[npo] Fix extraction (#20084)

youtube_dl/extractor/npo.py

@@ -12,11 +12,16 @@
     ExtractorError,
     fix_xml_ampersands,
     int_or_none,
+    merge_dicts,
     orderedSet,
     parse_duration,
     qualities,
+    str_or_none,
     strip_jsonp,
     unified_strdate,
+    unified_timestamp,
+    url_or_none,
+    urlencode_postdata,
 )
 
 
@@ -176,9 +181,118 @@ def suitable(cls, url):
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        return self._get_info(video_id)
+        try:
+            return self._get_info(url, video_id)
+        except ExtractorError:
+            return self._get_old_info(video_id)
 
-    def _get_info(self, video_id):
+    def _get_info(self, url, video_id):
+        token = self._download_json(
+            'https://www.npostart.nl/api/token', video_id,
+            'Downloading token', headers={
+                'Referer': url,
+                'X-Requested-With': 'XMLHttpRequest',
+            })['token']
+
+        player = self._download_json(
+            'https://www.npostart.nl/player/%s' % video_id, video_id,
+            'Downloading player JSON', data=urlencode_postdata({
+                'autoplay': 0,
+                'share': 1,
+                'pageUrl': url,
+                'hasAdConsent': 0,
+                '_token': token,
+            }))
+
+        player_token = player['token']
+
+        format_urls = set()
+        formats = []
+        for profile in ('hls', 'dash-widevine', 'dash-playready', 'smooth'):
+            streams = self._download_json(
+                'https://start-player.npo.nl/video/%s/streams' % video_id,
+                video_id, 'Downloading %s profile JSON' % profile, fatal=False,
+                query={
+                    'profile': profile,
+                    'quality': 'npo',
+                    'tokenId': player_token,
+                    'streamType': 'broadcast',
+                })
+            if not streams:
+                continue
+            stream = streams.get('stream')
+            if not isinstance(stream, dict):
+                continue
+            stream_url = url_or_none(stream.get('src'))
+            if not stream_url or stream_url in format_urls:
+                continue
+            format_urls.add(stream_url)
+            if stream.get('protection') is not None:
+                continue
+            stream_type = stream.get('type')
+            stream_ext = determine_ext(stream_url)
+            if stream_type == 'application/dash+xml' or stream_ext == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    stream_url, video_id, mpd_id='dash', fatal=False))
+            elif stream_type == 'application/vnd.apple.mpegurl' or stream_ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    stream_url, video_id, ext='mp4',
+                    entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
+            elif '.ism/Manifest' in stream_url:
+                formats.extend(self._extract_ism_formats(
+                    stream_url, video_id, ism_id='mss', fatal=False))
+            else:
+                formats.append({
+                    'url': stream_url,
+                })
+
+        self._sort_formats(formats)
+
+        info = {
+            'id': video_id,
+            'title': video_id,
+            'formats': formats,
+        }
+
+        embed_url = url_or_none(player.get('embedUrl'))
+        if embed_url:
+            webpage = self._download_webpage(
+                embed_url, video_id, 'Downloading embed page', fatal=False)
+            if webpage:
+                video = self._parse_json(
+                    self._search_regex(
+                        r'\bvideo\s*=\s*({.+?})\s*;', webpage, 'video',
+                        default='{}'), video_id)
+                if video:
+                    title = video.get('episodeTitle')
+                    subtitles = {}
+                    subtitles_list = video.get('subtitles')
+                    if isinstance(subtitles_list, list):
+                        for cc in subtitles_list:
+                            cc_url = url_or_none(cc.get('src'))
+                            if not cc_url:
+                                continue
+                            lang = str_or_none(cc.get('language')) or 'nl'
+                            subtitles.setdefault(lang, []).append({
+                                'url': cc_url,
+                            })
+                    return merge_dicts({
+                        'title': title,
+                        'description': video.get('description'),
+                        'thumbnail': url_or_none(
+                            video.get('still_image_url') or video.get('orig_image_url')),
+                        'duration': int_or_none(video.get('duration')),
+                        'timestamp': unified_timestamp(video.get('broadcastDate')),
+                        'creator': video.get('channel'),
+                        'series': video.get('title'),
+                        'episode': title,
+                        'episode_number': int_or_none(video.get('episodeNumber')),
+                        'subtitles': subtitles,
+                    }, info)
+
+        return info
+
+    def _get_old_info(self, video_id):
         metadata = self._download_json(
             'http://e.omroep.nl/metadata/%s' % video_id,
             video_id,
@@ -280,7 +394,7 @@ def add_format_url(format_url):
             # JSON
             else:
                 video_url = stream_info.get('url')
-            if not video_url or video_url in urls:
+            if not video_url or 'vodnotavailable.' in video_url or video_url in urls:
                 continue
             urls.add(video_url)
             if determine_ext(video_url) == 'm3u8':