[ard] improve extraction(closes #23761)

- simplify extraction
- extract age limit and series
- bypass geo-restriction

1 files changed, 131 insertions, 126 deletions

diff --git a/youtube_dl/extractor/ard.py b/youtube_dl/extractor/ard.py
index 8adae4644..09d3ab4f9 100644
--- a/youtube_dl/extractor/ard.py
+++ b/youtube_dl/extractor/ard.py
@@ -1,6 +1,7 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import json
 import re
 
 from .common import InfoExtractor
@@ -22,66 +23,28 @@ from ..utils import (
 from ..compat import compat_etree_fromstring
 
 
-class ARDMediathekIE(InfoExtractor):
-    IE_NAME = 'ARD:mediathek'
-    _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
-
-    _TESTS = [{
-        # available till 26.07.2022
-        'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
-        'info_dict': {
-            'id': '44726822',
-            'ext': 'mp4',
-            'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
-            'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
-            'duration': 1740,
-        },
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        }
-    }, {
-        'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
-        'only_matching': True,
-    }, {
-        # audio
-        'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
-        'only_matching': True,
-    }, {
-        'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
-        'only_matching': True,
-    }, {
-        # audio
-        'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
-        'only_matching': True,
-    }, {
-        'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
-        'only_matching': True,
-    }]
-
-    @classmethod
-    def suitable(cls, url):
-        return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
+class ARDMediathekBaseIE(InfoExtractor):
+    _GEO_COUNTRIES = ['DE']
 
     def _extract_media_info(self, media_info_url, webpage, video_id):
         media_info = self._download_json(
             media_info_url, video_id, 'Downloading media JSON')
+        return self._parse_media_info(media_info, video_id, '"fsk"' in webpage)
 
+    def _parse_media_info(self, media_info, video_id, fsk):
         formats = self._extract_formats(media_info, video_id)
 
         if not formats:
-            if '"fsk"' in webpage:
+            if fsk:
                 raise ExtractorError(
                     'This video is only available after 20:00', expected=True)
             elif media_info.get('_geoblocked'):
-                raise ExtractorError('This video is not available due to geo restriction', expected=True)
+                self.raise_geo_restricted(
+                    'This video is not available due to geoblocking',
+                    countries=self._GEO_COUNTRIES)
 
         self._sort_formats(formats)
 
-        duration = int_or_none(media_info.get('_duration'))
-        thumbnail = media_info.get('_previewImage')
-        is_live = media_info.get('_isLive') is True
-
         subtitles = {}
         subtitle_url = media_info.get('_subtitleUrl')
         if subtitle_url:
@@ -92,9 +55,9 @@ class ARDMediathekIE(InfoExtractor):
 
         return {
             'id': video_id,
-            'duration': duration,
-            'thumbnail': thumbnail,
-            'is_live': is_live,
+            'duration': int_or_none(media_info.get('_duration')),
+            'thumbnail': media_info.get('_previewImage'),
+            'is_live': media_info.get('_isLive') is True,
             'formats': formats,
             'subtitles': subtitles,
         }
@@ -123,11 +86,11 @@ class ARDMediathekIE(InfoExtractor):
                             update_url_query(stream_url, {
                                 'hdcore': '3.1.1',
                                 'plugin': 'aasp-3.1.1.69.124'
-                            }),
-                            video_id, f4m_id='hds', fatal=False))
+                            }), video_id, f4m_id='hds', fatal=False))
                     elif ext == 'm3u8':
                         formats.extend(self._extract_m3u8_formats(
-                            stream_url, video_id, 'mp4', m3u8_id='hls', fatal=False))
+                            stream_url, video_id, 'mp4', 'm3u8_native',
+                            m3u8_id='hls', fatal=False))
                     else:
                         if server and server.startswith('rtmp'):
                             f = {
@@ -140,7 +103,9 @@ class ARDMediathekIE(InfoExtractor):
                                 'url': stream_url,
                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
                             }
-                        m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
+                        m = re.search(
+                            r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$',
+                            stream_url)
                         if m:
                             f.update({
                                 'width': int(m.group('width')),
@@ -151,6 +116,48 @@ class ARDMediathekIE(InfoExtractor):
                         formats.append(f)
         return formats
 
