Merge f80ba18ee9 into 8e15177b41

Closes #9358
Authored by: jakeogh, minamotorin, shoxie007, bbilly1

Co-authored-by: minamotorin <76122224+minamotorin@users.noreply.github.com>
Co-authored-by: shoxie007 <74592022+shoxie007@users.noreply.github.com>
Co-authored-by: Simon <35427372+bbilly1@users.noreply.github.com>

README.md

@@ -666,7 +666,7 @@ ## Filesystem Options:
                                     The name of the browser to load cookies
                                     from. Currently supported browsers are:
                                     brave, chrome, chromium, edge, firefox,
-                                    opera, safari, vivaldi. Optionally, the
+                                    opera, safari, vivaldi, whale. Optionally, the
                                     KEYRING used for decrypting Chromium cookies
                                     on Linux, the name/path of the PROFILE to
                                     load cookies from, and the CONTAINER name

supportedsites.md

@@ -1449,6 +1449,8 @@ # Supported sites
  - **ThisVid**
  - **ThisVidMember**
  - **ThisVidPlaylist**
+ - **Threads**
+ - **ThreadsIOS**: Threads' iOS `barcelona://` URL
  - **ThreeSpeak**
  - **ThreeSpeakUser**
  - **TikTok**

yt_dlp/cookies.py

@@ -46,7 +46,7 @@
 from .utils._utils import _YDLLogger
 from .utils.networking import normalize_url
 
-CHROMIUM_BASED_BROWSERS = {'brave', 'chrome', 'chromium', 'edge', 'opera', 'vivaldi'}
+CHROMIUM_BASED_BROWSERS = {'brave', 'chrome', 'chromium', 'edge', 'opera', 'vivaldi', 'whale'}
 SUPPORTED_BROWSERS = CHROMIUM_BASED_BROWSERS | {'firefox', 'safari'}
 
 
@@ -219,6 +219,7 @@ def _get_chromium_based_browser_settings(browser_name):
             'edge': os.path.join(appdata_local, R'Microsoft\Edge\User Data'),
             'opera': os.path.join(appdata_roaming, R'Opera Software\Opera Stable'),
             'vivaldi': os.path.join(appdata_local, R'Vivaldi\User Data'),
+            'whale': os.path.join(appdata_local, R'Naver\Naver Whale\User Data'),
         }[browser_name]
 
     elif sys.platform == 'darwin':
@@ -230,6 +231,7 @@ def _get_chromium_based_browser_settings(browser_name):
             'edge': os.path.join(appdata, 'Microsoft Edge'),
             'opera': os.path.join(appdata, 'com.operasoftware.Opera'),
             'vivaldi': os.path.join(appdata, 'Vivaldi'),
+            'whale': os.path.join(appdata, 'Naver/Whale'),
         }[browser_name]
 
     else:
@@ -241,6 +243,7 @@ def _get_chromium_based_browser_settings(browser_name):
             'edge': os.path.join(config, 'microsoft-edge'),
             'opera': os.path.join(config, 'opera'),
             'vivaldi': os.path.join(config, 'vivaldi'),
+            'whale': os.path.join(config, 'naver-whale'),
         }[browser_name]
 
     # Linux keyring names can be determined by snooping on dbus while opening the browser in KDE:
@@ -252,6 +255,7 @@ def _get_chromium_based_browser_settings(browser_name):
         'edge': 'Microsoft Edge' if sys.platform == 'darwin' else 'Chromium',
         'opera': 'Opera' if sys.platform == 'darwin' else 'Chromium',
         'vivaldi': 'Vivaldi' if sys.platform == 'darwin' else 'Chrome',
+        'whale': 'Whale',
     }[browser_name]
 
     browsers_without_profiles = {'opera'}

yt_dlp/extractor/_extractors.py

@@ -1985,6 +1985,10 @@
     ThisVidMemberIE,
     ThisVidPlaylistIE,
 )
+from .threads import (
+    ThreadsIE,
+    ThreadsIOSIE
+)
 from .threespeak import (
     ThreeSpeakIE,
     ThreeSpeakUserIE,

yt_dlp/extractor/common.py

@@ -957,7 +957,8 @@ def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=
         if urlh is False:
             assert not fatal
             return False
-        content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding)
+        content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal,
+                                             encoding=encoding, data=data)
         return (content, urlh)
 
     @staticmethod
@@ -1005,8 +1006,10 @@ def __check_blocked(self, content):
                 'Visit http://blocklist.rkn.gov.ru/ for a block reason.',
                 expected=True)
 
