Merge 97c1053ba81c29cb34066e9ce3fd5c353d1fcf84 into 3eb8d22ddb8982ca4fb56bb7a8d6517538bf14c6

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

youtube_dl/extractor/bbc.py

@@ -25,6 +25,7 @@ from ..utils import (
     get_element_by_class,
     int_or_none,
     js_to_json,
+    parse_bitrate,
     parse_duration,
     parse_iso8601,
     strip_or_none,
@@ -647,9 +648,7 @@ class BBCIE(BBCCoUkIE):
             'skip_download': True,
         }
     }, {
-        # article with single video embedded with data-playable containing XML playlist
-        # with direct video links as progressiveDownloadUrl (for now these are extracted)
-        # and playlist with f4m and m3u8 as streamingUrl
+        # article with single video (formerly) embedded, now using SIMORGH_DATA JSON
         'url': 'http://www.bbc.com/turkce/haberler/2015/06/150615_telabyad_kentin_cogu',
         'info_dict': {
             'id': '150615_telabyad_kentin_cogu',
@@ -661,12 +660,13 @@ class BBCIE(BBCCoUkIE):
         },
         'params': {
             'skip_download': True,
-        }
+        },
+        'skip': 'Video no longer embedded, 2021',
     }, {
-        # single video embedded with data-playable containing XML playlists (regional section)
+        # single video embedded, legacy media, in promo object of SIMORGH_DATA JSON
         'url': 'http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw',
         'info_dict': {
-            'id': '150619_video_honduras_militares_hospitales_corrupcion_aw',
+            'id': '39275083',
             'ext': 'mp4',
             'title': 'Honduras militariza sus hospitales por nuevo escándalo de corrupción',
             'description': 'md5:1525f17448c4ee262b64b8f0c9ce66c8',
@@ -845,6 +845,18 @@ class BBCIE(BBCCoUkIE):
             'upload_date': '20190604',
             'categories': ['Psychology'],
         },
+    }, {
+        # BBC World Service etc: media nested in content object of SIMORGH_DATA JSON
+        'url': 'http://www.bbc.co.uk/scotland/articles/cm49v4x1r9lo',
+        'info_dict': {
+            'id': 'p06p040v',
+            'ext': 'mp4',
+            'title': 'Five things ants can teach us about management',
+            'description': 'They may be tiny, but us humans could learn a thing or two from ants.',
+            'duration': 191,
+            'thumbnail': r're:https?://.+/p06p0qzv.jpg',
+            'upload_date': '20181016',
+        },
     }]
 
     @classmethod
@@ -1107,6 +1119,99 @@ class BBCIE(BBCCoUkIE):
                     'subtitles': subtitles,
                 }
 
+        # simorgh-based playlist (see https://github.com/bbc/simorgh)
+        # JSON assigned to window.SIMORGH_DATA in a <script> element
+        simorgh_data = self._parse_json(
+            self._search_regex(
+                r'window\.SIMORGH_DATA\s*=\s*(\{[^<]+})\s*</',
+                webpage, 'simorgh playlist', default='{}'),
+            playlist_id, fatal=False)
+        # legacy media, video in promo object (eg, http://www.bbc.com/mundo/video_fotos/2015/06/150619_video_honduras_militares_hospitales_corrupcion_aw)
+        playlist = try_get(simorgh_data, lambda x: x['pageData']['promo']['media']['playlist']) or []
+        if playlist:
+            media = simorgh_data['pageData']['promo']
+            if media['media'].get('format') == 'video':
+                media.update(media['media'])
+                title = (dict_get(media.get('headlines') or {},
+                                  ('shortHeadline', 'headline'))
+                         or playlist_title),
+                programme_id = media.get('id')
+                if programme_id and title:
+                    formats = []
+                    keys = {'url', 'format', 'format_id', 'language', 'quality', 'tbr', 'resolution'}
+                    for format in playlist:
+                        if not (format.get('url') and format.get('format')):
+                            continue
+                        bitrate = format.pop('bitrate')
+                        format['tbr'] = int_or_none(bitrate, scale=1000) or parse_bitrate(bitrate)
+                        format['language'] = media.get('language')
+                        # format id: penultimate item from the url split on _ and .
+                        (fmt,) = re.split('[_.]', format['url'])[-2:][:1]
+                        format['format_id'] = '%s_%s' % (format['format'], fmt)
+                        # try to set resolution using any available data
+                        aspect_ratio = re.split(r'[xX:]', media.get('aspectRatio') or '')
+                        if len(aspect_ratio) != 2:
+                            aspect_ratio = None
+                        else:
+                            aspect_ratio = float_or_none(aspect_ratio[0], scale=aspect_ratio[1])
+                        # these may not be present, but try anyway
+                        width = int_or_none(format.get('width'))
+                        height = int_or_none(format.get('height'))
+                        if (not height) and aspect_ratio:
+                            height = int(width / aspect_ratio)
+                        elif (not width) and aspect_ratio:
+                            width = int(height * aspect_ratio)
+                        format['resolution'] = ('%dx%d' % (width, height) if width and height
+                                                else dict_get(format, ('resolution', 'res'), default=fmt))
+                        format['quality'] = -1
+                        formats.append(dict((k, format[k]) for k in keys))
+                    self._sort_formats(formats)
+                    return {
+                        'id': programme_id,
+                        'title': title,
+                        'description': media.get('summary') or playlist_description,
+                        'formats': formats,
+                        'subtitles': None,
+                        'thumbnail': try_get(media, lambda x: x['image']['href']),
+                        'timestamp': int_or_none(media.get('timestamp'), scale=1000)
+                    }
+
+        # general case: media nested in content object
+        # test: https://www.bbc.co.uk/scotland/articles/cm49v4x1r9lo
+        if simorgh_data:
+
+            def extract_media_from_simorgh(model):
+                if not isinstance(model, dict):
+                    return
+                for block in model.get('blocks') or {}:
+                    if block.get('type') == 'aresMediaMetadata':
+                        vpid = try_get(block, lambda x: x['model']['versions'][0]['versionId'])
+                        if vpid:
+                            formats, subtitles = self._download_media_selector(vpid)
+                            self._sort_formats(formats)
+                            model = block['model']
+                            version = model['versions'][0]
+                            thumbnail = model.get('imageUrl')
+                            return {
+                                'id': vpid,
+                                'title': model.get('title') or 'unnamed clip',
+                                'description': dict_get(model.get('synopses') or {}, ('long', 'medium', 'short')),
+                                'duration': (int_or_none(version.get('duration'))
+                                             or parse_duration(version.get('durationISO8601'))),
+                                'timestamp': version.get('availableFrom'),
+                                'thumbnail': urljoin(url, thumbnail.replace('$recipe', 'raw')) if thumbnail else None,
+                                'formats': formats,
+                                'subtitles': subtitles,
+                            }
+                    else:
+                        entry = extract_media_from_simorgh(block.get('model'))
+                        if entry:
+                            return entry
+
+            playlist = extract_media_from_simorgh(try_get(simorgh_data, lambda x: x['pageData']['content']['model']))
+            if playlist:
+                return playlist
+
         preload_state = self._parse_json(self._search_regex(
             r'window\.__PRELOADED_STATE__\s*=\s*({.+?});', webpage,
             'preload state', default='{}'), playlist_id, fatal=False)

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/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: