yt_dlp/extractor/ondemandkorea.py

@@ -1,6 +1,5 @@
 import functools
 import re
-import uuid
 
 from .common import InfoExtractor
 from ..networking import HEADRequest
@@ -12,6 +11,7 @@ from ..utils import (
     join_nonempty,
     parse_age_limit,
     parse_qs,
+    random_uuidv4,
     unified_strdate,
     url_or_none,
 )
@@ -39,7 +39,7 @@ class OnDemandKoreaIE(InfoExtractor):
         },
     }, {
         'url': 'https://www.ondemandkorea.com/player/vod/breakup-probation-a-week?contentId=1595796',
-        'md5': '57266c720006962be7ff415b24775caa',
+        'md5': '44e274d2b04977e03fc7f3941fbcb355',
         'info_dict': {
             'id': '1595796',
             'ext': 'mp4',
@@ -64,65 +64,58 @@ class OnDemandKoreaIE(InfoExtractor):
             'thumbnail': r're:^https?://.*\.(jpg|jpeg|png)',
             'age_limit': 18,
         },
-    }, {
-        'url': 'https://www.ondemandkorea.com/en/player/vod/capture-the-moment-how-is-that-possible?contentId=1605006',
-        'only_matching': True,
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        data = self._download_json(
+        data = self._download_json(f'https://odkmedia.io/odx/api/v3/playback/{video_id}/', video_id,
-            f'https://odkmedia.io/odx/api/v3/playback/{video_id}/', video_id, fatal=False,
+                                   fatal=False, headers={'service-name': 'odk'},
-            headers={'service-name': 'odk'}, query={'did': str(uuid.uuid4())}, expected_status=(403, 404))
+                                   query={'did': random_uuidv4()}, expected_status=(403, 404))
-        if not traverse_obj(data, ('result', {dict})):
+        if not data.get('result'):
-            msg = traverse_obj(data, ('messages', '__default'), 'title', expected_type=str)
+            raise ExtractorError(traverse_obj(data, ('messages', '__default'), 'title'), expected=True)
-            raise ExtractorError(msg or 'Got empty response from playback API', expected=True)
 
-        data = data['result']
+        potential_urls = traverse_obj(data, ('result', ('sources', 'manifest'), ..., 'url'))
-
+        # Try to bypass geo-restricted ad proxy
-        def try_geo_bypass(url):
+        potential_urls = [
-            return traverse_obj(url, ({parse_qs}, 'stream_url', 0, {url_or_none})) or url
+            alt_url if (alt_url := traverse_obj(url, ({parse_qs}, 'stream_url', 0, {url_or_none}))) else url
-
+            for url in potential_urls
-        def try_upgrade_quality(url):
+        ]
-            mod_url = re.sub(r'_720(p?)\.m3u8', r'_1080\1.m3u8', url)
+        # Try to upgrade quality
-            return mod_url if mod_url != url and self._request_webpage(
+        potential_urls = [
-                HEADRequest(mod_url), video_id, note='Checking for higher quality format',
+            mod_url if self._request_webpage(
-                errnote='No higher quality format found', fatal=False) else url
+                HEADRequest(mod_url := re.sub(r'_720(p?)\.m3u8', r'_1080\1.m3u8', url)), video_id,
-
+                note='Checking if higher quality format is available', fatal=False) else url
-        potential_urls = traverse_obj(
+            for url in potential_urls
-            data, (('sources', 'manifest'), ..., 'url', {url_or_none}, {try_geo_bypass}, {try_upgrade_quality}))
+        ]
 
         formats = []
         for url in potential_urls:
             formats.extend(self._extract_m3u8_formats(url, video_id, fatal=False))
 
         subtitles = {}
-        for track in traverse_obj(data, ('text_tracks', lambda _, v: url_or_none(v['url']))):
+        for track in traverse_obj(data, ('result', 'text_tracks', lambda _, v: url_or_none(v['url']))):
             subtitles.setdefault(track.get('language', 'und'), []).append({
                 'url': track['url'],
                 'ext': track.get('codec'),
                 'name': track.get('label'),
             })
 
-        def if_series(key=None):
-            return lambda obj: obj[key] if key and obj['kind'] == 'series' else None
-
         return {
             'id': video_id,
             'title': join_nonempty(
-                ('episode', 'program', 'title'),
+                ('result', 'episode', 'program', 'title'),
-                ('episode', 'title'), from_dict=data, delim=': '),
+                ('result', 'episode', 'title'), from_dict=data, delim=': '),
-            **traverse_obj(data, {
+            **traverse_obj(data, ('result', {
                 'thumbnail': ('episode', 'images', 'thumbnail', {url_or_none}),
                 'release_date': ('episode', 'release_date', {lambda x: x.replace('-', '')}, {unified_strdate}),
                 'duration': ('duration', {functools.partial(float_or_none, scale=1000)}),
                 'age_limit': ('age_rating', 'name', {lambda x: x.replace('R', '')}, {parse_age_limit}),
-                'series': ('episode', {if_series(key='program')}, 'title'),
+                'series': ('episode', {lambda x: x['program'] if x['kind'] == 'series' else None}, 'title'),
-                'series_id': ('episode', {if_series(key='program')}, 'id'),
+                'series_id': ('episode', {lambda x: x['program'] if x['kind'] == 'series' else None}, 'id'),
-                'episode': ('episode', {if_series(key='title')}),
+                'episode': ('episode', {lambda x: x['title'] if x['kind'] == 'series' else None},),
-                'episode_number': ('episode', {if_series(key='number')}, {int_or_none}),
+                'episode_number': ('episode', {lambda x: x['number'] if x['kind'] == 'series' else None}, {int_or_none}),
-            }, get_all=False),
+            }), get_all=False),
             'formats': formats,
             'subtitles': subtitles,
         }
@@ -155,11 +148,12 @@ class OnDemandKoreaProgramIE(InfoExtractor):
             headers={'service-name': 'odk'}, query={
                 'page': page,
                 'page_size': self._PAGE_SIZE,
-            }, note=f'Downloading page {page}', expected_status=404)
+            }, note=f'Downloading page {page}')
-        for episode in traverse_obj(page_data, ('result', 'results', ...)):
+        for episode in traverse_obj(page_data, ('result', 'results')):
             yield self.url_result(
                 f'https://www.ondemandkorea.com/player/vod/{display_id}?contentId={episode["id"]}',
-                ie=OnDemandKoreaIE, video_title=episode.get('title'))
+                ie=OnDemandKoreaIE, video_title=episode.get('title')
+            )
 
     def _real_extract(self, url):
         display_id = self._match_id(url)