+
+class ARDMediathekIE(ARDMediathekBaseIE):
+    IE_NAME = 'ARD:mediathek'
+    _VALID_URL = r'^https?://(?:(?:(?:www|classic)\.)?ardmediathek\.de|mediathek\.(?:daserste|rbb-online)\.de|one\.ard\.de)/(?:.*/)(?P<video_id>[0-9]+|[^0-9][^/\?]+)[^/\?]*(?:\?.*)?'
+
+    _TESTS = [{
+        # available till 26.07.2022
+        'url': 'http://www.ardmediathek.de/tv/S%C3%9CDLICHT/Was-ist-die-Kunst-der-Zukunft-liebe-Ann/BR-Fernsehen/Video?bcastId=34633636&documentId=44726822',
+        'info_dict': {
+            'id': '44726822',
+            'ext': 'mp4',
+            'title': 'Was ist die Kunst der Zukunft, liebe Anna McCarthy?',
+            'description': 'md5:4ada28b3e3b5df01647310e41f3a62f5',
+            'duration': 1740,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        }
+    }, {
+        'url': 'https://one.ard.de/tv/Mord-mit-Aussicht/Mord-mit-Aussicht-6-39-T%C3%B6dliche-Nach/ONE/Video?bcastId=46384294&documentId=55586872',
+        'only_matching': True,
+    }, {
+        # audio
+        'url': 'http://www.ardmediathek.de/tv/WDR-H%C3%B6rspiel-Speicher/Tod-eines-Fu%C3%9Fballers/WDR-3/Audio-Podcast?documentId=28488308&bcastId=23074086',
+        'only_matching': True,
+    }, {
+        'url': 'http://mediathek.daserste.de/sendungen_a-z/328454_anne-will/22429276_vertrauen-ist-gut-spionieren-ist-besser-geht',
+        'only_matching': True,
+    }, {
+        # audio
+        'url': 'http://mediathek.rbb-online.de/radio/Hörspiel/Vor-dem-Fest/kulturradio/Audio?documentId=30796318&topRessort=radio&bcastId=9839158',
+        'only_matching': True,
+    }, {
+        'url': 'https://classic.ardmediathek.de/tv/Panda-Gorilla-Co/Panda-Gorilla-Co-Folge-274/Das-Erste/Video?bcastId=16355486&documentId=58234698',
+        'only_matching': True,
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if ARDBetaMediathekIE.suitable(url) else super(ARDMediathekIE, cls).suitable(url)
+
     def _real_extract(self, url):
         # determine video id from url
         m = re.match(self._VALID_URL, url)
@@ -302,19 +309,20 @@ class ARDIE(InfoExtractor):
         }
 
 
-class ARDBetaMediathekIE(InfoExtractor):
-    _VALID_URL = r'https://(?:beta|www)\.ardmediathek\.de/[^/]+/(?:player|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
+class ARDBetaMediathekIE(ARDMediathekBaseIE):
+    _VALID_URL = r'https://(?:beta|www)\.ardmediathek\.de/(?P<client>[^/]+)/(?:player|live)/(?P<video_id>[a-zA-Z0-9]+)(?:/(?P<display_id>[^/?#]+))?'
     _TESTS = [{
         'url': 'https://beta.ardmediathek.de/ard/player/Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE/die-robuste-roswita',
-        'md5': '2d02d996156ea3c397cfc5036b5d7f8f',
+        'md5': 'dfdc87d2e7e09d073d5a80770a9ce88f',
         'info_dict': {
             'display_id': 'die-robuste-roswita',
-            'id': 'Y3JpZDovL2Rhc2Vyc3RlLmRlL3RhdG9ydC9mYmM4NGM1NC0xNzU4LTRmZGYtYWFhZS0wYzcyZTIxNGEyMDE',
-            'title': 'Tatort: Die robuste Roswita',
+            'id': '70153354',
+            'title': 'Die robuste Roswita',
             'description': r're:^Der Mord.*trüber ist als die Ilm.',
             'duration': 5316,
-            'thumbnail': 'https://img.ardmediathek.de/standard/00/55/43/59/34/-1774185891/16x9/960?mandant=ard',
-            'upload_date': '20180826',
+            'thumbnail': 'https://img.ardmediathek.de/standard/00/70/15/33/90/-1852531467/16x9/960?mandant=ard',
+            'timestamp': 1577047500,
+            'upload_date': '20191222',
             'ext': 'mp4',
         },
     }, {
@@ -330,71 +338,68 @@ class ARDBetaMediathekIE(InfoExtractor):
         video_id = mobj.group('video_id')
         display_id = mobj.group('display_id') or video_id
 
