From 17ec8bcfa90b80913072fdcb0cafc09c1ad79849 Mon Sep 17 00:00:00 2001 From: Damiano Amatruda Date: Fri, 22 Oct 2021 02:04:00 +0200 Subject: [PATCH] [microsoftstream] Add extractor (#1201) Based on: https://github.com/ytdl-org/youtube-dl/pull/24649 Fixes: https://github.com/ytdl-org/youtube-dl/issues/24440 Authored by: damianoamatruda, nixklai --- test/test_utils.py | 7 +- yt_dlp/extractor/extractors.py | 1 + yt_dlp/extractor/microsoftstream.py | 125 ++++++++++++++++++++++++++++ yt_dlp/utils.py | 4 +- 4 files changed, 133 insertions(+), 4 deletions(-) create mode 100644 yt_dlp/extractor/microsoftstream.py diff --git a/test/test_utils.py b/test/test_utils.py index d84c3d3eef..810ed3de4c 100644 --- a/test/test_utils.py +++ b/test/test_utils.py @@ -1163,12 +1163,15 @@ def test_parse_count(self): def test_parse_resolution(self): self.assertEqual(parse_resolution(None), {}) self.assertEqual(parse_resolution(''), {}) - self.assertEqual(parse_resolution('1920x1080'), {'width': 1920, 'height': 1080}) - self.assertEqual(parse_resolution('1920×1080'), {'width': 1920, 'height': 1080}) + self.assertEqual(parse_resolution(' 1920x1080'), {'width': 1920, 'height': 1080}) + self.assertEqual(parse_resolution('1920×1080 '), {'width': 1920, 'height': 1080}) self.assertEqual(parse_resolution('1920 x 1080'), {'width': 1920, 'height': 1080}) self.assertEqual(parse_resolution('720p'), {'height': 720}) self.assertEqual(parse_resolution('4k'), {'height': 2160}) self.assertEqual(parse_resolution('8K'), {'height': 4320}) + self.assertEqual(parse_resolution('pre_1920x1080_post'), {'width': 1920, 'height': 1080}) + self.assertEqual(parse_resolution('ep1x2'), {}) + self.assertEqual(parse_resolution('1920, 1080'), {'width': 1920, 'height': 1080}) def test_parse_bitrate(self): self.assertEqual(parse_bitrate(None), None) diff --git a/yt_dlp/extractor/extractors.py b/yt_dlp/extractor/extractors.py index 8ea7d2ed8e..ef2b25c930 100644 --- a/yt_dlp/extractor/extractors.py +++ b/yt_dlp/extractor/extractors.py @@ -760,6 +760,7 @@ from .mgoon import MgoonIE from .mgtv import MGTVIE from .miaopai import MiaoPaiIE +from .microsoftstream import MicrosoftStreamIE from .microsoftvirtualacademy import ( MicrosoftVirtualAcademyIE, MicrosoftVirtualAcademyCourseIE, diff --git a/yt_dlp/extractor/microsoftstream.py b/yt_dlp/extractor/microsoftstream.py new file mode 100644 index 0000000000..4d5a9df1ff --- /dev/null +++ b/yt_dlp/extractor/microsoftstream.py @@ -0,0 +1,125 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from base64 import b64decode + +from .common import InfoExtractor +from ..utils import ( + merge_dicts, + parse_iso8601, + parse_duration, + parse_resolution, + try_get, + url_basename, +) + + +class MicrosoftStreamIE(InfoExtractor): + IE_NAME = 'microsoftstream' + IE_DESC = 'Microsoft Stream' + _VALID_URL = r'https?://(?:web|www|msit)\.microsoftstream\.com/video/(?P[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})' + + _TESTS = [{ + 'url': 'https://web.microsoftstream.com/video/6e51d928-4f46-4f1c-b141-369925e37b62?list=user&userId=f5491e02-e8fe-4e34-b67c-ec2e79a6ecc0', + 'only_matching': True, + }, { + 'url': 'https://msit.microsoftstream.com/video/b60f5987-aabd-4e1c-a42f-c559d138f2ca', + 'only_matching': True, + }] + + def _get_all_subtitles(self, api_url, video_id, headers): + subtitles = {} + automatic_captions = {} + text_tracks = self._download_json( + f'{api_url}/videos/{video_id}/texttracks', video_id, + note='Downloading subtitles JSON', fatal=False, headers=headers, + query={'api-version': '1.4-private'}).get('value') or [] + for track in text_tracks: + if not track.get('language') or not track.get('url'): + continue + sub_dict = automatic_captions