From ebea243ea248ad92bdcb3a21fc904f3164b3941e Mon Sep 17 00:00:00 2001
From: Florent DELAHAYE <florent@delahaye.me>
Date: Sun, 27 Jul 2025 10:41:03 +0200
Subject: [PATCH] feat(extractor/applepodcasts): add playlist support

---
 yt_dlp/extractor/_extractors.py   |   5 +-
 yt_dlp/extractor/applepodcasts.py | 142 +++++++++++++++++++++++++++++-
 2 files changed, 142 insertions(+), 5 deletions(-)

diff --git a/yt_dlp/extractor/_extractors.py b/yt_dlp/extractor/_extractors.py
index 617c2c5ce0..84608b2efc 100644
--- a/yt_dlp/extractor/_extractors.py
+++ b/yt_dlp/extractor/_extractors.py
@@ -137,7 +137,10 @@ from .aol import AolIE
 from .apa import APAIE
 from .aparat import AparatIE
 from .appleconnect import AppleConnectIE
-from .applepodcasts import ApplePodcastsIE
+from .applepodcasts import (
+    ApplePodcastsIE,
+    ApplePodcastsPlaylistIE
+)
 from .appletrailers import (
     AppleTrailersIE,
     AppleTrailersSectionIE,
diff --git a/yt_dlp/extractor/applepodcasts.py b/yt_dlp/extractor/applepodcasts.py
index b99d24e0eb..5c20dbaff1 100644
--- a/yt_dlp/extractor/applepodcasts.py
+++ b/yt_dlp/extractor/applepodcasts.py
@@ -5,10 +5,16 @@ from ..utils import (
     parse_iso8601,
 )
 from ..utils.traversal import traverse_obj
+import re
+from urllib.parse import urljoin
 
+class ApplePodcastsBaseIE(InfoExtractor):
+    _BASE_URL_REGEX = r'https?://podcasts\.apple\.com/(?:[^/]+/)?podcast(?:/[^/]+){1,2}'
+    _BASE_HTML_JSON_LOCATION = r'<script [^>]*\bid=["\']serialized-server-data["\'][^>]*>'
+    _BASE_HTML_JSON_PATTERN = r'\[{(?s:.+)}\]'
 
-class ApplePodcastsIE(InfoExtractor):
-    _VALID_URL = r'https?://podcasts\.apple\.com/(?:[^/]+/)?podcast(?:/[^/]+){1,2}.*?\bi=(?P<id>\d+)'
+class ApplePodcastsIE(ApplePodcastsBaseIE):
+    _VALID_URL = ApplePodcastsBaseIE._BASE_URL_REGEX + r'.*?\bi=(?P<id>\d+)'
     _TESTS = [{
         'url': 'https://podcasts.apple.com/us/podcast/ferreck-dawn-to-the-break-of-dawn-117/id1625658232?i=1000665010654',
         'md5': '82cc219b8cc1dcf8bfc5a5e99b23b172',
@@ -55,8 +61,8 @@ class ApplePodcastsIE(InfoExtractor):
         episode_id = self._match_id(url)
         webpage = self._download_webpage(url, episode_id)
         server_data = self._search_json(
-            r'<script [^>]*\bid=["\']serialized-server-data["\'][^>]*>', webpage,
-            'server data', episode_id, contains_pattern=r'\[{(?s:.+)}\]')[0]['data']
+            ApplePodcastsBaseIE._BASE_HTML_JSON_LOCATION, webpage,
+            'server data', episode_id, contains_pattern=ApplePodcastsBaseIE._BASE_HTML_JSON_PATTERN)[0]['data']
         model_data = traverse_obj(server_data, (
             'headerButtonItems', lambda _, v: v['$kind'] == 'bookmark' and v['modelType'] == 'EpisodeOffer',
             'model', {dict}, any))
@@ -75,3 +81,131 @@ class ApplePodcastsIE(InfoExtractor):
             'thumbnail': self._og_search_thumbnail(webpage),
             'vcodec': 'none',
         }
+
+class ApplePodcastsPlaylistIE(ApplePodcastsBaseIE):
+    # Apple podcast items are partially described (last episodes only) in the embedded json from main page therefore API calls are mandatory to get a full list
+
+    _VALID_URL = ApplePodcastsBaseIE._BASE_URL_REGEX + r'/id(?P<id>\d+)'
+
+    _TESTS = [{
+        'url': 'https://podcasts.apple.com/fr/podcast/id1691740320',
+        'info_dict': {
+            'id': '1691740320',
+            'title': 'LEGEND',
+            'playlist_uploader': 'Guillaume Pley'
+        },
+        'playlist_mincount': 400,
+            'playlist_entries': [
+                {
+                    'id': '1000718966711',
