lgcosta/paperless-ngx
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Improvements to pdf merge

Browse Source
This commit is contained in:
shamoon 2024-03-15 20:50:01 -07:00
parent 45c27dbc0b
commit c1b5a6598c
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
src/documents/bulk_edit.py
View File

@ -200,11 +200,13 @@ def merge(doc_ids: list[int], metadata_document_id: Optional[int] = None):
import pikepdf import pikepdf
merged_pdf = pikepdf.new() merged_pdf = pikepdf.new()
version = merged_pdf.pdf_version
# use doc_ids to preserve order # use doc_ids to preserve order
for doc_id in doc_ids: for doc_id in doc_ids:
doc = qs.get(id=doc_id) doc = qs.get(id=doc_id)
try: try:
with pikepdf.open(str(doc.source_path)) as pdf: with pikepdf.open(str(doc.source_path)) as pdf:
version = max(version, pdf.pdf_version)
merged_pdf.pages.extend(pdf.pages) merged_pdf.pages.extend(pdf.pages)
affected_docs.append(doc.id) affected_docs.append(doc.id)
except Exception as e: except Exception as e:
@ -219,7 +221,8 @@ def merge(doc_ids: list[int], metadata_document_id: Optional[int] = None):
settings.SCRATCH_DIR, settings.SCRATCH_DIR,
f"{'_'.join([str(doc_id) for doc_id in doc_ids])[:100]}_merged.pdf", f"{'_'.join([str(doc_id) for doc_id in doc_ids])[:100]}_merged.pdf",
) )
merged_pdf.save(filepath) merged_pdf.remove_unreferenced_resources()
merged_pdf.save(filepath, min_version=version)
merged_pdf.close() merged_pdf.close()
if metadata_document_id: if metadata_document_id: