AI_POC/TERES_fastapi_backend
12
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

v0.21.1-fastapi

Browse Source
This commit is contained in:
dangzerong
2025-11-04 16:06:36 +08:00
parent 3e58c3d0e9
commit d57b5d76ae
218 changed files with 19617 additions and 72339 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
rag/app/one.py
View File

@@ -23,6 +23,7 @@ from deepdoc.parser.utils import get_text
from rag.app import naive
from rag.nlp import rag_tokenizer, tokenize
from deepdoc.parser import PdfParser, ExcelParser, PlainParser, HtmlParser
from deepdoc.parser.figure_parser import vision_figure_parser_pdf_wrapper,vision_figure_parser_docx_wrapper
class Pdf(PdfParser):
@@ -57,13 +58,8 @@ class Pdf(PdfParser):
sections = [(b["text"], self.get_position(b, zoomin))
for i, b in enumerate(self.boxes)]
for (img, rows), poss in tbls:
if not rows:
continue
sections.append((rows if isinstance(rows, str) else rows[0],
[(p[0] + 1 - from_page, p[1], p[2], p[3], p[4]) for p in poss]))
return [(txt, "") for txt, _ in sorted(sections, key=lambda x: (
x[-1][0][0], x[-1][0][3], x[-1][0][1]))], None
x[-1][0][0], x[-1][0][3], x[-1][0][1]))], tbls
def chunk(filename, binary=None, from_page=0, to_page=100000,
@@ -80,6 +76,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
if re.search(r"\.docx$", filename, re.IGNORECASE):
callback(0.1, "Start to parse.")
sections, tbls = naive.Docx()(filename, binary)
tbls=vision_figure_parser_docx_wrapper(sections=sections,tbls=tbls,callback=callback,**kwargs)
sections = [s for s, _ in sections if s]
for (_, html), _ in tbls:
sections.append(html)
@@ -89,8 +86,14 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
pdf_parser = Pdf()
if parser_config.get("layout_recognize", "DeepDOC") == "Plain Text":
pdf_parser = PlainParser()
sections, _ = pdf_parser(
sections, tbls = pdf_parser(
filename if not binary else binary, to_page=to_page, callback=callback)
tbls=vision_figure_parser_pdf_wrapper(tbls=tbls,callback=callback,**kwargs)
for (img, rows), poss in tbls:
if not rows:
continue
sections.append((rows if isinstance(rows, str) else rows[0],
[(p[0] + 1 - from_page, p[1], p[2], p[3], p[4]) for p in poss]))
sections = [s for s, _ in sections if s]
elif re.search(r"\.xlsx?$", filename, re.IGNORECASE):