-    def _request_dump_filename(self, url, video_id):
-        basen = f'{video_id}_{url}'
+    def _request_dump_filename(self, url, video_id, data=None):
+        if data is not None:
+            data = hashlib.md5(data).hexdigest()
+        basen = join_nonempty(video_id, data, url, delim='_')
         trim_length = self.get_param('trim_file_name') or 240
         if len(basen) > trim_length:
             h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest()
@@ -1028,16 +1031,18 @@ def __decode_webpage(self, webpage_bytes, encoding, headers):
         except LookupError:
             return webpage_bytes.decode('utf-8', 'replace')
 
-    def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None):
+    def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True,
+                              prefix=None, encoding=None, data=None):
         webpage_bytes = urlh.read()
         if prefix is not None:
             webpage_bytes = prefix + webpage_bytes
+        url_or_request = self._create_request(url_or_request, data)
         if self.get_param('dump_intermediate_pages', False):
             self.to_screen('Dumping request to ' + urlh.url)
             dump = base64.b64encode(webpage_bytes).decode('ascii')
             self._downloader.to_screen(dump)
         if self.get_param('write_pages'):
-            filename = self._request_dump_filename(urlh.url, video_id)
+            filename = self._request_dump_filename(urlh.url, video_id, url_or_request.data)
             self.to_screen(f'Saving request to {filename}')
             with open(filename, 'wb') as outf:
                 outf.write(webpage_bytes)
@@ -1098,7 +1103,7 @@ def download_content(self, url_or_request, video_id, note=note, errnote=errnote,
                              impersonate=None, require_impersonation=False):
             if self.get_param('load_pages'):
                 url_or_request = self._create_request(url_or_request, data, headers, query)
-                filename = self._request_dump_filename(url_or_request.url, video_id)
+                filename = self._request_dump_filename(url_or_request.url, video_id, url_or_request.data)
                 self.to_screen(f'Loading request from {filename}')
                 try:
                     with open(filename, 'rb') as dumpf:

