[ie/twitch:stream] Fix extractor (#14988)

Closes #14987
Authored by: seproDev

yt_dlp/extractor/floatplane.py

@@ -6,15 +6,15 @@
     OnDemandPagedList,
     clean_html,
     determine_ext,
+    float_or_none,
     format_field,
     int_or_none,
     join_nonempty,
-    parse_codecs,
     parse_iso8601,
     url_or_none,
     urljoin,
 )
-from ..utils.traversal import traverse_obj
+from ..utils.traversal import require, traverse_obj
 
 
 class FloatplaneBaseIE(InfoExtractor):
@@ -50,37 +50,31 @@ def _real_extract(self, url):
             media_id = media['id']
             media_typ = media.get('type') or 'video'
 
-            metadata = self._download_json(
-                f'{self._BASE_URL}/api/v3/content/{media_typ}', media_id, query={'id': media_id},
-                note=f'Downloading {media_typ} metadata', impersonate=self._IMPERSONATE_TARGET)
-
             stream = self._download_json(
-                f'{self._BASE_URL}/api/v2/cdn/delivery', media_id, query={
-                    'type': 'vod' if media_typ == 'video' else 'aod',
-                    'guid': metadata['guid'],
-                }, note=f'Downloading {media_typ} stream data',
+                f'{self._BASE_URL}/api/v3/delivery/info', media_id,
+                query={'scenario': 'onDemand', 'entityId': media_id},
+                note=f'Downloading {media_typ} stream data',
                 impersonate=self._IMPERSONATE_TARGET)
 
-            path_template = traverse_obj(stream, ('resource', 'uri', {str}))
+            metadata = self._download_json(
+                f'{self._BASE_URL}/api/v3/content/{media_typ}', media_id,
+                f'Downloading {media_typ} metadata', query={'id': media_id},
+                fatal=False, impersonate=self._IMPERSONATE_TARGET)
 
-            def format_path(params):
-                path = path_template
-                for i, val in (params or {}).items():
-                    path = path.replace(f'{{qualityLevelParams.{i}}}', val)
-                return path
+            cdn_base_url = traverse_obj(stream, (
+                'groups', 0, 'origins', ..., 'url', {url_or_none}, any, {require('cdn base url')}))
 
             formats = []
-            for quality in traverse_obj(stream, ('resource', 'data', 'qualityLevels', ...)):
-                url = urljoin(stream['cdn'], format_path(traverse_obj(
-                    stream, ('resource', 'data', 'qualityLevelParams', quality['name'], {dict}))))
-                format_id = traverse_obj(quality, ('name', {str}))
+            for variant in traverse_obj(stream, ('groups', 0, 'variants', lambda _, v: v['url'])):
+                format_url = urljoin(cdn_base_url, variant['url'])
+                format_id = traverse_obj(variant, ('name', {str}))
                 hls_aes = {}
                 m3u8_data = None
 
                 # If we need impersonation for the API, then we need it for HLS keys too: extract in advance
                 if self._IMPERSONATE_TARGET is not None:
                     m3u8_data = self._download_webpage(
-                        url, media_id, fatal=False, impersonate=self._IMPERSONATE_TARGET, headers=self._HEADERS,
+                        format_url, media_id, fatal=False, impersonate=self._IMPERSONATE_TARGET, headers=self._HEADERS,
                         note=join_nonempty('Downloading', format_id, 'm3u8 information', delim=' '),
                         errnote=join_nonempty('Failed to download', format_id, 'm3u8 information', delim=' '))
                     if not m3u8_data:
@@ -98,14 +92,19 @@ def format_path(params):
                             hls_aes['key'] = urlh.read().hex()
 
