pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-08 20:42:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

apply pep8 formalize (#155)

Browse Source
This commit is contained in:
KevinHuSh
2024-03-27 11:33:46 +08:00
committed by GitHub
parent a02e836790
commit fd7fcb5baf
55 changed files with 1568 additions and 753 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
deepdoc/parser/docx_parser.py
View File

@ -99,12 +99,15 @@ class HuDocxParser:
return ["\n".join(lines)]
def __call__(self, fnm, from_page=0, to_page=100000):
self.doc = Document(fnm) if isinstance(fnm, str) else Document(BytesIO(fnm))
self.doc = Document(fnm) if isinstance(
fnm, str) else Document(BytesIO(fnm))
pn = 0
secs = []
for p in self.doc.paragraphs:
if pn > to_page: break
if from_page <= pn < to_page and p.text.strip(): secs.append((p.text, p.style.name))
if pn > to_page:
break
if from_page <= pn < to_page and p.text.strip():
secs.append((p.text, p.style.name))
for run in p.runs:
if 'lastRenderedPageBreak' in run._element.xml:
pn += 1