pengjunjie/ragflow
1
0
Fork 1
mirror of https://github.com/infiniflow/ragflow.git synced 2025-12-31 17:15:32 +08:00
Code Issues Packages Projects Releases Wiki Activity

Refine: image/table context. (#12336)

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

#12303

### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
This commit is contained in:
Kevin Hu
2025-12-30 20:24:27 +08:00
committed by GitHub
parent 348265afc1
commit 52f91c2388
5 changed files with 116 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
deepdoc/parser/pdf_parser.py
View File

@ -476,11 +476,13 @@ class RAGFlowPdfParser:
self.boxes = bxs
def _naive_vertical_merge(self, zoomin=3):
bxs = self._assign_column(self.boxes, zoomin)
#bxs = self._assign_column(self.boxes, zoomin)
bxs = self.boxes
grouped = defaultdict(list)
for b in bxs:
grouped[(b["page_number"], b.get("col_id", 0))].append(b)
# grouped[(b["page_number"], b.get("col_id", 0))].append(b)
grouped[(b["page_number"], "x")].append(b)
merged_boxes = []
for (pg, col), bxs in grouped.items():
@ -551,7 +553,7 @@ class RAGFlowPdfParser:
merged_boxes.extend(bxs)
self.boxes = sorted(merged_boxes, key=lambda x: (x["page_number"], x.get("col_id", 0), x["top"]))
#self.boxes = sorted(merged_boxes, key=lambda x: (x["page_number"], x.get("col_id", 0), x["top"]))
def _final_reading_order_merge(self, zoomin=3):
if not self.boxes:
@ -1206,7 +1208,7 @@ class RAGFlowPdfParser:
start = timer()
self._text_merge()
self._concat_downward()
#self._naive_vertical_merge(zoomin)
self._naive_vertical_merge(zoomin)
if callback:
callback(0.92, "Text merged ({:.2f}s)".format(timer() - start))