Merge ac410f1354 into f2a4983df7

Closes #11526
Authored by: jshumphrey

yt_dlp/extractor/_extractors.py

@@ -118,7 +118,6 @@ from .amazon import (
 from .amazonminitv import (
     AmazonMiniTVIE,
     AmazonMiniTVSeasonIE,
-    AmazonMiniTVSeriesIE,
 )
 from .amcnetworks import AMCNetworksIE
 from .americastestkitchen import (

yt_dlp/extractor/amazonminitv.py

@@ -1,5 +1,3 @@
-import json
-
 from .common import InfoExtractor
 from ..utils import ExtractorError, int_or_none, traverse_obj, try_get
 
@@ -9,35 +7,28 @@ class AmazonMiniTVBaseIE(InfoExtractor):
         self._download_webpage(
             'https://www.amazon.in/minitv', None,
             note='Fetching guest session cookies')
-        AmazonMiniTVBaseIE.session_id = self._get_cookies('https://www.amazon.in')['session-id'].value
+        AmazonMiniTVBaseIE.urtk = self._get_cookies('https://www.amazon.in')['urtk'].value
 
     def _call_api(self, asin, data=None, note=None):
-        device = {'clientId': 'ATVIN', 'deviceLocale': 'en_GB'}
+        query = {
+            'contentId': asin,
+        }
         if data:
-            data['variables'].update({
+            query.update(data)
-                'contentType': 'VOD',
-                'sessionIdToken': self.session_id,
-                **device,
-            })
 
         resp = self._download_json(
-            f'https://www.amazon.in/minitv/api/web/{"graphql" if data else "prs"}',
+            'https://www.amazon.in/minitv-pr/api/web/page/title',
             asin, note=note, headers={
                 'Content-Type': 'application/json',
+                'accounttype': 'NEW_GUEST_ACCOUNT',
                 'currentpageurl': '/',
                 'currentplatform': 'dWeb',
-            }, data=json.dumps(data).encode() if data else None,
+            }, data=None,
-            query=None if data else {
+            query=query)
-                'deviceType': 'A1WMMUXPCUJL4N',
-                'contentId': asin,
-                **device,
-            })
 
         if resp.get('errors'):
             raise ExtractorError(f'MiniTV said: {resp["errors"][0]["message"]}')
-        elif not data:
         return resp
-        return resp['data'][data['operationName']]
 
 
 class AmazonMiniTVIE(AmazonMiniTVBaseIE):
@@ -89,104 +80,46 @@ class AmazonMiniTVIE(AmazonMiniTVBaseIE):
         'only_matching': True,
     }]
 
-    _GRAPHQL_QUERY_CONTENT = '''
-query content($sessionIdToken: String!, $deviceLocale: String, $contentId: ID!, $contentType: ContentType!, $clientId: String) {
-  content(
-    applicationContextInput: {deviceLocale: $deviceLocale, sessionIdToken: $sessionIdToken, clientId: $clientId}
-    contentId: $contentId
-    contentType: $contentType
-  ) {
-    contentId
-    name
-    ... on Episode {
-      contentId
-      vodType
-      name
-      images
-      description {
-        synopsis
-        contentLengthInSeconds
-      }
-      publicReleaseDateUTC
-      audioTracks
-      seasonId
-      seriesId
-      seriesName
-      seasonNumber
-      episodeNumber
-      timecode {
-        endCreditsTime
-      }
-    }
-    ... on MovieContent {
-      contentId
-      vodType
-      name
-      description {
-        synopsis
-        contentLengthInSeconds
-      }
-      images
-      publicReleaseDateUTC
-      audioTracks
-    }
-  }
-}'''
-
     def _real_extract(self, url):
         asin = f'amzn1.dv.gti.{self._match_id(url)}'
         prs = self._call_api(asin, note='Downloading playback info')
+        playback_info = traverse_obj(prs, ('widgets', 0, 'data', 'playbackAssets', 'manifestData'))
+        title_info = traverse_obj(prs, ('widgets', 0, 'data', 'contentDetails'))
+        title_info_ = traverse_obj(prs, ('metaData', 'contentDetails'))
 
