Merge ce604ae7dce406aaf8fde8f25ee40bd390255b58 into 3eb8d22ddb8982ca4fb56bb7a8d6517538bf14c6

* thx seproDev, yt-dlp/yt-dlp/pull/12694/commits/ae69e3c
* also add documenting comment in `InfoExtractor`

youtube_dl/downloader/external.py

@@ -393,6 +393,19 @@ class FFmpegFD(ExternalFD):
             # https://github.com/ytdl-org/youtube-dl/issues/11800#issuecomment-275037127
             # http://trac.ffmpeg.org/ticket/6125#comment:10
             args += ['-seekable', '1' if seekable else '0']
+        http_seekable = info_dict.get('_http_seekable')
+        if http_seekable is not None:
+            # setting -http_seekable prevents ffmpeg from guessing if the server
+            # supports seeking in other kinds of requests (by adding the same header
+            # as above: `Range: bytes=0-`)
+            args += ['-http_seekable', '1' if http_seekable else '0']
+        icy = info_dict.get('_icy')
+        if icy is not None:
+            # setting -icy 0 prevents ffmpeg from sending the header `Icy-Metadata: 1`,
+            # which can cause also problems
+            # https://github.com/ytdl-org/youtube-dl/pull/29688
+            # https://trac.ffmpeg.org/ticket/5460#comment:5
+            args += ['-icy', '1' if icy else '0']
 
         args += self._configuration_args()
 

youtube_dl/extractor/bokecc.py

@@ -32,7 +32,7 @@ class BokeCCBaseIE(InfoExtractor):
 
 
 class BokeCCIE(BokeCCBaseIE):
-    _IE_DESC = 'CC视频'
+    IE_DESC = 'CC视频'
     _VALID_URL = r'https?://union\.bokecc\.com/playvideo\.bo\?(?P<query>.*)'
 
     _TESTS = [{

youtube_dl/extractor/cloudy.py

@@ -9,7 +9,7 @@ from ..utils import (
 
 
 class CloudyIE(InfoExtractor):
-    _IE_DESC = 'cloudy.ec'
+    IE_DESC = 'cloudy.ec'
     _VALID_URL = r'https?://(?:www\.)?cloudy\.ec/(?:v/|embed\.php\?.*?\bid=)(?P<id>[A-Za-z0-9]+)'
     _TESTS = [{
         'url': 'https://www.cloudy.ec/v/af511e2527aac',

youtube_dl/extractor/common.py

@@ -422,6 +422,8 @@ class InfoExtractor(object):
     _GEO_COUNTRIES = None
     _GEO_IP_BLOCKS = None
     _WORKING = True
+    # supply this in public subclasses: used in supported sites list, etc
+    # IE_DESC = 'short description of IE'
 
     def __init__(self, downloader=None):
         """Constructor. Receives an optional downloader."""

youtube_dl/extractor/itv.py

@@ -35,15 +35,6 @@ from ..utils import (
 
 class ITVBaseIE(InfoExtractor):
 
-    def _search_nextjs_data(self, webpage, video_id, **kw):
-        transform_source = kw.pop('transform_source', None)
-        fatal = kw.pop('fatal', True)
-        return self._parse_json(
-            self._search_regex(
-                r'''<script\b[^>]+\bid=('|")__NEXT_DATA__\1[^>]*>(?P<js>[^<]+)</script>''',
-                webpage, 'next.js data', group='js', fatal=fatal, **kw),
-            video_id, transform_source=transform_source, fatal=fatal)
-
     def __handle_request_webpage_error(self, err, video_id=None, errnote=None, fatal=True):
         if errnote is False:
             return False
@@ -109,7 +100,9 @@ class ITVBaseIE(InfoExtractor):
 
 class ITVIE(ITVBaseIE):
     _VALID_URL = r'https?://(?:www\.)?itv\.com/(?:(?P<w>watch)|hub)/[^/]+/(?(w)[\w-]+/)(?P<id>\w+)'
-    _IE_DESC = 'ITVX'
+    IE_DESC = 'ITVX'
+    _WORKING = False
+
     _TESTS = [{
         'note': 'Hub URLs redirect to ITVX',
         'url': 'https://www.itv.com/hub/liar/2a4547a0012',
@@ -270,7 +263,7 @@ class ITVIE(ITVBaseIE):
                 'ext': determine_ext(href, 'vtt'),
             })
 
