From 4d5180ee6181de8b07abfcdab0af21dbae822272 Mon Sep 17 00:00:00 2001 From: Research Assistant Date: Fri, 19 Jun 2026 14:57:10 +0800 Subject: [PATCH] fix: restore post-boundary body text markers, add curly quote publisher note --- paperforge/worker/ocr_document.py | 6 ++++++ paperforge/worker/ocr_roles.py | 2 ++ 2 files changed, 8 insertions(+) diff --git a/paperforge/worker/ocr_document.py b/paperforge/worker/ocr_document.py index 9fb97919..e35a53b0 100644 --- a/paperforge/worker/ocr_document.py +++ b/paperforge/worker/ocr_document.py @@ -2773,6 +2773,12 @@ def _looks_like_backmatter_body_text(text: str) -> bool: "conflict of interest", "declaration", "publisher", + "author contributions", + "funding", + "acknowledg", + "data availability", + "supplement", + "ethics", "copyright", ) return any(marker in lower for marker in markers) diff --git a/paperforge/worker/ocr_roles.py b/paperforge/worker/ocr_roles.py index 0d0929e3..d39f7557 100644 --- a/paperforge/worker/ocr_roles.py +++ b/paperforge/worker/ocr_roles.py @@ -35,6 +35,7 @@ _BACKMATTER_TITLE_DENY_LIST: set[str] = set() _BACKMATTER_HEADINGS = { "publisher's note", + "publisher\u2019s note", } _PAGE1_ARTICLE_TYPE_LABELS = frozenset( @@ -557,6 +558,7 @@ def resolve_final_role( phrase in text_lower for phrase in ( "conflict of interest", + "publisher's note", "publisher\u2019s note", "the remaining authors declare", "copyright",