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

fix table desc bugs, add positions to chunks (#91)

Browse Source
This commit is contained in:
KevinHuSh
2024-03-04 14:42:26 +08:00
committed by GitHub
parent 8a726fb04b
commit 685b4d8a95
13 changed files with 117 additions and 46 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
rag/app/manual.py
View File

@ -2,7 +2,7 @@ import copy
import re
from api.db import ParserType
from rag.nlp import huqie, tokenize, tokenize_table
from rag.nlp import huqie, tokenize, tokenize_table, add_positions
from deepdoc.parser import PdfParser
from rag.utils import num_tokens_from_string
@ -14,6 +14,7 @@ class Pdf(PdfParser):
def __call__(self, filename, binary=None, from_page=0,
to_page=100000, zoomin=3, callback=None):
callback(msg="OCR is running...")
self.__images__(
filename if not binary else binary,
zoomin,
@ -32,7 +33,7 @@ class Pdf(PdfParser):
self._concat_downward(concat_between_pages=False)
self._filter_forpages()
callback(0.77, "Text merging finished")
tbls = self._extract_table_figure(True, zoomin, False)
tbls = self._extract_table_figure(True, zoomin, False, True)
# clean mess
for b in self.boxes:
@ -91,7 +92,8 @@ def chunk(filename, binary=None, from_page=0, to_page=100000, lang="Chinese", ca
d = copy.deepcopy(doc)
ck = "\n".join(chunk)
tokenize(d, pdf_parser.remove_tag(ck), pdf_parser.is_english)
d["image"] = pdf_parser.crop(ck)
d["image"], poss = pdf_parser.crop(ck, need_position=True)
add_positions(d, poss)
res.append(d)
chunk = []
tk_cnt = 0