-        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default='{}')
+        next_data = self._search_nextjs_data(webpage, video_id, fatal=False, default={})
         video_data.update(traverse_obj(next_data, ('props', 'pageProps', ('title', 'episode')), expected_type=dict)[0] or {})
         title = traverse_obj(video_data, 'headerTitle', 'episodeTitle')
         info = self._og_extract(webpage, require_title=not title)
@@ -323,7 +316,7 @@ class ITVIE(ITVBaseIE):
 
 class ITVBTCCIE(ITVBaseIE):
     _VALID_URL = r'https?://(?:www\.)?itv\.com/(?!(?:watch|hub)/)(?:[^/]+/)+(?P<id>[^/?#&]+)'
-    _IE_DESC = 'ITV articles: News, British Touring Car Championship'
+    IE_DESC = 'ITV articles: News, British Touring Car Championship'
     _TESTS = [{
         'note': 'British Touring Car Championship',
         'url': 'https://www.itv.com/btcc/articles/btcc-2018-all-the-action-from-brands-hatch',

youtube_dl/extractor/nbc.py

@@ -477,43 +477,94 @@ class NBCOlympicsIE(InfoExtractor):
 class NBCOlympicsStreamIE(AdobePassIE):
     IE_NAME = 'nbcolympics:stream'
     _VALID_URL = r'https?://stream\.nbcolympics\.com/(?P<id>[0-9a-z-]+)'
-    _TEST = {
+    _TESTS = [
-        'url': 'http://stream.nbcolympics.com/2018-winter-olympics-nbcsn-evening-feb-8',
+        # "Tokenized" .m3u8 source URL
-        'info_dict': {
+        {
-            'id': '203493',
+            'url': 'https://stream.nbcolympics.com/womens-soccer-group-round-11',
-            'ext': 'mp4',
+            'info_dict': {
-            'title': 're:Curling, Alpine, Luge [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+                'id': '2019740',
+                'ext': 'mp4',
+                'title': r"re:Women's Group Stage - Netherlands vs\. Brazil [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$",
+            },
+            'params': {
+                # m3u8 download
+                'skip_download': True,
+            },
         },
-        'params': {
+        # Plain .m3u8 source URL
-            # m3u8 download
+        {
-            'skip_download': True,
+            'url': 'https://stream.nbcolympics.com/gymnastics-event-finals-mens-floor-pommel-horse-womens-vault-bars',
+            'info_dict': {
+                'id': '2021729',
+                'ext': 'mp4',
+                'title': r're:Event Finals: M Floor, W Vault, M Pommel, W Uneven Bars [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$',
+            },
+            'params': {
+                # m3u8 download
+                'skip_download': True,
+            },
         },
-    }
+    ]
     _DATA_URL_TEMPLATE = 'http://stream.nbcolympics.com/data/%s_%s.json'
+    _LEAP_URL_TEMPLATE = 'https://api-leap.nbcsports.com/feeds/assets/%s?application=NBCOlympics&platform=%s&format=nbc-player&env=staging'
 
