Merge 9d314fa700 into f2a4983df7

Closes #11526
Authored by: jshumphrey

yt_dlp/extractor/_extractors.py

@@ -1687,6 +1687,7 @@ from .rcti import (
     RCTIPlusTVIE,
 )
 from .rds import RDSIE
+from .rdsit import RDSITIE
 from .redbee import (
     RTBFIE,
     ParliamentLiveUKIE,

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/rdsit.py

@@ -0,0 +1,79 @@
+import re
+
+from .common import InfoExtractor
+
+
+class RDSITIE(InfoExtractor):
+    IE_NAME = 'rds.it'
+    _VALID_URL = r'https?://(?:www\.)?rds\.(?:it|radio)'
+    _TESTS = [{
+        'url': 'https://rds.it',
+        'info_dict': {
+            'id': 'rds',
+            'ext': 'mp4',
+            'title': r're:RDS',
+            'description': 'Radio Dimensione Suono',
+            'is_live': True,
+            'track_id': int,
+            'track': str,
+            'thumbnail': r're:https://web\.rds\.it/',
+            'concurrent_view_count': int,
+            'live_status': 'is_live',
+            'like_count': int,
+        },
+    }]
+    _PATH_CODECS = {
+        'rds': 'mp3',
+        'rds_aac': 'aac',
+        'rds_aac64': 'aac',
+    }
+
+    def _real_extract(self, url):
+        sources = self._download_json('https://icecast.rds.radio/status-json.xsl', 'rds', 'Downloading JSON status')['icestats']['source']
+        current_song = self._download_json('https://cdnapi.rds.it/v2/site/get_player_info', 'rds', 'Downloading JSON player')['song_status']['current_song']
+        formats = self._extract_m3u8_formats('https://stream.rdstv.radio/index.m3u8', 'rds', m3u8_id='hls-rdstv')
+        listeners = 0
+
+        for s in sources:
+            if s['server_name'] == 'RDS':
+                if s.get('listeners'):
+                    listeners += s['listeners']
+                path = re.search(r'[^/]+$', s['listenurl']).group(0)
+                formats.append({
+                    'format_id': f'https-{path}',
+                    'url': f'https://icstream.rds.radio/{path}',
+                    'abr': s['bitrate'],
+                    'acodec': self._PATH_CODECS.get(path),
+                    'ext': self._PATH_CODECS.get(path),
+                    'vcodec': 'none',
+                    'format_note': 'HLS without m3u8',
+                    'language': 'it',
+                    'preference': -10,
+                })
+        # The codecs in the index are wrong
+        mp3_formats = self._extract_m3u8_formats('https://stream.rds.radio/audio/rds.stream/index.m3u8', 'rds', 'mp3', quality=0, m3u8_id='hls-rds')
+        for fmt in mp3_formats:
+            fmt['acodec'] = 'mp3'
+        formats.extend(mp3_formats)
+        aac_formats = self._extract_m3u8_formats('https://stream.rds.radio/audio/rds.stream_aac/index.m3u8', 'rds', 'aac', m3u8_id='hls-rds_aac')
+        for fmt in aac_formats:
+            fmt['acodec'] = 'aac'
+        formats.extend(aac_formats)
+        aac64_formats = self._extract_m3u8_formats('https://stream.rds.radio/audio/rds.stream_aac64/index.m3u8', 'rds', 'aac', m3u8_id='hls-rds_aac')
+        for fmt in aac64_formats:
+            fmt['acodec'] = 'aac'
+        formats.extend(aac64_formats)
+
+        return {
+            'id': 'rds',
+            'title': 'RDS',
+            'description': 'Radio Dimensione Suono',
+            'is_live': True,
+            'formats': formats,
+            'concurrent_view_count': listeners,
+            'thumbnail': current_song.get('cover'),
+            'track': current_song.get('title'),
+            'track_id': current_song.get('id'),
+            'artists': current_song.get('artists').split(', ') if current_song.get('artists') else None,
+            'like_count': current_song.get('likes'),
+        }