[extractor/hrfensehen] Fix extractor (#5096)

Authored by: snapdgn

diff --git a/yt_dlp/extractor/hrfensehen.py b/yt_dlp/extractor/hrfensehen.py
index 6f7ed9b4bc92a0c5018cd98b63b2cb6024ee2317..dd72d86d77c33068317dd9fb28c62498a7502a3a 100644 (file)
--- a/yt_dlp/extractor/hrfensehen.py
+++ b/yt_dlp/extractor/hrfensehen.py
@@ -1,14 +1,19 @@
 import json
 import re
 
-from ..utils import int_or_none, unified_timestamp, unescapeHTML
+from ..utils import (
+    int_or_none,
+    traverse_obj,
+    try_call,
+    unescapeHTML,
+    unified_timestamp,
+)
 from .common import InfoExtractor
 
 
 class HRFernsehenIE(InfoExtractor):
     IE_NAME = 'hrfernsehen'
     _VALID_URL = r'^https?://www\.(?:hr-fernsehen|hessenschau)\.de/.*,video-(?P<id>[0-9]{6})\.html'
-
     _TESTS = [{
         'url': 'https://www.hessenschau.de/tv-sendung/hessenschau-vom-26082020,video-130546.html',
         'md5': '5c4e0ba94677c516a2f65a84110fc536',
@@ -21,10 +26,11 @@ class HRFernsehenIE(InfoExtractor):
             'subtitles': {'de': [{
                 'url': 'https://hr-a.akamaihd.net/video/as/hessenschau/2020_08/hrLogo_200826200407_L385592_512x288-25p-500kbit.vtt'
             }]},
-            'timestamp': 1598470200,
+            'timestamp': 1598400000,
             'upload_date': '20200826',
-            'thumbnail': 'https://www.hessenschau.de/tv-sendung/hs_ganz-1554~_t-1598465545029_v-16to9__medium.jpg',
-            'title': 'hessenschau vom 26.08.2020'
+            'thumbnail': 'https://www.hessenschau.de/tv-sendung/hs_ganz-1554~_t-1598465545029_v-16to9.jpg',
+            'title': 'hessenschau vom 26.08.2020',
+            'duration': 1654
         }
     }, {
         'url': 'https://www.hr-fernsehen.de/sendungen-a-z/mex/sendungen/fair-und-gut---was-hinter-aldis-eigenem-guetesiegel-steckt,video-130544.html',
@@ -33,25 +39,18 @@ class HRFernsehenIE(InfoExtractor):
 
     _GEO_COUNTRIES = ['DE']
 
-    def extract_airdate(self, loader_data):
-        airdate_str = loader_data.get('mediaMetadata', {}).get('agf', {}).get('airdate')
-
-        if airdate_str is None:
-            return None
-
-        return unified_timestamp(airdate_str)
-
     def extract_formats(self, loader_data):
         stream_formats = []
-        for stream_obj in loader_data["videoResolutionLevels"]:
+        data = loader_data['mediaCollection']['streams'][0]['media']
+        for inner in data[1:]:
             stream_format = {
-                'format_id': str(stream_obj['verticalResolution']) + "p",
-                'height': stream_obj['verticalResolution'],
-                'url': stream_obj['url'],
+                'format_id': try_call(lambda: f'{inner["maxHResolutionPx"]}p'),
+                'height': inner.get('maxHResolutionPx'),
+                'url': inner['url'],
             }
 
             quality_information = re.search(r'([0-9]{3,4})x([0-9]{3,4})-([0-9]{2})p-([0-9]{3,4})kbit',
-                                            stream_obj['url'])
+                                            inner['url'])
             if quality_information:
                 stream_format['width'] = int_or_none(quality_information.group(1))
                 stream_format['height'] = int_or_none(quality_information.group(2))
@@ -72,22 +71,22 @@ def _real_extract(self, url):
         description = self._html_search_meta(
             ['description'], webpage)
 
-        loader_str = unescapeHTML(self._search_regex(r"data-new-hr-mediaplayer-loader='([^']*)'", webpage, "ardloader"))
+        loader_str = unescapeHTML(self._search_regex(r"data-(?:new-)?hr-mediaplayer-loader='([^']*)'", webpage, 'ardloader'))
         loader_data = json.loads(loader_str)
 
+        subtitle = traverse_obj(loader_data, ('mediaCollection', 'subTitles', 0, 'sources', 0, 'url'))
+
         info = {
             'id': video_id,
             'title': title,
             'description': description,
             'formats': self.extract_formats(loader_data),
-            'timestamp': self.extract_airdate(loader_data)
+            'subtitles': {'de': [{'url': subtitle}]},
+            'timestamp': unified_timestamp(self._search_regex(
+                r'<time\sdatetime="(\d{4}\W\d{1,2}\W\d{1,2})', webpage, 'datetime', fatal=False)),
+            'duration': int_or_none(traverse_obj(
+                loader_data, ('playerConfig', 'pluginData', 'trackingAti@all', 'richMedia', 'duration'))),
+            'thumbnail': self._search_regex(r'thumbnailUrl\W*([^"]+)', webpage, 'thumbnail', default=None),
         }
 
-        if "subtitle" in loader_data:
-            info["subtitles"] = {"de": [{"url": loader_data["subtitle"]}]}
-
-        thumbnails = list(set([t for t in loader_data.get("previewImageUrl", {}).values()]))
-        if len(thumbnails) > 0:
-            info["thumbnails"] = [{"url": t} for t in thumbnails]
-
         return info