     def _real_extract(self, url):
         display_id = self._match_id(url)
         webpage = self._download_webpage(url, display_id)
         pid = self._search_regex(r'pid\s*=\s*(\d+);', webpage, 'pid')
-        resource = self._search_regex(
+
-            r"resource\s*=\s*'(.+)';", webpage,
-            'resource').replace("' + pid + '", pid)
         event_config = self._download_json(
             self._DATA_URL_TEMPLATE % ('event_config', pid),
-            pid)['eventConfig']
+            pid,
+            'Downloading event config',
+        )['eventConfig']
+        resource = event_config.get('resourceId', 'NBCOlympics')
         title = self._live_title(event_config['eventTitle'])
-        source_url = self._download_json(
+
-            self._DATA_URL_TEMPLATE % ('live_sources', pid),
+        leap_config = self._download_json(
-            pid)['videoSources'][0]['sourceUrl']
+            self._LEAP_URL_TEMPLATE % (pid, 'desktop'),
+            pid,
+            'Downloading leap config',
+        )
+        source_url = leap_config['videoSources'][0]['cdnSources']['primary'][0]['sourceUrl']
+
+        ap_resource = self._get_mvpd_resource(
+            resource,
+            re.sub(r'[^\w\d ]+', '', event_config['eventTitle']),
+            pid,
+            event_config.get('ratingId', 'NO VALUE'),
+        )
         media_token = self._extract_mvpd_auth(
-            url, pid, event_config.get('requestorId', 'NBCOlympics'), resource)
+            url, pid, event_config.get('requestorId', 'NBCOlympics'), ap_resource)
-        formats = self._extract_m3u8_formats(self._download_webpage(
+
-            'http://sp.auth.adobe.com/tvs/v1/sign', pid, query={
+        if event_config.get('cdnToken') is True:
-                'cdn': 'akamai',
+            source_url = self._download_json(
-                'mediaToken': base64.b64encode(media_token.encode()),
+                'https://tokens.playmakerservices.com/',
-                'resource': base64.b64encode(resource.encode()),
+                pid,
-                'url': source_url,
+                'Retrieving tokenized URL',
-            }), pid, 'mp4')
+                data=json.dumps({
+                    'application': 'NBCSports',
+                    'authentication-type': 'adobe-pass',
+                    'cdn': 'akamai',
+                    # Indicates that the player communicates its token not via the path
+                    # but via a cookie? NBC's player specifies `'false'` but field just
+                    # doesn't seem to have an effect.
+                    # 'inPath': 'false',
+                    'pid': pid,
+                    'platform': 'desktop',
+                    'requestorId': 'NBCOlympics',
+                    'resourceId': base64.b64encode(ap_resource.encode()).decode(),
+                    'token': base64.b64encode(media_token.encode()).decode(),
+                    'url': source_url,
+                    'version': 'v1',
+                }).encode(),
+            )['akamai'][0]['tokenizedUrl']
+
+        formats = self._extract_m3u8_formats(source_url, pid, 'mp4')
+        for f in formats:
+            f['_seekable'] = False
+            f['_http_seekable'] = False
+            f['_icy'] = False
         self._sort_formats(formats)
 
         return {

youtube_dl/extractor/senateisvp.py

@@ -47,7 +47,7 @@ class SenateISVPIE(InfoExtractor):
         ['vetaff', '76462', 'http://vetaff-f.akamaihd.net'],
         ['arch', '', 'http://ussenate-f.akamaihd.net/']
     ]
-    _IE_NAME = 'senate.gov'
+    IE_NAME = 'senate.gov'
     _VALID_URL = r'https?://(?:www\.)?senate\.gov/isvp/?\?(?P<qs>.+)'
     _TESTS = [{
         'url': 'http://www.senate.gov/isvp/?comm=judiciary&type=live&stt=&filename=judiciary031715&auto_play=false&wmode=transparent&poster=http%3A%2F%2Fwww.judiciary.senate.gov%2Fthemes%2Fjudiciary%2Fimages%2Fvideo-poster-flash-fit.png',

youtube_dl/jsinterp.py

@@ -686,6 +686,8 @@ class JSInterpreter(object):
             raise self.Exception('Cannot get index {idx!r:.100}'.format(**locals()), expr=repr(obj), cause=e)
 
     def _dump(self, obj, namespace):
+        if obj is JS_Undefined:
+            return 'undefined'
         try:
             return json.dumps(obj)
         except TypeError: