCESS_TOKEN is None: raise ExtractorError('Access token is None. Did you forget to call "_extract_access_token"?') - return self._ACCESS_TOKEN - @property - def video_id(self): - if self._VIDEO_ID is None: - raise('Variable "_VIDEO_ID" is not defined. Did you make the main extraction call?') - return self._VIDEO_ID - @property def headers(self): return {'Authorization': 'Bearer %s' % self.access_token} - @property - def texttrack_info_endpoint(self): - return "%s/videos/%s/texttracks?api-version=1.3-private" % (self.api_gateway, self.video_id) - - @property - def media_info_endpoint(self): - return "%s/videos/%s?$expand=creator,tokens,status,liveEvent,extensions&api-version=1.3-private" % (self.api_gateway, self.video_id) - - def _request_texttracks(self): - """ - Make an additional request to Microsoft Stream for the subtitle and auto-caption - """ - # Map default variable - self._TEXTTRACKS_RESPONSE = self._download_json(self.texttrack_info_endpoint, self.video_id, headers=self.headers).get('value') - return self._TEXTTRACKS_RESPONSE + """ + Utils function + """ def _determine_protocol(self, mime): """ @@ -122,6 +76,12 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): else: return None + """ + Remapper + + Remap data into correct field + """ + def _remap_thumbnails(self, thumbnail_dict_list): output = [] preference_index = ['extraSmall', 'small', 'medium', 'large'] @@ -133,7 +93,7 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): }) return output - def _remap_playback(self, master_playlist_urls): + def _remap_playback(self, master_playlist_urls, video_id=None): """ A parser for the HLS and MPD playlists from the API endpoint. """ @@ -143,11 +103,11 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): protocol = self._determine_protocol(master_playlist_url['mimeType']) # Handle HLS Master playlist if protocol == 'm3u8': - varient_playlists = self._extract_m3u8_formats(master_playlist_url['playbackUrl'], video_id=self.video_id, headers=self.headers) + varient_playlists = self._extract_m3u8_formats(master_playlist_url['playbackUrl'], video_id=video_id, headers=self.headers) # For MPEG-DASH Master playlists elif protocol == 'http_dash_segments': - varient_playlists = self._extract_mpd_formats(master_playlist_url['playbackUrl'], video_id=self.video_id, headers=self.headers) + varient_playlists = self._extract_mpd_formats(master_playlist_url['playbackUrl'], video_id=video_id, headers=self.headers) # For other Master playlists (like Microsoft Smooth Streaming) else: @@ -160,7 +120,7 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): output.append(varient_playlist) return output - def _extract_subtitle(self, tracks, is_auto_generated): + def _remap_subtitle(self, tracks, video_id, is_auto_generated): """ An internal method for filtering and remapping text tracks """ @@ -177,18 +137,83 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): subtitle_subset[track_language] = [] # Scaffold an empty list for the object to insert into # Since the subtitle is token protected, a get request will fire here. - data = self._download_webpage(url_or_request=track.get('url'), video_id=self.video_id, headers=self.headers) + data = self._download_webpage(url_or_request=track.get('url'), video_id=video_id, headers=self.headers) subtitle_subset[track_language].append({'data': data, "ext": "vtt"}) return subtitle_subset + def _remap_video(self, video): + return { + 'id': video['id'], + 'title': video['name'], + 'description': video.get('description'), + 'uploader': video.get('creator').get('name'), + 'thumbnails': self._remap_thumbnails(video.get('posterImage') or []), + 'formats': self._remap_playback(video['playbackUrls'], video_id=video['id']), + 'is_live': False + } + """ + Formatter + """ + + def _format_texttrack_endpoint(self, video_id): + return "%s/videos/%s/texttracks?api-version=1.4-private" % (self.api_gateway, video_id) + + def _request_texttracks(self, video_id): + """ + Make an additional request to Microsoft Stream for the subtitle and auto-caption + """ + self._TEXTTRACKS_RESPONSE = self._download_json(self._format_texttrack_endpoint(video_id), video_id, headers=self.headers).get('value') + return self._TEXTTRACKS_RESPONSE + + +class MicrosoftStreamIE(MicrosoftStreamBaseIE): + """ + Extractor for single Microsoft Stream video + """ + IE_NAME = 'microsoftstream' + _VALID_URL = r'https?://(?:(?:web|www)\.)?microsoftstream\.com/video/(?P[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12})' # https://regex101.com/r/K1mlgK/1/ + _NETRC_MACHINE = 'microsoftstream' + _TEXTTRACKS_RESPONSE = None + _VIDEO_ID = None + + _TEST = [{ + 'url': 'https://web.microsoftstream.com/video/c883c6a5-9895-4900-9a35-62f4b5d506c9', + 'info_dict': { + 'id': 'c883c6a5-9895-4900-9a35-62f4b5d506c9', + 'ext': 'mp4', + 'title': 'Webinar for Researchers: Use of GitLab', + 'thumbnail': r're:^https?://.