                 formats.append({
-                    **traverse_obj(quality, {
+                    **traverse_obj(variant, {
                         'format_note': ('label', {str}),
-                        'width': ('width', {int}),
-                        'height': ('height', {int}),
+                        'width': ('meta', 'video', 'width', {int_or_none}),
+                        'height': ('meta', 'video', 'height', {int_or_none}),
+                        'vcodec': ('meta', 'video', 'codec', {str}),
+                        'acodec': ('meta', 'audio', 'codec', {str}),
+                        'vbr': ('meta', 'video', 'bitrate', 'average', {int_or_none(scale=1000)}),
+                        'abr': ('meta', 'audio', 'bitrate', 'average', {int_or_none(scale=1000)}),
+                        'audio_channels': ('meta', 'audio', 'channelCount', {int_or_none}),
+                        'fps': ('meta', 'video', 'fps', {float_or_none}),
                     }),
-                    **parse_codecs(quality.get('codecs')),
-                    'url': url,
-                    'ext': determine_ext(url.partition('/chunk.m3u8')[0], 'mp4'),
+                    'url': format_url,
+                    'ext': determine_ext(format_url.partition('/chunk.m3u8')[0], 'mp4'),
                     'format_id': format_id,
                     'hls_media_playlist_data': m3u8_data,
                     'hls_aes': hls_aes or None,

yt_dlp/extractor/tubetugraz.py

@@ -136,8 +136,10 @@ class TubeTuGrazIE(TubeTuGrazBaseIE):
     IE_DESC = 'tube.tugraz.at'
 
     _VALID_URL = r'''(?x)
-        https?://tube\.tugraz\.at/paella/ui/watch.html\?id=
-        (?P<id>[0-9a-fA-F]{8}-(?:[0-9a-fA-F]{4}-){3}[0-9a-fA-F]{12})
+        https?://tube\.tugraz\.at/(?:
+            paella/ui/watch\.html\?(?:[^#]*&)?id=|
+            portal/watch/
+        )(?P<id>[0-9a-fA-F]{8}-(?:[0-9a-fA-F]{4}-){3}[0-9a-fA-F]{12})
     '''
     _TESTS = [
         {
@@ -152,6 +154,7 @@ class TubeTuGrazIE(TubeTuGrazBaseIE):
                 'creator': 'Safran C',
                 'duration': 3295818,
                 'series_id': 'b1192fff-2aa7-4bf0-a5cf-7b15c3bd3b34',
+                'creators': ['Safran C'],
             },
         }, {
             'url': 'https://tube.tugraz.at/paella/ui/watch.html?id=2df6d787-e56a-428d-8ef4-d57f07eef238',
@@ -162,6 +165,10 @@ class TubeTuGrazIE(TubeTuGrazBaseIE):
                 'ext': 'mp4',
             },
             'expected_warnings': ['Extractor failed to obtain "title"'],
+        }, {
+            # Portal URL format
+            'url': 'https://tube.tugraz.at/portal/watch/ab28ec60-8cbe-4f1a-9b96-a95add56c612',
+            'only_matching': True,
         },
     ]
 

yt_dlp/extractor/twitch.py

@@ -46,7 +46,7 @@ class TwitchBaseIE(InfoExtractor):
         'ClipsCards__User': 'b73ad2bfaecfd30a9e6c28fada15bd97032c83ec77a0440766a56fe0bd632777',
         'ShareClipRenderStatus': 'e0a46b287d760c6890a39d1ccd736af5ec9479a267d02c710e9ac33326b651d2',
         'ChannelCollectionsContent': '447aec6a0cc1e8d0a8d7732d47eb0762c336a2294fdb009e9c9d854e49d484b9',
-        'StreamMetadata': 'a647c2a13599e5991e175155f798ca7f1ecddde73f7f341f39009c14dbf59962',
+        'StreamMetadata': 'b57f9b910f8cd1a4659d894fe7550ccc81ec9052c01e438b290fd66a040b9b93',
         'ComscoreStreamingQuery': 'e1edae8122517d013405f237ffcc124515dc6ded82480a88daef69c83b53ac01',
         'VideoPreviewOverlay': '3006e77e51b128d838fa4e835723ca4dc9a05c5efd4466c1085215c6e437e65c',
         'VideoMetadata': '49b5b8f268cdeb259d75b58dcb0c1a748e3b575003448a2333dc5cdafd49adad',
@@ -1050,7 +1050,10 @@ def _real_extract(self, url):
         gql = self._download_gql(
             channel_name, [{
                 'operationName': 'StreamMetadata',
-                'variables': {'channelLogin': channel_name},
+                'variables': {
+                    'channelLogin': channel_name,
+                    'includeIsDJ': True,
+                },
             }, {
                 'operationName': 'ComscoreStreamingQuery',
                 'variables': {