From 12f95cff91813a2143bc8fab5bc3e89df1bd0fdb Mon Sep 17 00:00:00 2001 From: Codex Bot Date: Wed, 18 Nov 2026 04:21:31 +0800 Subject: [PATCH] Reject incomplete MM title markup before HTML parsing --- paper_atlas/collectors/mm_publications2026.py | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/paper_atlas/collectors/mm_publications2026.py b/paper_atlas/collectors/mm_publications2026.py index 90792ba..c3f341e 100644 --- a/paper_atlas/collectors/mm_publications2026.py +++ b/paper_atlas/collectors/mm_publications2026.py @@ -155,6 +155,11 @@ class _TitleFormatting(HTMLParser): def formatting_text(title): + # HTMLParser may silently discard a trailing incomplete start tag. + # Check lexical coverage before parsing so no title content can disappear. + literal=re.sub(r'','',title,flags=re.I) + if '<' in literal or '>' in literal: + return None parser=_TitleFormatting() parser.feed(title);parser.close() text=''.join(parser.parts)