Merge 7ceb05cfcc into f2a4983df7

Closes #11526
Authored by: jshumphrey

yt_dlp/extractor/_extractors.py

@@ -1448,6 +1448,7 @@ from .onet import (
     OnetPlIE,
 )
 from .onionstudios import OnionStudiosIE
+from .onsen import OnsenIE
 from .opencast import (
     OpencastIE,
     OpencastPlaylistIE,

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/chaturbate.py

@@ -9,7 +9,7 @@ from ..utils import (
 
 
 class ChaturbateIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:[^/]+\.)?chaturbate\.com/(?:fullvideo/?\?.*?\bb=)?(?P<id>[^/?&#]+)'
+    _VALID_URL = r'https?://(?:[^/]+\.)?chaturbate\.(?P<tld>com|eu|global)/(?:fullvideo/?\?.*?\bb=)?(?P<id>[^/?&#]+)'
     _TESTS = [{
         'url': 'https://www.chaturbate.com/siswet19/',
         'info_dict': {
@@ -29,15 +29,24 @@ class ChaturbateIE(InfoExtractor):
     }, {
         'url': 'https://en.chaturbate.com/siswet19/',
         'only_matching': True,
+    }, {
+        'url': 'https://chaturbate.eu/siswet19/',
+        'only_matching': True,
+    }, {
+        'url': 'https://chaturbate.eu/fullvideo/?b=caylin',
+        'only_matching': True,
+    }, {
+        'url': 'https://chaturbate.global/siswet19/',
+        'only_matching': True,
     }]
 
     _ROOM_OFFLINE = 'Room is currently offline'
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        video_id, tld = self._match_valid_url(url).group('id', 'tld')
 
         webpage = self._download_webpage(
-            f'https://chaturbate.com/{video_id}/', video_id,
+            f'https://chaturbate.{tld}/{video_id}/', video_id,
             headers=self.geo_verification_headers())
 
         found_m3u8_urls = []