*$' + }, + 'skip': 'Requires Microsoft 365 account credentials', + }, { + 'url': 'https://www.microsoftstream.com/video/1541f3f9-7fed-4901-ae70-0f7cb775679f', + 'only_matching': True, + }] + + """ + Getters + + The following getters include helpful message to prompt developers for potential errors. + """ + + @property + def video_id(self): + if self._VIDEO_ID is None: + raise('Variable "_VIDEO_ID" is not defined. Did you make the main extraction call?') + return self._VIDEO_ID + + @property + def media_info_endpoint(self): + return "%s/videos/%s?$expand=creator,tokens,status,liveEvent,extensions&api-version=1.4-private" % (self.api_gateway, self.video_id) + def _get_subtitles(self, tracks=None): # Fulfill abstract method tracks = self._TEXTTRACKS_RESPONSE if tracks is None else tracks - return self._extract_subtitle(tracks, False) + return self._remap_subtitle(tracks, is_auto_generated=False, video_id=self.video_id) def _get_automatic_captions(self, tracks=None): # Fulfill abstract method tracks = self._TEXTTRACKS_RESPONSE if tracks is None else tracks - return self._extract_subtitle(tracks, True) + return self._remap_subtitle(tracks, is_auto_generated=True, video_id=self.video_id) def _real_extract(self, url): self._VIDEO_ID = self._match_id(url) @@ -202,19 +227,91 @@ class MicrosoftStreamIE(MicrosoftStreamBaseIE): self._extract_api_gateway(webpage) # "GET" api for video information - apiResponse = self._download_json(self.media_info_endpoint, self.video_id, headers=self.headers) + video = self._download_json(self.media_info_endpoint, self.video_id, headers=self.headers) + texttracks = self._request_texttracks(self.video_id) - texttracks = self._request_texttracks() - - return { - 'id': self.video_id, - 'title': apiResponse['name'], - 'description': apiResponse.get('description'), - 'uploader': apiResponse.get('creator').get('name'), - 'thumbnails': self._remap_thumbnails(apiResponse.get('posterImage')), - 'formats': self._remap_playback(apiResponse['playbackUrls']), + return merge_dicts(self._remap_video(video), { 'subtitles': self._get_subtitles(texttracks), 'automatic_captions': self._get_automatic_captions(texttracks), - 'is_live': False - # 'duration': apiResponse['media']['duration'], + }) + + +class MicrosoftStreamChannelIE(MicrosoftStreamBaseIE): + """ + Extractor for single channel of Microsoft Stream video + """ + IE_NAME = 'microsoftstream:channel' + _VALID_URL = r'https?://(?:(?:web|www)\.)?microsoftstream\.com/channel/(?P[0-9a-f]{8}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{4}-[0-9a-f]{12})' # https://regex101.com/r/K1mlgK/1/ + _NETRC_MACHINE = 'microsoftstream' + _CHANNEL_ID = None + + _TEST = [{ + 'url': 'https://web.microsoftstream.com/channel/c883c6a5-9895-4900-9a35-62f4b5d506c9', + 'only_matching': True, + }, { + 'url': 'https://www.microsoftstream.com/channel/0ceffe58-07b1-4098-8ed9-a15f4e8231f7', + 'only_matching': True, + }] + + def _format_channel_video_endpoint(self, skip=0, top=100): + parameters = { + '$top': top, + '$skip': skip, + '$expand': 'creator,events', + '$orderby': 'created asc', + 'api-version': '1.4-private' } + + return update_url_query('%schannels/%s/videos' % (self.api_gateway, self._CHANNEL_ID), parameters) + + def _iterate_video(self, video): + subtitle_dict = {} + + if (self._downloader.params.get('writesubtitles', False) + or self._downloader.params.get('writeautomaticsub', False) + or self._downloader.params.get('listsubtitles')): + texttracks = self._request_texttracks(video['id']) + + if (self._downloader.params.get('writesubtitles', False) + or self._downloader.params.get('listsubtitles')): + subtitle_dict['subtitles'] = self._remap_subtitle(texttracks, video_id=video['id'], is_auto_generated=False) + + if (self._downloader.params.get('writeautomaticsub', False) + or self._downloader.params.get('listsubtitles')): + subtitle_dict['automatic_captions'] = self._remap_subtitle(texttracks, video_id=video['id'], is_auto_generated=False) + + return merge_dicts(self._remap_video(video), subtitle_dict) + + def _fetch_video(self): + found_all_video = False + current_skip = 0 + top = 100 + all_video = [] + + while found_all_video is False: + video_info_endpoint = self._format_channel_video_endpoint(current_skip, top) + channel_video_subset = self._download_json(video_info_endpoint, self._CHANNEL_ID, headers=self.headers)['value'] + + for video in channel_video_subset: + all_video.append(self._iterate_video(video)) # Remap the video + + found_all_video = bool(len(channel_video_subset) < top) # Break out from the iteration if all content is downloaded. + current_skip += top # Adjust starting position + + return all_video + + def _real_extract(self, url): + self._CHANNEL_ID = self._match_id(url) + + webpage = self._download_webpage(url, self._CHANNEL_ID) + if not self.is_logged_in(webpage): + return self.raise_login_required() + + # Extract access token from webpage + self._extract_access_token(webpage) + self._extract_api_gateway(webpage) + + entries = self._fetch_video() + + return {'_type': 'playlist', + 'entries': entries}