+                    'title': 'MICHAEL YOUN : LES MOMENTS LES PLUS FOUS DE SES 25 ANS DE CARRIÈRE (MORNING LIVE, FATAL 2…)',
+                    'uploader': 'Guillaume Pley',
+                    'description': "Retrouvez la boutique LEGEND ➡️: https://shop.legend-group.fr/\nMerci à Michaël Youn d'être venu nous voir sur LEGEND. Il est venu nous raconter plus de 25 ans de carrière, en tant qu’acteur, réalisateur et artiste. Il a été révélé par l'émission Morning Live sur M6, il nous a livré les anecdotes les plus folles qu’il a vécues à cette époque. Il est aussi venu nous raconter comment il a rencontré sa femme et ce que ses enfants ont changé dans sa vie.\nPour voir la bande annonce du film « Certains l’aiment chauve » déjà disponible au cinéma ➡️ https://www.allocine.fr/film/fichefilm_gen_cfilm=1000007354.html\nRetrouvez l'interview complète sur YouTube ➡️ https://youtu.be/_TXBz1dSfBw\nPour toutes demandes de partenariats : legend@influxcrew.com\nRetrouvez-nous sur tous les réseaux LEGEND !\nFacebook : https://www.facebook.com/legendmediafr\nInstagram : https://www.instagram.com/legendmedia/\nTikTok : https://www.tiktok.com/@legend\nTwitter : https://twitter.com/legendmediafr\nSnapchat : https://t.snapchat.com/CgEvsbWV\n Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.",
+                    'release_timestamp': 1753434168,
+                    'duration': 6856,
+                    'url': 'https://podcasts.apple.com/fr/podcast/michael-youn-les-moments-les-plus-fous-de-ses-25-ans/id1691740320?i=1000718966711'
+                },
+                {
+                    'id': '1000718672235',
+                    'title': 'AMBULANCIER DU SAMU: SES INTERVENTIONS IMPROBABLES (SUIC*DES, FAUX MALADES, ENFANTS DR0GUÉS)',
+                    'uploader': 'Guillaume Pley',
+                    'description': "Retrouvez la boutique LEGEND ➡️: https://shop.legend-group.fr/\nMerci à Thomas d’être passé nous voir chez LEGEND ! Thomas est ambulancier et urgentiste au SMUR depuis 10 ans. Il est venu partager avec nous ses anecdotes les plus marquantes.\nIl a vécu des interventions difficiles, comme sur une scène de crime où une mère avait tué ses deux enfants, ou encore ce jour où il a pris en charge une victime coupée en deux par un hachoir.\nMais son métier, c’est aussi des moments plus légers, parfois même drôles, comme cette fois où il a dû intervenir sur le tournage d’un film X pour secourir des acteurs.\nPour toutes demandes de partenariats : legend@influxcrew.com\nRetrouvez-nous sur tous les réseaux LEGEND !\nRetrouvez l'interview complète sur YouTube ➡️ https://youtu.be/ye5cVoc7hIc\nFacebook : https://www.facebook.com/legendmediafr\nInstagram : https://www.instagram.com/legendmedia/\nTikTok : https://www.tiktok.com/@legend\nTwitter : https://twitter.com/legendmediafr\nSnapchat : https://t.snapchat.com/CgEvsbWV\n Hébergé par Acast. Visitez acast.com/privacy pour plus d'informations.",
+                    'release_timestamp': 1753272000,
+                    'duration': 4165,
+                    'url': 'https://podcasts.apple.com/fr/podcast/ambulancier-du-samu-ses-interventions-improbables-suic/id1691740320?i=1000718672235'
+                }
+            ]
+        }]
+
+    # Extract token (supposedly JWT) from javascript
+    # Note: javascript file number/names and token variable name may change
+    def _extract_token(self, webpage):
+        js_urls = re.findall(r'<script[^>]+src=["\'](/assets/[^"\']+\.js)["\']', webpage)
+        js_urls = [urljoin('https://podcasts.apple.com', u) for u in js_urls]
+
+        auth_token = None
+        for js_url in js_urls:
