Merge a111b3ab58 into f2a4983df7

Closes #11526
Authored by: jshumphrey

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

yt_dlp/extractor/generic.py

@@ -137,10 +137,13 @@ class GenericIE(InfoExtractor):
             'playlist': [{
                 'info_dict': {
                     'ext': 'mov',
-                    'id': 'pdv_maddow_netcast_mov-12-03-2020-223726',
-                    'title': 'MSNBC Rachel Maddow (video) - 12-03-2020-223726',
+                    'id': 'https://nbcnewsencoding.akamaized.net/nbcnews/video/podcast/pdv_maddow_netcast_mov-11-18-2022-224832.mov',
+                    'title': 'MSNBC Rachel Maddow (video) - 11-18-2022-224832',
                     'description': 're:.*her unique approach to storytelling.*',
-                    'upload_date': '20201204',
+                    'upload_date': '20221119',
+                    'direct': True,
+                    'duration': 1152.0,
+                    'timestamp': 1668826112.0,
                 },
             }],
             'skip': 'Dead link',
@@ -183,7 +186,27 @@ class GenericIE(InfoExtractor):
                 'description': 'CGP Grey and Brady Haran talk about YouTube, life, work, whatever.',
                 'title': 'Hello Internet',
             },
+            'playlist': [{
+                'info_dict': {
+                    'ext': 'mp3',
+                    'id': '52d66949e4b0a8cec3bcdd46:52d67282e4b0cca8969714fa:5e58de8a37459e0d069efda0',
+                    'title': 'H.I. #136: Dog Bingo',
+                    'description': 'md5:8e5d0bc18204a1451d5e32937ecf7d6d',
+                    'thumbnail': 'https://images.squarespace-cdn.com/content/v1/52d66949e4b0a8cec3bcdd46/1582882549057-1XW68A67UR858INW8IW5/HI+slides.002.png?format=1500w',
+                    'timestamp': 1582884807.0,
+                    'episode': 'H.I. #136: Dog Bingo',
+                    'episode_number': 136,
+                    'age_limit': 0,
+                    'duration': 4914.0,
+                    'upload_date': '20200228',
+                    'direct': True,
+                    'uploader': 'Hello Internet',
+                },
+            }],
             'playlist_mincount': 100,
+            'params': {
+                'skip_download': True,
+            },
         },
         # RSS feed with guid
         {
@@ -2211,6 +2234,7 @@ class GenericIE(InfoExtractor):
     def _extract_rss(self, url, video_id, doc):
         NS_MAP = {
             'itunes': 'http://www.itunes.com/dtds/podcast-1.0.dtd',
+            'dc': 'http://dublincore.org/specifications/dublin-core/dcmes-xml/2001-04-11/dcmes-xml-dtd.dtd',
         }
 
         entries = []
@@ -2222,17 +2246,20 @@ class GenericIE(InfoExtractor):
                 continue
 
             guid = try_call(lambda: it.find('guid').text)
-            if guid:
-                next_url = smuggle_url(next_url, {'force_videoid': guid})
+            next_url_new = smuggle_url(next_url, {'force_videoid': guid}) if guid else next_url
 
             def itunes(key):
                 return xpath_text(it, xpath_with_ns(f'./itunes:{key}', NS_MAP), default=None)
 
+            def dc(key):
+                return xpath_text(it, xpath_with_ns(f'./dc:{key}', NS_MAP), default=None)
+
             entries.append({
                 '_type': 'url_transparent',
-                'url': next_url,
+                'url': next_url_new,
                 'title': try_call(lambda: it.find('title').text),
                 'description': xpath_text(it, 'description', default=None),
+                'uploader': xpath_text(it, 'author', default=None) or itunes('author') or dc('creator'),
                 'timestamp': unified_timestamp(xpath_text(it, 'pubDate', default=None)),
                 'duration': parse_duration(itunes('duration')),
                 'thumbnail': url_or_none(xpath_attr(it, xpath_with_ns('./itunes:image', NS_MAP), 'href')),
@@ -2240,6 +2267,7 @@ class GenericIE(InfoExtractor):
                 'episode_number': int_or_none(itunes('episode')),
                 'season_number': int_or_none(itunes('season')),
                 'age_limit': {'true': 18, 'yes': 18, 'false': 0, 'no': 0}.get((itunes('explicit') or '').lower()),
+                'webpage_url': next_url,
             })
 
         return {