Fix IDE warnings (#12281)

### What problem does this PR solve?

As title

### Type of change

- [x] Refactoring

---------

Signed-off-by: Jin Hai <haijin.chn@gmail.com>
This commit is contained in:
Jin Hai
2025-12-29 12:01:18 +08:00
committed by GitHub
parent 647fb115a0
commit 01f0ced1e6
43 changed files with 817 additions and 637 deletions

View File

@ -26,6 +26,7 @@ from deepdoc.parser.figure_parser import vision_figure_parser_docx_wrapper
from rag.app.naive import by_plaintext, PARSERS
from common.parser_config_utils import normalize_layout_recognizer
class Pdf(PdfParser):
def __call__(self, filename, binary=None, from_page=0,
to_page=100000, zoomin=3, callback=None):
@ -95,14 +96,14 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
callback(0.1, "Start to parse.")
sections, tbls, pdf_parser = parser(
filename = filename,
binary = binary,
from_page = from_page,
to_page = to_page,
lang = lang,
callback = callback,
pdf_cls = Pdf,
layout_recognizer = layout_recognizer,
filename=filename,
binary=binary,
from_page=from_page,
to_page=to_page,
lang=lang,
callback=callback,
pdf_cls=Pdf,
layout_recognizer=layout_recognizer,
mineru_llm_name=parser_model_name,
**kwargs
)
@ -112,9 +113,9 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
if name in ["tcadp", "docling", "mineru"]:
parser_config["chunk_token_num"] = 0
callback(0.8, "Finish parsing.")
for (img, rows), poss in tbls:
if not rows:
continue
@ -172,7 +173,9 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
if __name__ == "__main__":
import sys
def dummy(prog=None, msg=""):
pass
chunk(sys.argv[1], from_page=0, to_page=10, callback=dummy)