yt_dlp/extractor/threads.py

@@ -0,0 +1,157 @@
+from .common import InfoExtractor
+from ..utils import (
+    strftime_or_none,
+    traverse_obj,
+    remove_end,
+    strip_or_none
+)
+
+
+class ThreadsIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?threads\.net/(?P<uploader>[^/]+)/post/(?P<id>[^/?#&]+)/?(?P<embed>embed.*?)?'
+
+    _TESTS = [{
+        'url': 'https://www.threads.net/@tntsportsbr/post/C6cqebdCfBi',
+        'info_dict': {
+            'id': 'C6cqebdCfBi',
+            'ext': 'mp4',
+            'title': 'md5:062673d04195aa2d99b8d7a11798cb9d',
+            'description': 'md5:fe0c73f9a892fb92efcc67cc075561b0',
+            'uploader': 'TNT Sports Brasil',
+            'uploader_id': 'tntsportsbr',
+            'uploader_url': 'https://www.threads.net/@tntsportsbr',
+            'channel': 'tntsportsbr',
+            'channel_url': 'https://www.threads.net/@tntsportsbr',
+            'timestamp': 1714613811,
+            'upload_date': '20240502',
+            'like_count': int,
+            'channel_is_verified': bool,
+            'thumbnail': r're:^https?://.*\.jpg'
+        }
+    }, {
+        'url': 'https://www.threads.net/@felipebecari/post/C6cM_yNPHCF',
+        'info_dict': {
+            'id': 'C6cM_yNPHCF',
+            'ext': 'mp4',
+            'title': '@felipebecari • Sobre o futuro dos dois últimos resgatados: tem muita notícia boa! 🐶❤️',
+            'description': 'Sobre o futuro dos dois últimos resgatados: tem muita notícia boa! 🐶❤️',
+            'uploader': 'Felipe Becari',
+            'uploader_id': 'felipebecari',
+            'uploader_url': 'https://www.threads.net/@felipebecari',
+            'channel': 'felipebecari',
+            'channel_url': 'https://www.threads.net/@felipebecari',
+            'timestamp': 1714598318,
+            'upload_date': '20240501',
+            'like_count': int,
+            'channel_is_verified': bool,
+            'thumbnail': r're:^https?://.*\.jpg'
+        }
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        metadata = {}
+
+        # Try getting videos from json
+        json_data = self._search_regex(
+            r'<script[^>]+>(.*"code":"%s".*)</script>' % video_id,
+            webpage, 'main json', fatal=True)
+
+        result = self._search_json(
+            r'"result":', json_data,
+            'result data', video_id, fatal=True)
+
+        edges = traverse_obj(result, ('data', 'data', 'edges'))
+
+        for node in edges:
+            items = traverse_obj(node, ('node', 'thread_items'))
+
+            for item in items:
+                post = item.get('post')
+
+                if post and post.get('code') == video_id:
+                    formats = []
+                    thumbnails = []
+
+                    # Videos
+                    if (post.get('carousel_media') is not None):  # Handle multiple videos posts
+                        media_list = post.get('carousel_media')
+                    else:
+                        media_list = [post]
+
+                    for media in media_list:
+                        videos = media.get('video_versions')
+
+                        for video in videos:
+                            formats.append({
+                                'format_id': '%s-%s' % (media.get('pk'), video['type']),  # id-type
+                                'url': video['url'],
+                                'width': media.get('original_width'),
+                                'height': media.get('original_height'),
+                            })
+
+                    # Thumbnails
+                    thumbs = traverse_obj(post, ('image_versions2', 'candidates'))
+
+                    for thumb in thumbs:
+                        thumbnails.append({
+                            'url': thumb['url'],
+                            'width': thumb['width'],
+                            'height': thumb['height'],
+                        })
+
+                    # Metadata
+                    metadata.setdefault('uploader_id', traverse_obj(post, ('user', 'username')))
+                    metadata.setdefault('channel_is_verified', traverse_obj(post, ('user', 'is_verified')))
+                    metadata.setdefault('uploader_url', 'https://www.threads.net/@%s' % traverse_obj(post, ('user', 'username')))
+                    metadata.setdefault('timestamp', post.get('taken_at'))
+                    metadata.setdefault('like_count', post.get('like_count'))
+
+        # Try getting metadata
+        metadata['id'] = video_id
+        metadata['title'] = strip_or_none(remove_end(self._html_extract_title(webpage), '• Threads'))
+        metadata['description'] = self._og_search_description(webpage)
+
+        metadata['channel'] = metadata.get('uploader_id')
+        metadata['channel_url'] = metadata.get('uploader_url')
+        metadata['uploader'] = self._search_regex(r'(.*?) \(', self._og_search_title(webpage), 'uploader')
+        metadata['upload_date'] = strftime_or_none(metadata.get('timestamp'))
+
+        return {
+            **metadata,
+            'formats': formats,
+            'thumbnails': thumbnails
+        }
+
+
+class ThreadsIOSIE(InfoExtractor):
+    IE_DESC = 'IOS barcelona:// URL'
+    _VALID_URL = r'barcelona://media\?shortcode=(?P<id>[^/?#&]+)'
+    _TESTS = [{
+        'url': 'barcelona://media?shortcode=C6fDehepo5D',
+        'info_dict': {
+            'id': 'C6fDehepo5D',
+            'ext': 'mp4',
+            'title': 'md5:dc92f960981b8b3a33eba9681e9fdfc6',
+            'description': 'md5:0c36a7e67e1517459bc0334dba932164',
+            'uploader': 'Sa\u0303o Paulo Futebol Clube',
+            'uploader_id': 'saopaulofc',
+            'uploader_url': 'https://www.threads.net/@saopaulofc',
+            'channel': 'saopaulofc',
+            'channel_url': 'https://www.threads.net/@saopaulofc',
+            'timestamp': 1714694014,
+            'upload_date': '20240502',
+            'like_count': int,
+            'channel_is_verified': bool,
+            'thumbnail': r're:^https?://.*\.jpg'
+        },
+        'add_ie': ['Threads']
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        # Threads doesn't care about the user url, it redirects to the right one
+        # So we use ** instead so that we don't need to find it
+        return self.url_result(f'http://www.threads.net/**/post/{video_id}', ThreadsIE, video_id)

yt_dlp/extractor/youtube.py

@@ -3317,7 +3317,36 @@ def _extract_heatmap(self, data):
                 'value': ('intensityScoreNormalized', {float_or_none}),
             })) or None
 