         formats, subtitles = [], {}
-        for type_, asset in prs['playbackAssets'].items():
+        for mpd in playback_info:
-            if not traverse_obj(asset, 'manifestUrl'):
-                continue
-            if type_ == 'hls':
-                m3u8_fmts, m3u8_subs = self._extract_m3u8_formats_and_subtitles(
-                    asset['manifestUrl'], asin, ext='mp4', entry_protocol='m3u8_native',
-                    m3u8_id=type_, fatal=False)
-                formats.extend(m3u8_fmts)
-                subtitles = self._merge_subtitles(subtitles, m3u8_subs)
-            elif type_ == 'dash':
             mpd_fmts, mpd_subs = self._extract_mpd_formats_and_subtitles(
-                    asset['manifestUrl'], asin, mpd_id=type_, fatal=False)
+                mpd['manifestURL'], asin, mpd_id=mpd['codec'], fatal=False)
             formats.extend(mpd_fmts)
             subtitles = self._merge_subtitles(subtitles, mpd_subs)
-            else:
-                self.report_warning(f'Unknown asset type: {type_}')
 
-        title_info = self._call_api(
+        credits_time = try_get(title_info, lambda x: x['skipData']['INTRO']['endTime'])
-            asin, note='Downloading title info', data={
+        is_episode = title_info_.get('vodType') == 'EPISODE'
-                'operationName': 'content',
-                'variables': {'contentId': asin},
-                'query': self._GRAPHQL_QUERY_CONTENT,
-            })
-        credits_time = try_get(title_info, lambda x: x['timecode']['endCreditsTime'] / 1000)
-        is_episode = title_info.get('vodType') == 'EPISODE'
 
         return {
             'id': asin,
-            'title': title_info.get('name'),
+            'title': title_info_.get('name'),
             'formats': formats,
             'subtitles': subtitles,
             'language': traverse_obj(title_info, ('audioTracks', 0)),
             'thumbnails': [{
-                'id': type_,
+                'id': 'imageSrc',
-                'url': url,
+                'url': title_info_.get('imageSrc'),
-            } for type_, url in (title_info.get('images') or {}).items()],
+            }] if title_info_.get('imageSrc') else [],
-            'description': traverse_obj(title_info, ('description', 'synopsis')),
+            'description': traverse_obj(title_info_, ('synopsis')),
-            'release_timestamp': int_or_none(try_get(title_info, lambda x: x['publicReleaseDateUTC'] / 1000)),
+            'release_timestamp': int_or_none(try_get(title_info_, lambda x: x['publicReleaseDateUTC'] / 1000)),
-            'duration': traverse_obj(title_info, ('description', 'contentLengthInSeconds')),
+            'duration': traverse_obj(title_info_, ('contentLengthInSeconds')),
             'chapters': [{
                 'start_time': credits_time,
                 'title': 'End Credits',
             }] if credits_time else [],
-            'series': title_info.get('seriesName'),
+            'series': title_info_.get('seasonName') if is_episode else None,
-            'series_id': title_info.get('seriesId'),
+            'series_id': title_info.get('seriesId') if is_episode else None,
-            'season_number': title_info.get('seasonNumber'),
+            'season_number': title_info.get('seasonNumber') if is_episode else None,
-            'season_id': title_info.get('seasonId'),
+            'season_id': title_info.get('seasonId') if is_episode else None,
             'episode': title_info.get('name') if is_episode else None,
-            'episode_number': title_info.get('episodeNumber'),
+            'episode_number': title_info.get('episodeNumber') if is_episode else None,
             'episode_id': asin if is_episode else None,
         }
 
@@ -206,88 +139,17 @@ class AmazonMiniTVSeasonIE(AmazonMiniTVBaseIE):
         'only_matching': True,
     }]
 
-    _GRAPHQL_QUERY = '''
-query getEpisodes($sessionIdToken: String!, $clientId: String, $episodeOrSeasonId: ID!, $deviceLocale: String) {
-  getEpisodes(
-    applicationContextInput: {sessionIdToken: $sessionIdToken, deviceLocale: $deviceLocale, clientId: $clientId}
-    episodeOrSeasonId: $episodeOrSeasonId
-  ) {
-    episodes {
-      ... on Episode {
-        contentId
-        name
-        images
-        seriesName
-        seasonId
-        seriesId
-        seasonNumber
-        episodeNumber
-        description {
-          synopsis
-          contentLengthInSeconds
-        }
-        publicReleaseDateUTC
-      }
-    }
-  }
-}
-'''
-
     def _entries(self, asin):
         season_info = self._call_api(
-            asin, note='Downloading season info', data={
+            asin, note='Downloading season info',
-                'operationName': 'getEpisodes',
+            data={'cursor': '8e0cefec-e190-46ba-854d-1f3ca7978b4a:::'},
-                'variables': {'episodeOrSeasonId': asin},
+        )
-                'query': self._GRAPHQL_QUERY,
-            })
 
