[extractor/youtube] Fix comment loop detection for pinned comments (#6714)

Pinned comments may repeat a second time - this is expected.

Fixes https://github.com/yt-dlp/yt-dlp/issues/6712

Authored by: coletdjnz

diff --git a/yt_dlp/extractor/youtube.py b/yt_dlp/extractor/youtube.py
index ca56f112bb7e8a19cada9569bd441bc0f0404094..6dc36f9b998db35131d1e8daf5f1ed84d23a3134 100644 (file)
--- a/yt_dlp/extractor/youtube.py
+++ b/yt_dlp/extractor/youtube.py
@@ -3316,9 +3316,17 @@ def extract_thread(contents):
                 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:
+                    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['seen_comment_ids']:
+                    if comment_id in tracker['pinned_comment_ids'] and not is_pinned:
+                        # Pinned comments may appear a second time in newest first sort
+                        # See: https://github.com/yt-dlp/yt-dlp/issues/6712
+                        continue
                     self.report_warning('Detected YouTube comments looping. Stopping comment extraction as we probably cannot get any more.')
                     yield
                 else:
@@ -3348,7 +3356,9 @@ def extract_thread(contents):
                 current_page_thread=0,
                 total_parent_comments=0,
                 total_reply_comments=0,
-                seen_comment_ids=set())
+                seen_comment_ids=set(),
+                pinned_comment_ids=set()
+            )
 
         # TODO: Deprecated
         # YouTube comments have a max depth of 2