[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

diff --git a/test/test_utils.py b/test/test_utils.py
index d84c3d3eefe15b637e21bd5bff54508b8f7b0735..810ed3de4cb43a8efa5195d759c67b98812994c7 100644 (file)
--- 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 8ea7d2ed8ebda0979cda336eb4bb299e66acad6b..ef2b25c9308b6107a500ea79737550db3ea6fd7b 100644 (file)
--- 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 (file)
index 0000000..4d5a9df
--- /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<id>[\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 '<title>Microsoft Stream</title>' 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 319f6979ba2b0072a3e81ab610c6c275c50aeb47..e05677d08eb0f219c97b1b0e32df946b6cb09415 100644 (file)
--- 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<w>\d+)\s*[xX×]\s*(?P<h>\d+)\b', s)
+    mobj = re.search(r'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\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'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s)
     if mobj:
         return {'height': int(mobj.group(1))}