-        webpage = self._download_webpage(url, display_id)
-        data_json = self._search_regex(r'window\.__APOLLO_STATE__\s*=\s*(\{.*);\n', webpage, 'json')
-        data = self._parse_json(data_json, display_id)
-
-        res = {
-            'id': video_id,
-            'display_id': display_id,
+        player_page = self._download_json(
+            'https://api.ardmediathek.de/public-gateway',
+            display_id, data=json.dumps({
+                'query': '''{
+  playerPage(client:"%s", clipId: "%s") {
+    blockedByFsk
+    broadcastedOn
+    maturityContentRating
+    mediaCollection {
+      _duration
+      _geoblocked
+      _isLive
+      _mediaArray {
+        _mediaStreamArray {
+          _quality
+          _server
+          _stream
         }
-        formats = []
-        subtitles = {}
-        geoblocked = False
-        for widget in data.values():
-            if widget.get('_geoblocked') is True:
-                geoblocked = True
-            if '_duration' in widget:
-                res['duration'] = int_or_none(widget['_duration'])
-            if 'clipTitle' in widget:
-                res['title'] = widget['clipTitle']
-            if '_previewImage' in widget:
-                res['thumbnail'] = widget['_previewImage']
-            if 'broadcastedOn' in widget:
-                res['timestamp'] = unified_timestamp(widget['broadcastedOn'])
-            if 'synopsis' in widget:
-                res['description'] = widget['synopsis']
-            subtitle_url = url_or_none(widget.get('_subtitleUrl'))
-            if subtitle_url:
-                subtitles.setdefault('de', []).append({
-                    'ext': 'ttml',
-                    'url': subtitle_url,
-                })
-            if '_quality' in widget:
-                format_url = url_or_none(try_get(
-                    widget, lambda x: x['_stream']['json'][0]))
-                if not format_url:
-                    continue
-                ext = determine_ext(format_url)
-                if ext == 'f4m':
-                    formats.extend(self._extract_f4m_formats(
-                        format_url + '?hdcore=3.11.0',
-                        video_id, f4m_id='hds', fatal=False))
-                elif ext == 'm3u8':
-                    formats.extend(self._extract_m3u8_formats(
-                        format_url, video_id, 'mp4', m3u8_id='hls',
-                        fatal=False))
-                else:
-                    # HTTP formats are not available when geoblocked is True,
-                    # other formats are fine though
-                    if geoblocked:
-                        continue
-                    quality = str_or_none(widget.get('_quality'))
-                    formats.append({
-                        'format_id': ('http-' + quality) if quality else 'http',
-                        'url': format_url,
-                        'preference': 10,  # Plain HTTP, that's nice
-                    })
-
-        if not formats and geoblocked:
-            self.raise_geo_restricted(
-                msg='This video is not available due to geoblocking',
-                countries=['DE'])
-
-        self._sort_formats(formats)
-        res.update({
-            'subtitles': subtitles,
-            'formats': formats,
+      }
+      _previewImage
+      _subtitleUrl
+      _type
+    }
+    show {
+      title
+    }
+    synopsis
+    title
+    tracking {
+      atiCustomVars {
+        contentId
+      }
+    }
+  }
+}''' % (mobj.group('client'), video_id),
+            }).encode(), headers={
+                'Content-Type': 'application/json'
+            })['data']['playerPage']
+        title = player_page['title']
+        content_id = str_or_none(try_get(
+            player_page, lambda x: x['tracking']['atiCustomVars']['contentId']))
+        media_collection = player_page.get('mediaCollection') or {}
+        if not media_collection and content_id:
+            media_collection = self._download_json(
+                'https://www.ardmediathek.de/play/media/' + content_id,
+                content_id, fatal=False) or {}
+        info = self._parse_media_info(
+            media_collection, content_id or video_id,
+            player_page.get('blockedByFsk'))
+        age_limit = None
+        description = player_page.get('synopsis')
+        maturity_content_rating = player_page.get('maturityContentRating')
+        if maturity_content_rating:
+            age_limit = int_or_none(maturity_content_rating.lstrip('FSK'))
+        if not age_limit:
+            age_limit = int_or_none(self._search_regex(r'\(FSK\s*(\d+)\)\s*$', description, 'age limit', default=None))
+        info.update({
+            'age_limit': age_limit,
+            'display_id': display_id,
+            'title': title,
+            'description': description,
+            'timestamp': unified_timestamp(player_page.get('broadcastedOn')),
+            'series': try_get(player_page, lambda x: x['show']['title']),
         })
-
-        return res
+        return info