Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1380,7 +1380,15 @@ def _patch_ocr_with_textlayer(page_md: str, pdf_path: str, page_num: int) -> str
|
|
| 1380 |
needs_checks = "checks" not in result.lower() or "<th>Number</th>" not in result
|
| 1381 |
needs_dab = "daily account balance" not in result.lower()
|
| 1382 |
if needs_checks or needs_dab:
|
| 1383 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1384 |
jb_extra = _extract_jb_summary_sections(raw_tl, needs_checks, needs_dab)
|
| 1385 |
if jb_extra:
|
| 1386 |
result = result.rstrip() + "\n\n" + jb_extra
|
|
|
|
| 1380 |
needs_checks = "checks" not in result.lower() or "<th>Number</th>" not in result
|
| 1381 |
needs_dab = "daily account balance" not in result.lower()
|
| 1382 |
if needs_checks or needs_dab:
|
| 1383 |
+
# Use get_text() directly (not clip-based) to preserve multi-column layout order
|
| 1384 |
+
try:
|
| 1385 |
+
import pymupdf as _fitz
|
| 1386 |
+
_doc = _fitz.open(pdf_path)
|
| 1387 |
+
_page = _doc[page_num]
|
| 1388 |
+
raw_tl = _page.get_text()
|
| 1389 |
+
_doc.close()
|
| 1390 |
+
except Exception:
|
| 1391 |
+
raw_tl = ""
|
| 1392 |
jb_extra = _extract_jb_summary_sections(raw_tl, needs_checks, needs_dab)
|
| 1393 |
if jb_extra:
|
| 1394 |
result = result.rstrip() + "\n\n" + jb_extra
|