-        for episode in season_info['episodes']:
+        for season in season_info['widgets'][0]['data']['options']:
+            if season['active']:
+                for episode in season['value']['data']['widgets'][0]['data']['widgets']:
                     yield self.url_result(
-                f'amazonminitv:{episode["contentId"]}', AmazonMiniTVIE, episode['contentId'])
+                        f'amazonminitv:{episode["data"]["contentId"]}', AmazonMiniTVIE, episode['data']['contentId'])
-
-    def _real_extract(self, url):
-        asin = f'amzn1.dv.gti.{self._match_id(url)}'
-        return self.playlist_result(self._entries(asin), asin)
-
-
-class AmazonMiniTVSeriesIE(AmazonMiniTVBaseIE):
-    IE_NAME = 'amazonminitv:series'
-    _VALID_URL = r'amazonminitv:series:(?:amzn1\.dv\.gti\.)?(?P<id>[a-f0-9-]+)'
-    IE_DESC = 'Amazon MiniTV Series, "minitv:series:" prefix'
-    _TESTS = [{
-        'url': 'amazonminitv:series:amzn1.dv.gti.56521d46-b040-4fd5-872e-3e70476a04b0',
-        'playlist_mincount': 3,
-        'info_dict': {
-            'id': 'amzn1.dv.gti.56521d46-b040-4fd5-872e-3e70476a04b0',
-        },
-    }, {
-        'url': 'amazonminitv:series:56521d46-b040-4fd5-872e-3e70476a04b0',
-        'only_matching': True,
-    }]
-
-    _GRAPHQL_QUERY = '''
-query getSeasons($sessionIdToken: String!, $deviceLocale: String, $episodeOrSeasonOrSeriesId: ID!, $clientId: String) {
-  getSeasons(
-    applicationContextInput: {deviceLocale: $deviceLocale, sessionIdToken: $sessionIdToken, clientId: $clientId}
-    episodeOrSeasonOrSeriesId: $episodeOrSeasonOrSeriesId
-  ) {
-    seasons {
-      seasonId
-    }
-  }
-}
-'''
-
-    def _entries(self, asin):
-        season_info = self._call_api(
-            asin, note='Downloading series info', data={
-                'operationName': 'getSeasons',
-                'variables': {'episodeOrSeasonOrSeriesId': asin},
-                'query': self._GRAPHQL_QUERY,
-            })
-
-        for season in season_info['seasons']:
-            yield self.url_result(f'amazonminitv:season:{season["seasonId"]}', AmazonMiniTVSeasonIE, season['seasonId'])
 
     def _real_extract(self, url):
         asin = f'amzn1.dv.gti.{self._match_id(url)}'

yt_dlp/extractor/archiveorg.py

@@ -205,6 +205,26 @@ class ArchiveOrgIE(InfoExtractor):
                 },
             },
         ],
+    }, {
+        # The reviewbody is None for one of the reviews; just need to extract data without crashing
+        'url': 'https://archive.org/details/gd95-04-02.sbd.11622.sbeok.shnf/gd95-04-02d1t04.shn',
+        'info_dict': {
+            'id': 'gd95-04-02.sbd.11622.sbeok.shnf/gd95-04-02d1t04.shn',
+            'ext': 'mp3',
+            'title': 'Stuck Inside of Mobile with the Memphis Blues Again',
+            'creators': ['Grateful Dead'],
+            'duration': 338.31,
+            'track': 'Stuck Inside of Mobile with the Memphis Blues Again',
+            'description': 'md5:764348a470b986f1217ffd38d6ac7b72',
+            'display_id': 'gd95-04-02d1t04.shn',
+            'location': 'Pyramid Arena',
+            'uploader': 'jon@archive.org',
+            'album': '1995-04-02 - Pyramid Arena',
+            'upload_date': '20040519',
+            'track_number': 4,
+            'release_date': '19950402',
+            'timestamp': 1084927901,
+        },
     }]
 