+            js_code = self._download_webpage(js_url, 'Generic authentification token', fatal=False, note=f'Scanning {js_url}')
+            if not js_code:
+                continue
+            match = re.search(r'const\s+Ml\s*=\s*"((?:eyJ)[^"]+)"', js_code)
+            if match:
+                auth_token = match.group(1)
+                break
+
+        if not auth_token:
+            raise ExtractorError('Generic authorization token not found in any JS files')
+
+        return auth_token
+
+    # Call backend API pages and merge them as a single list
+    def _paginate_episodes(self, playlist_id, token):
+        base_url = 'https://amp-api.podcasts.apple.com/v1/catalog/fr/podcasts/'
+        headers = {
+            'Authorization': f'Bearer {token}',
+            'Origin': 'https://podcasts.apple.com'
+        }
+
+        all_episodes = []
+        offset = 0
+        limit = 25 # Limit in use by website but other values seem to be accepted
+
+        while True:
+            episodes_url = (
+                f'{base_url}{playlist_id}/episodes?'
+                f'l=fr-FR&offset={offset}&limit={limit}'
+            )
+            episodes_json = self._download_json(episodes_url, playlist_id, headers=headers, note=f'Downloading episodes offset {offset}')
+            all_episodes.extend(episodes_json.get('data', []))
+            if 'next' not in episodes_json:
+                break
+            offset += limit
+
+        return all_episodes
+
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+        webpage = self._download_webpage(url, playlist_id)
+        server_data = self._search_json(
+            ApplePodcastsBaseIE._BASE_HTML_JSON_LOCATION, webpage,
+            'server data', playlist_id, contains_pattern=ApplePodcastsBaseIE._BASE_HTML_JSON_PATTERN)[0]['data']
+        playlist_data = traverse_obj(server_data,
+                (..., lambda _, v: v.get('contentType') == 'showHeaderRegular', 'items', 0),
+                expected_type=dict, get_all=False)
+
+        token = self._extract_token(webpage)
+        episodes = self._paginate_episodes(playlist_id, token)
+
+        entries = []
+        for e in episodes:
+            episode_data = traverse_obj(e, {
+                'id': ('id', {str}),
+                'title': ('attributes', 'name', {str}),
+                'uploader': ('attributes', 'artistName', {str}),
+                'description': ('attributes', 'description', 'standard', {str}),
+                'url': ('attributes', 'url', {clean_podcast_url}),
+                'release_timestamp': ('attributes', 'releaseDateTime', {parse_iso8601}),
+                'duration': ('attributes', 'durationInMilliseconds', {lambda x: int(x) // 1000}),
+                'thumbnail_template': ('artwork', 'url', {str}),
+                'thumb_width': ('artwork', 'width', {int}),
+                'thumb_height': ('artwork', 'height', {int}),
+            })
+
+            if not episode_data.get('url'):
+                continue
+
+            entries.append({
+                '_type': 'url',
+                'ie_key': 'ApplePodcasts',
+                **episode_data,
+            })
+
+        return self.playlist_result(entries,
+                                    playlist_id,
+                                    **traverse_obj(playlist_data, {
+                                        'playlist_title': ('title', {str}),
+                                        'playlist_description': ('description', {str}),
+                                        'playlist_uploader': ('providerTitle', {str}),
+                                    })
+               )