obsidian-plugin/lllin000_PaperForge
1
0
Fork 0
mirror of https://github.com/lllin000/PaperForge.git synced 2026-07-22 06:50:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix: reader cards skip caption for matched figures, suppress backmatter at all heading levels, restore figure embeds alongside cards

Browse source
This commit is contained in:
Research Assistant 2026-06-11 19:58:23 +08:00
parent 3bd51dee25
commit bef3ff540e
1 changed files with 21 additions and 20 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

41
paperforge/worker/ocr_render.py
View file

@ -647,7 +647,9 @@ def _render_reader_figure_card(figure: dict) -> list[str]:
lines.append(f"> **Figure {fn}** — {status}")
else:
lines.append(f"> **Figure (unmatched)** — {status}")
if caption_text:
# Only emit caption text when there is no embedded figure asset to show it.
# EXACT_MATCH / SEQUENCE_MATCH already render the caption via figure_XXX.md.
if status not in ("EXACT_MATCH", "SEQUENCE_MATCH") and caption_text:
lines.append(f"> {caption_text}")
return lines
@ -670,17 +672,16 @@ def _emit_page_objects(
"""
has_reader = bool(reader_figures_by_page.get(page))
if not has_reader:
for fig in figures_by_page.get(page, []):
if str(fig['figure_id']).startswith("unmatched_legend_"):
continue
lines.append(f"![[render/figures/{fig['figure_id']}.md]]")
lines.append("")
for cluster_id in unresolved_clusters_by_page.get(page, []):
if str(cluster_id).startswith("unresolved_cluster_"):
continue
lines.append(f"![[render/figures/{cluster_id}.md]]")
lines.append("")
for fig in figures_by_page.get(page, []):
if str(fig['figure_id']).startswith("unmatched_legend_"):
continue
lines.append(f"![[render/figures/{fig['figure_id']}.md]]")
lines.append("")
for cluster_id in unresolved_clusters_by_page.get(page, []):
if str(cluster_id).startswith("unresolved_cluster_"):
continue
lines.append(f"![[render/figures/{cluster_id}.md]]")
lines.append("")
for tbl_id in tables_by_page.get(page, []):
lines.append(f"![[render/tables/{tbl_id}.md]]")
@ -1030,19 +1031,19 @@ def render_fulltext_markdown(
elif role in ("subsection_heading", "sub_subsection_heading", "section_heading"):
last_structured_insert_page = None
last_structured_insert_bbox = None
_BACKMATTER_HEADING_KEYWORDS = frozenset({
"author contributions", "data availability", "funding", "acknowledg",
"conflict of interest", "competing interests", "supplementary material",
"ethics statement", "publisher", "biographies",
})
_heading_lower = text.strip().lower()
if any(kw in _heading_lower for kw in _BACKMATTER_HEADING_KEYWORDS):
continue
if role == "section_heading":
if text.strip().lower() in FRONTMATTER_NOISE:
continue
if "published online" in text.strip().lower():
continue
_BACKMATTER_HEADING_KEYWORDS = frozenset({
"author contributions", "data availability", "funding", "acknowledg",
"conflict of interest", "competing interests", "supplementary material",
"ethics statement", "publisher", "biographies",
})
_heading_lower = text.strip().lower()
if any(kw in _heading_lower for kw in _BACKMATTER_HEADING_KEYWORDS):
continue
if _is_bogus_heading(text):
if text:
lines.append(text)