     @staticmethod
@@ -335,7 +355,7 @@ class ArchiveOrgIE(InfoExtractor):
                 info['comments'].append({
                     'id': review.get('review_id'),
                     'author': review.get('reviewer'),
-                    'text': str_or_none(review.get('reviewtitle'), '') + '\n\n' + review.get('reviewbody'),
+                    'text': join_nonempty('reviewtitle', 'reviewbody', from_dict=review, delim='\n\n'),
                     'timestamp': unified_timestamp(review.get('createdate')),
                     'parent': 'root'})
 

yt_dlp/extractor/facebook.py

@@ -563,13 +563,13 @@ class FacebookIE(InfoExtractor):
                 return extract_video_data(try_get(
                     js_data, lambda x: x['jsmods']['instances'], list) or [])
 
-        def extract_dash_manifest(video, formats):
+        def extract_dash_manifest(vid_data, formats, mpd_url=None):
             dash_manifest = traverse_obj(
-                video, 'dash_manifest', 'playlist', 'dash_manifest_xml_string', expected_type=str)
+                vid_data, 'dash_manifest', 'playlist', 'dash_manifest_xml_string', 'manifest_xml', expected_type=str)
             if dash_manifest:
                 formats.extend(self._parse_mpd_formats(
                     compat_etree_fromstring(urllib.parse.unquote_plus(dash_manifest)),
-                    mpd_url=url_or_none(video.get('dash_manifest_url'))))
+                    mpd_url=url_or_none(video.get('dash_manifest_url')) or mpd_url))
 
         def process_formats(info):
             # Downloads with browser's User-Agent are rate limited. Working around
@@ -619,9 +619,12 @@ class FacebookIE(InfoExtractor):
                         video = video['creation_story']
                         video['owner'] = traverse_obj(video, ('short_form_video_context', 'video_owner'))
                         video.update(reel_info)
-                    fmt_data = traverse_obj(video, ('videoDeliveryLegacyFields', {dict})) or video
+
                     formats = []
                     q = qualities(['sd', 'hd'])
+
+                    # Legacy formats extraction
+                    fmt_data = traverse_obj(video, ('videoDeliveryLegacyFields', {dict})) or video
                     for key, format_id in (('playable_url', 'sd'), ('playable_url_quality_hd', 'hd'),
                                            ('playable_url_dash', ''), ('browser_native_hd_url', 'hd'),
                                            ('browser_native_sd_url', 'sd')):
@@ -629,7 +632,7 @@ class FacebookIE(InfoExtractor):
                         if not playable_url:
                             continue
                         if determine_ext(playable_url) == 'mpd':
-                            formats.extend(self._extract_mpd_formats(playable_url, video_id))
+                            formats.extend(self._extract_mpd_formats(playable_url, video_id, fatal=False))
                         else:
                             formats.append({
                                 'format_id': format_id,
@@ -638,6 +641,28 @@ class FacebookIE(InfoExtractor):
                                 'url': playable_url,
                             })
                     extract_dash_manifest(fmt_data, formats)
+
+                    # New videoDeliveryResponse formats extraction
+                    fmt_data = traverse_obj(video, ('videoDeliveryResponseFragment', 'videoDeliveryResponseResult'))
+                    mpd_urls = traverse_obj(fmt_data, ('dash_manifest_urls', ..., 'manifest_url', {url_or_none}))
+                    dash_manifests = traverse_obj(fmt_data, ('dash_manifests', lambda _, v: v['manifest_xml']))
+                    for idx, dash_manifest in enumerate(dash_manifests):
+                        extract_dash_manifest(dash_manifest, formats, mpd_url=traverse_obj(mpd_urls, idx))
+                    if not dash_manifests:
+                        # Only extract from MPD URLs if the manifests are not already provided
+                        for mpd_url in mpd_urls:
+                            formats.extend(self._extract_mpd_formats(mpd_url, video_id, fatal=False))
+                    for prog_fmt in traverse_obj(fmt_data, ('progressive_urls', lambda _, v: v['progressive_url'])):
+                        format_id = traverse_obj(prog_fmt, ('metadata', 'quality', {str.lower}))
+                        formats.append({
+                            'format_id': format_id,
+                            # sd, hd formats w/o resolution info should be deprioritized below DASH
+                            'quality': q(format_id) - 3,
+                            'url': prog_fmt['progressive_url'],
+                        })
+                    for m3u8_url in traverse_obj(fmt_data, ('hls_playlist_urls', ..., 'hls_playlist_url', {url_or_none})):
+                        formats.extend(self._extract_m3u8_formats(m3u8_url, video_id, 'mp4', fatal=False, m3u8_id='hls'))
+
                     if not formats:
                         # Do not append false positive entry w/o any formats
                         return