KMS/ragflow
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

Fix pdfparser content confusion (#1700)

Browse Source 
### What problem does this PR solve?

#1407 #1656 

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
This commit is contained in:
H 2024-07-25 11:40:23 +08:00 committed by GitHub
parent 6fee2962cb
commit b24abee364
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 2 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
deepdoc/parser/pdf_parser.py
View File

@ -285,24 +285,10 @@ class RAGFlowPdfParser:
"page_number": pagenum} for b, t in bxs if b[0][0] <= b[1][0] and b[0][1] <= b[-1][1]],
self.mean_height[-1] / 3
)
# solve char content confusion
record_error_length, ct = 0, 0.001
for c in chars[0:128]:
ii = Recognizer.find_overlapped(c, bxs)
if ii is None:
continue
record_error_length += abs((bxs[ii]["bottom"] + bxs[ii]["top"] - c["bottom"] - c["top"]) / 2)
ct += 1
record_error_length = record_error_length / ct
for char in chars:
char["top"] -= record_error_length
char["bottom"] -= record_error_length
# merge chars in the same rect
for c in Recognizer.sort_X_firstly(
chars, self.mean_width[pagenum - 1] // 4):
for c in Recognizer.sort_Y_firstly(
chars, self.mean_height[pagenum - 1] // 4):
ii = Recognizer.find_overlapped(c, bxs)
if ii is None:
self.lefted_chars.append(c)