if not comment_id:
return
- text = self._get_text(comment_renderer, 'contentText')
+ info = {
+ 'id': comment_id,
+ 'text': self._get_text(comment_renderer, 'contentText'),
+ 'like_count': self._get_count(comment_renderer, 'voteCount'),
+ 'author_id': traverse_obj(comment_renderer, ('authorEndpoint', 'browseEndpoint', 'browseId', {self.ucid_or_none})),
+ 'author': self._get_text(comment_renderer, 'authorText'),
+ 'author_thumbnail': traverse_obj(comment_renderer, ('authorThumbnail', 'thumbnails', -1, 'url', {url_or_none})),
+ 'parent': parent or 'root',
+ }
# Timestamp is an estimate calculated from the current time and time_text
time_text = self._get_text(comment_renderer, 'publishedTimeText') or ''
timestamp = self._parse_time_text(time_text)
- author = self._get_text(comment_renderer, 'authorText')
- author_id = try_get(comment_renderer,
- lambda x: x['authorEndpoint']['browseEndpoint']['browseId'], str)
+ info.update({
+ # FIXME: non-standard, but we need a way of showing that it is an estimate.
+ '_time_text': time_text,
+ 'timestamp': timestamp,
+ })
- votes = parse_count(try_get(comment_renderer, (lambda x: x['voteCount']['simpleText'],
- lambda x: x['likeCount']), str)) or 0
- author_thumbnail = try_get(comment_renderer,
- lambda x: x['authorThumbnail']['thumbnails'][-1]['url'], str)
+ info['author_url'] = urljoin(
+ 'https://www.youtube.com', traverse_obj(comment_renderer, ('authorEndpoint', (
+ ('browseEndpoint', 'canonicalBaseUrl'), ('commandMetadata', 'webCommandMetadata', 'url'))),
+ expected_type=str, get_all=False))
- author_is_uploader = try_get(comment_renderer, lambda x: x['authorIsChannelOwner'], bool)
- is_favorited = 'creatorHeart' in (try_get(
- comment_renderer, lambda x: x['actionButtons']['commentActionButtonsRenderer'], dict) or {})
- return {
- 'id': comment_id,
- 'text': text,
- 'timestamp': timestamp,
- 'time_text': time_text,
- 'like_count': votes,
- 'is_favorited': is_favorited,
- 'author': author,
- 'author_id': author_id,
- 'author_thumbnail': author_thumbnail,
- 'author_is_uploader': author_is_uploader,
- 'parent': parent or 'root'
- }
+ author_is_uploader = traverse_obj(comment_renderer, 'authorIsChannelOwner')
+ if author_is_uploader is not None:
+ info['author_is_uploader'] = author_is_uploader
+
+ comment_abr = traverse_obj(
+ comment_renderer, ('actionsButtons', 'commentActionButtonsRenderer'), expected_type=dict)
+ if comment_abr is not None:
+ info['is_favorited'] = 'creatorHeart' in comment_abr
+
+ comment_ab_icontype = traverse_obj(
+ comment_renderer, ('authorCommentBadge', 'authorCommentBadgeRenderer', 'icon', 'iconType'))
+ if comment_ab_icontype is not None:
+ info['author_is_verified'] = comment_ab_icontype in ('CHECK_CIRCLE_THICK', 'OFFICIAL_ARTIST_BADGE')
+
+ is_pinned = traverse_obj(comment_renderer, 'pinnedCommentBadge')
+ if is_pinned:
+ info['is_pinned'] = True
+
+ return info
def _comment_entries(self, root_continuation_data, ytcfg, video_id, parent=None, tracker=None):
expected_comment_count = self._get_count(
comments_header_renderer, 'countText', 'commentsCount')
- if expected_comment_count:
+ if expected_comment_count is not None:
tracker['est_total'] = expected_comment_count
self.to_screen(f'Downloading ~{expected_comment_count} comments')
comment_sort_index = int(get_single_config_arg('comment_sort') != 'top') # 1 = new, 0 = top
comment = self._extract_comment(comment_renderer, parent)
if not comment:
continue
- is_pinned = bool(traverse_obj(comment_renderer, 'pinnedCommentBadge'))
comment_id = comment['id']
- if is_pinned:
+ if comment.get('is_pinned'):
tracker['pinned_comment_ids'].add(comment_id)
# Sometimes YouTube may break and give us infinite looping comments.
# See: https://github.com/yt-dlp/yt-dlp/issues/6290
if comment_id in tracker['seen_comment_ids']:
- if comment_id in tracker['pinned_comment_ids'] and not is_pinned:
+ if comment_id in tracker['pinned_comment_ids'] and not comment.get('is_pinned'):
# Pinned comments may appear a second time in newest first sort
# See: https://github.com/yt-dlp/yt-dlp/issues/6712
continue
if not tracker:
tracker = dict(
running_total=0,
- est_total=0,
+ est_total=None,
current_page_thread=0,
total_parent_comments=0,
total_reply_comments=0,
continuation = self._build_api_continuation_query(self._generate_comment_continuation(video_id))
is_forced_continuation = True
+ continuation_items_path = (
+ 'onResponseReceivedEndpoints', ..., ('reloadContinuationItemsCommand', 'appendContinuationItemsAction'), 'continuationItems')
for page_num in itertools.count(0):
if not continuation:
break
headers = self.generate_api_headers(ytcfg=ytcfg, visitor_data=self._extract_visitor_data(response))
- comment_prog_str = f"({tracker['running_total']}/{tracker['est_total']})"
+ comment_prog_str = f"({tracker['running_total']}/~{tracker['est_total']})"
if page_num == 0:
if is_first_continuation:
note_prefix = 'Downloading comment section API JSON'
note_prefix = '%sDownloading comment%s API JSON page %d %s' % (
' ' if parent else '', ' replies' if parent else '',
page_num, comment_prog_str)
+
+ # Do a deep check for incomplete data as sometimes YouTube may return no comments for a continuation
+ # Ignore check if YouTube says the comment count is 0.
+ 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'))]]
try:
response = self._extract_response(
item_id=None, query=continuation,
ep='next', ytcfg=ytcfg, headers=headers, note=note_prefix,
- check_get_keys='onResponseReceivedEndpoints' if not is_forced_continuation else None)
+ check_get_keys=check_get_keys)
except ExtractorError as e:
# Ignore incomplete data error for replies if retries didn't work.
# This is to allow any other parent comments and comment threads to be downloaded.
else:
raise
is_forced_continuation = False
- continuation_contents = traverse_obj(
- response, 'onResponseReceivedEndpoints', expected_type=list, default=[])
-
continuation = None
- for continuation_section in continuation_contents:
- continuation_items = traverse_obj(
- continuation_section,
- (('reloadContinuationItemsCommand', 'appendContinuationItemsAction'), 'continuationItems'),
- get_all=False, expected_type=list) or []
+ for continuation_items in traverse_obj(response, continuation_items_path, expected_type=list, default=[]):
if is_first_continuation:
continuation = extract_header(continuation_items)
is_first_continuation = False
projects / yt-dlp.git / blobdiff