if track.get('autoGenerated') else subtitles + sub_dict.setdefault(track['language'], []).append({ + 'ext': 'vtt', + 'url': track.get('url') + }) + return { + 'subtitles': subtitles, + 'automatic_captions': automatic_captions + } + + def extract_all_subtitles(self, *args, **kwargs): + if (self.get_param('writesubtitles', False) + or self.get_param('writeautomaticsub', False) + or self.get_param('listsubtitles')): + return self._get_all_subtitles(*args, **kwargs) + return {} + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + if 'Microsoft Stream' not in webpage: + self.raise_login_required(method='cookies') + + access_token = self._html_search_regex(r'"AccessToken":"(.+?)"', webpage, 'access token') + api_url = self._html_search_regex(r'"ApiGatewayUri":"(.+?)"', webpage, 'api url') + + headers = {'Authorization': f'Bearer {access_token}'} + + video_data = self._download_json( + f'{api_url}/videos/{video_id}', video_id, + headers=headers, query={ + '$expand': 'creator,tokens,status,liveEvent,extensions', + 'api-version': '1.4-private' + }) + video_id = video_data.get('id') or video_id + language = video_data.get('language') + + thumbnails = [] + for thumbnail_id in ('extraSmall', 'small', 'medium', 'large'): + thumbnail_url = try_get(video_data, lambda x: x['posterImage'][thumbnail_id]['url'], str) + if not thumbnail_url: + continue + thumb = { + 'id': thumbnail_id, + 'url': thumbnail_url, + } + thumb_name = url_basename(thumbnail_url) + thumb_name = str(b64decode(thumb_name + '=' * (-len(thumb_name) % 4))) + thumb.update(parse_resolution(thumb_name)) + thumbnails.append(thumb) + + formats = [] + for playlist in video_data['playbackUrls']: + if playlist['mimeType'] == 'application/vnd.apple.mpegurl': + formats.extend(self._extract_m3u8_formats( + playlist['playbackUrl'], video_id, + ext='mp4', entry_protocol='m3u8_native', m3u8_id='hls', + fatal=False, headers=headers)) + elif playlist['mimeType'] == 'application/dash+xml': + formats.extend(self._extract_mpd_formats( + playlist['playbackUrl'], video_id, mpd_id='dash', + fatal=False, headers=headers)) + elif playlist['mimeType'] == 'application/vnd.ms-sstr+xml': + formats.extend(self._extract_ism_formats( + playlist['playbackUrl'], video_id, ism_id='mss', + fatal=False, headers=headers)) + formats = [merge_dicts(f, {'language': language}) for f in formats] + self._sort_formats(formats) + + return { + 'id': video_id, + 'title': video_data['name'], + 'description': video_data.get('description'), + 'uploader': try_get(video_data, lambda x: x['creator']['name'], str), + 'uploader_id': try_get(video_data, (lambda x: x['creator']['mail'], + lambda x: x['creator']['id']), str), + 'thumbnails': thumbnails, + **self.extract_all_subtitles(api_url, video_id, headers), + 'timestamp': parse_iso8601(video_data.get('created')), + 'duration': parse_duration(try_get(video_data, lambda x: x['media']['duration'])), + 'webpage_url': f'https://web.microsoftstream.com/video/{video_id}', + 'view_count': try_get(video_data, lambda x: x['metrics']['views'], int), + 'like_count': try_get(video_data, lambda x: x['metrics']['likes'], int), + 'comment_count': try_get(video_data, lambda x: x['metrics']['comments'], int), + 'formats': formats, + } diff --git a/yt_dlp/utils.py b/yt_dlp/utils.py index 319f6979ba..e05677d08e 100644 --- a/yt_dlp/utils.py +++ b/yt_dlp/utils.py @@ -3714,14 +3714,14 @@ def parse_resolution(s): if s is None: return {} - mobj = re.search(r'\b(?P\d+)\s*[xX×]\s*(?P\d+)\b', s) + mobj = re.search(r'(?\d+)\s*[xX×,]\s*(?P\d+)(?![a-zA-Z0-9])', s) if mobj: return { 'width': int(mobj.group('w')), 'height': int(mobj.group('h')), } - mobj = re.search(r'\b(\d+)[pPiI]\b', s) + mobj = re.search(r'(?