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/onsen.py

@@ -0,0 +1,138 @@
+import base64
+import json
+
+from .common import InfoExtractor
+from ..utils import (
+    ExtractorError,
+    parse_qs,
+    str_or_none,
+    strip_or_none,
+)
+from ..utils.traversal import traverse_obj
+
+
+class OnsenIE(InfoExtractor):
+    IE_NAME = 'onsen'
+    IE_DESC = '音泉'
+
+    _BASE_URL = 'https://www.onsen.ag/'
+    _HEADERS = {'Referer': _BASE_URL}
+    _NETRC_MACHINE = 'onsen'
+    _VALID_URL = r'https?://(?:(?:share|www)\.)onsen\.ag/program/(?P<id>[\w-]+)'
+    _TESTS = [{
+        'url': 'https://share.onsen.ag/program/tricolor?p=393&c=MTk2NjE',
+        'info_dict': {
+            'id': '19661',
+            'title': '第0回',
+            'cast': ['礒部花凜', '土屋李央', '林鼓子'],
+            'ext': 'm4a',
+            'description': 'md5:8435d68dcb7a43bc2c993911b0db245b',
+            'display_id': 'MTk2NjE=',
+            'http_headers': {'Referer': 'https://www.onsen.ag/'},
+            'media_type': 'sound',
+            'tags': ['かりこ'],
+            'thumbnail': 'https://d3bzklg4lms4gh.cloudfront.net/program_info/image/default/production/31/ea/c1db117c9b41655120d3a212b2038d15811f/image',
+            'section_start': 0,
+            'series': '礒部花凜・土屋李央・林鼓子 トリコロールカラー',
+            'series_id': 'tricolor',
+            'upload_date': '20240907',
+            'webpage_url': 'https://www.onsen.ag/program/tricolor?c=MTk2NjE=',
+        },
+    }, {
+        'url': 'https://share.onsen.ag/program/girls-band-cry-radio?p=370&c=MTgwMDE',
+        'info_dict': {
+            'id': '18001',
+            'title': '第4回',
+            'cast': ['夕莉', '理名', '朱李', '凪都', '美怜'],
+            'ext': 'mp4',
+            'description': 'md5:1d7f6a2f1f5a3e2a8ada4e9f652262dd',
+            'display_id': 'MTgwMDE=',
+            'http_headers': {'Referer': 'https://www.onsen.ag/'},
+            'media_type': 'movie',
+            'tags': ['ガールズバンドクライ', 'ガルクラ', 'ガルクラジオ'],
+            'thumbnail': 'https://d3bzklg4lms4gh.cloudfront.net/program_info/image/default/production/95/a7/6a848c87bebf3ec085d8890f3ce038f9b4dd/image',
+            'section_start': 0,
+            'series': 'TVアニメ『ガールズバンドクライ』WEBラジオ「ガールズバンドクライ～ラジオにも全部ぶち込め。～」',
+            'series_id': 'girls-band-cry-radio',
+            'upload_date': '20240425',
+            'webpage_url': 'https://www.onsen.ag/program/girls-band-cry-radio?c=MTgwMDE=',
+        },
+        'skip': 'Only available for premium supporters',
+    }, {
+        'url': 'https://www.onsen.ag/program/g-witch',
+        'info_dict': {
+            'id': 'g-witch',
+            'title': '機動戦士ガンダム 水星の魔女～アスティカシア高等専門学園 ラジオ委員会～',
+        },
+        'playlist_mincount': 7,
+    }]
+
+    def _perform_login(self, username, password):
+        signin = self._download_json(
+            f'{self._BASE_URL}web_api/signin', None, 'Logging in', headers={
+                'content-type': 'application/json; charset=UTF-8',
+            }, data=json.dumps({
+                'session': {
+                    'email': username,
+                    'password': password,
+                },
+            }).encode(), expected_status=401)
+
+        if signin.get('error'):
+            raise ExtractorError('Invalid username or password', expected=True)
+
+    def _get_info(self, program, program_id, metadata):
+        m3u8 = program['streaming_url']
+        rd = self._search_regex(f'{program_id}(\\d{{6}})', m3u8, 'release_date', default=None)
+        display_id = base64.b64encode(str(program['id']).encode()).decode()
+
+        return {
+            'display_id': display_id,
+            'formats': self._extract_m3u8_formats(m3u8, program_id, headers=self._HEADERS),
+            'http_headers': self._HEADERS,
+            'upload_date': f'20{rd}' if rd else None,
+            'webpage_url': f'{self._BASE_URL}program/{program_id}?c={display_id}',
+            **metadata,
+            **traverse_obj(program, {
+                'id': ('id', {str_or_none}),
+                'title': ('title', {strip_or_none}),
+                'thumbnail': ('poster_image_url', {lambda x: x.split('?')[0]}),
+                'media_type': ('media_type', {str}),
+            }),
+            'cast': metadata['cast'] + traverse_obj(program, ('guests', ..., 'name', {str})),
+        }
+
+    def _real_extract(self, url):
+        program_id = self._match_id(url)
+        qs = {k: v[0] for k, v in parse_qs(url).items() if v}
+        programs = self._download_json(
+            f'{self._BASE_URL}web_api/programs/{program_id}', program_id)
+
+        metadata = {
+            'cast': traverse_obj(programs, ('performers', ..., 'name', {str})),
+            'section_start': int(qs.get('t', 0)),
+            'series_id': program_id,
+            **traverse_obj(programs['program_info'], {
+                'description': ('description', {str}),
+                'series': ('title', {str}),
+                'tags': ('hashtag_list', {list}),
+            }),
+        }
+
+        if 'c' in qs:
+            p_id = base64.b64decode(qs['c'] + '=' * (-len(qs['c']) % 4)).decode()
+            program = traverse_obj(
+                programs, ('contents', lambda _, v: v['id'] == int(p_id)), get_all=False)
+            if not program:
+                raise ExtractorError('This program is no longer available', expected=True)
+            if not program['streaming_url']:
+                self.raise_login_required('This program is only available for premium supporters')
+
+            return self._get_info(program, program_id, metadata)
+        else:
+            entries = [
+                self._get_info(program, program_id, metadata)
+                for program in programs['contents'] if program['streaming_url']
+            ]
+
+            return self.playlist_result(entries, program_id, metadata['series'])