[nova:embed] Fix extraction (closes #24700)

youtube_dl/extractor/nova.py

@@ -6,6 +6,7 @@
 from .common import InfoExtractor
 from ..utils import (
     clean_html,
+    determine_ext,
     int_or_none,
     js_to_json,
     qualities,
@@ -33,6 +34,40 @@ def _real_extract(self, url):
 
         webpage = self._download_webpage(url, video_id)
 
+        duration = None
+        formats = []
+
+        player = self._parse_json(
+            self._search_regex(
+                r'Player\.init\s*\([^,]+,\s*({.+?})\s*,\s*{.+?}\s*\)\s*;',
+                webpage, 'player', default='{}'), video_id, fatal=False)
+        if player:
+            for format_id, format_list in player['tracks'].items():
+                if not isinstance(format_list, list):
+                    format_list = [format_list]
+                for format_dict in format_list:
+                    if not isinstance(format_dict, dict):
+                        continue
+                    format_url = url_or_none(format_dict.get('src'))
+                    format_type = format_dict.get('type')
+                    ext = determine_ext(format_url)
+                    if (format_type == 'application/x-mpegURL'
+                            or format_id == 'HLS' or ext == 'm3u8'):
+                        formats.extend(self._extract_m3u8_formats(
+                            format_url, video_id, 'mp4',
+                            entry_protocol='m3u8_native', m3u8_id='hls',
+                            fatal=False))
+                    elif (format_type == 'application/dash+xml'
+                          or format_id == 'DASH' or ext == 'mpd'):
+                        formats.extend(self._extract_mpd_formats(
+                            format_url, video_id, mpd_id='dash', fatal=False))
+                    else:
+                        formats.append({
+                            'url': format_url,
+                        })
+            duration = int_or_none(player.get('duration'))
+        else:
+            # Old path, not actual as of 08.04.2020
             bitrates = self._parse_json(
                 self._search_regex(
                     r'(?s)(?:src|bitrates)\s*=\s*({.+?})\s*;', webpage, 'formats'),
@@ -41,7 +76,6 @@ def _real_extract(self, url):
             QUALITIES = ('lq', 'mq', 'hq', 'hd')
             quality_key = qualities(QUALITIES)
 
-        formats = []
             for format_id, format_list in bitrates.items():
                 if not isinstance(format_list, list):
                     format_list = [format_list]
@@ -69,6 +103,7 @@ def _real_extract(self, url):
                             break
                     f['format_id'] = f_id
                     formats.append(f)
+
         self._sort_formats(formats)
 
         title = self._og_search_title(
@@ -81,7 +116,8 @@ def _real_extract(self, url):
             r'poster\s*:\s*(["\'])(?P<value>(?:(?!\1).)+)\1', webpage,
             'thumbnail', fatal=False, group='value')
         duration = int_or_none(self._search_regex(
-            r'videoDuration\s*:\s*(\d+)', webpage, 'duration', fatal=False))
+            r'videoDuration\s*:\s*(\d+)', webpage, 'duration',
+            default=duration))
 
         return {
             'id': video_id,