-    def _extract_comment(self, comment_renderer, parent=None):
+    def _extract_comment(self, entities, parent=None):
+        comment_entity_payload = get_first(entities, ('payload', 'commentEntityPayload', {dict}))
+        if not (comment_id := traverse_obj(comment_entity_payload, ('properties', 'commentId', {str}))):
+            return
+
+        toolbar_entity_payload = get_first(entities, ('payload', 'engagementToolbarStateEntityPayload', {dict}))
+        time_text = traverse_obj(comment_entity_payload, ('properties', 'publishedTime', {str})) or ''
+
+        return {
+            'id': comment_id,
+            'parent': parent or 'root',
+            **traverse_obj(comment_entity_payload, {
+                'text': ('properties', 'content', 'content', {str}),
+                'like_count': ('toolbar', 'likeCountA11y', {parse_count}),
+                'author_id': ('author', 'channelId', {self.ucid_or_none}),
+                'author': ('author', 'displayName', {str}),
+                'author_thumbnail': ('author', 'avatarThumbnailUrl', {url_or_none}),
+                'author_is_uploader': ('author', 'isCreator', {bool}),
+                'author_is_verified': ('author', 'isVerified', {bool}),
+                'author_url': ('author', 'channelCommand', 'innertubeCommand', (
+                    ('browseEndpoint', 'canonicalBaseUrl'), ('commandMetadata', 'webCommandMetadata', 'url')
+                ), {lambda x: urljoin('https://www.youtube.com', x)}),
+            }, get_all=False),
+            'is_favorited': (None if toolbar_entity_payload is None else
+                             toolbar_entity_payload.get('heartState') == 'TOOLBAR_HEART_STATE_HEARTED'),
+            '_time_text': time_text,  # FIXME: non-standard, but we need a way of showing that it is an estimate.
+            'timestamp': self._parse_time_text(time_text),
+        }
+
+    def _extract_comment_old(self, comment_renderer, parent=None):
         comment_id = comment_renderer.get('commentId')
         if not comment_id:
             return
@@ -3398,21 +3427,39 @@ def extract_header(contents):
                 break
             return _continuation
 
-        def extract_thread(contents):
+        def extract_thread(contents, entity_payloads):
             if not parent:
                 tracker['current_page_thread'] = 0
             for content in contents:
                 if not parent and tracker['total_parent_comments'] >= max_parents:
                     yield
                 comment_thread_renderer = try_get(content, lambda x: x['commentThreadRenderer'])
-                comment_renderer = get_first(
-                    (comment_thread_renderer, content), [['commentRenderer', ('comment', 'commentRenderer')]],
-                    expected_type=dict, default={})
 
-                comment = self._extract_comment(comment_renderer, parent)
+                # old comment format
+                if not entity_payloads:
+                    comment_renderer = get_first(
+                        (comment_thread_renderer, content), [['commentRenderer', ('comment', 'commentRenderer')]],
+                        expected_type=dict, default={})
+
+                    comment = self._extract_comment_old(comment_renderer, parent)
+
+                # new comment format
+                else:
+                    view_model = (
+                        traverse_obj(comment_thread_renderer, ('commentViewModel', 'commentViewModel', {dict}))
+                        or traverse_obj(content, ('commentViewModel', {dict})))
+                    comment_keys = traverse_obj(view_model, (('commentKey', 'toolbarStateKey'), {str}))
+                    if not comment_keys:
+                        continue
+                    entities = traverse_obj(entity_payloads, lambda _, v: v['entityKey'] in comment_keys)
+                    comment = self._extract_comment(entities, parent)
+                    if comment:
+                        comment['is_pinned'] = traverse_obj(view_model, ('pinnedText', {str})) is not None
+
                 if not comment:
                     continue
                 comment_id = comment['id']
+
                 if comment.get('is_pinned'):
                     tracker['pinned_comment_ids'].add(comment_id)
                 # Sometimes YouTube may break and give us infinite looping comments.
@@ -3505,7 +3552,7 @@ def extract_thread(contents):
             check_get_keys = None
             if not is_forced_continuation and not (tracker['est_total'] == 0 and tracker['running_total'] == 0):
                 check_get_keys = [[*continuation_items_path, ..., (
-                    'commentsHeaderRenderer' if is_first_continuation else ('commentThreadRenderer', 'commentRenderer'))]]
+                    'commentsHeaderRenderer' if is_first_continuation else ('commentThreadRenderer', 'commentViewModel', 'commentRenderer'))]]
             try:
                 response = self._extract_response(
                     item_id=None, query=continuation,
@@ -3529,6 +3576,7 @@ def extract_thread(contents):
                 raise
             is_forced_continuation = False
             continuation = None
+            mutations = traverse_obj(response, ('frameworkUpdates', 'entityBatchUpdate', 'mutations', ..., {dict}))
             for continuation_items in traverse_obj(response, continuation_items_path, expected_type=list, default=[]):
                 if is_first_continuation:
                     continuation = extract_header(continuation_items)
@@ -3537,7 +3585,7 @@ def extract_thread(contents):
                         break
                     continue
 
-                for entry in extract_thread(continuation_items):
+                for entry in extract_thread(continuation_items, mutations):
                     if not entry:
                         return
                     yield entry