From b897d6de2e3fbd26c6b6bcd9c09932e99e14fead Mon Sep 17 00:00:00 2001
From: Trenton H <797416+stumpylog@users.noreply.github.com>
Date: Mon, 21 Nov 2022 14:45:20 -0800
Subject: [PATCH] Don't use the sidecar file when redoing the OCR, it only
contains new text
---
src/paperless_tesseract/parsers.py | 16 ++++++++++++----
1 file changed, 12 insertions(+), 4 deletions(-)
diff --git a/src/paperless_tesseract/parsers.py b/src/paperless_tesseract/parsers.py
index 405df07ce..aa3ad64fa 100644
--- a/src/paperless_tesseract/parsers.py
+++ b/src/paperless_tesseract/parsers.py
@@ -95,7 +95,13 @@ class RasterisedDocumentParser(DocumentParser):
return None
def extract_text(self, sidecar_file, pdf_file):
- if sidecar_file and os.path.isfile(sidecar_file):
+ # When re-doing OCR, the sidecar contains ONLY the new text, not
+ # the whole text, so do not utilize it in that case
+ if (
+ sidecar_file is not None
+ and os.path.isfile(sidecar_file)
+ and settings.OCR_MODE != "redo"
+ ):
with open(sidecar_file) as f:
text = f.read()
@@ -142,7 +148,7 @@ class RasterisedDocumentParser(DocumentParser):
"input_file": input_file,
"output_file": output_file,
# need to use threads, since this will be run in daemonized
- # processes by django-q.
+ # processes via the task library.
"use_threads": True,
"jobs": settings.THREADS_PER_WORKER,
"language": settings.OCR_LANGUAGE,
@@ -165,9 +171,11 @@ class RasterisedDocumentParser(DocumentParser):
if settings.OCR_MODE == "redo":
ocrmypdf_args["clean"] = True
else:
+ # --clean-final is not compatible with --redo-ocr
ocrmypdf_args["clean_final"] = True
- if settings.OCR_DESKEW and not settings.OCR_MODE == "redo":
+ if settings.OCR_DESKEW and settings.OCR_MODE != "redo":
+ # --deskew is not compatible with --redo-ocr
ocrmypdf_args["deskew"] = True
if settings.OCR_ROTATE_PAGES:
@@ -263,7 +271,7 @@ class RasterisedDocumentParser(DocumentParser):
# Either no text was in the original or there should be an archive
# file created, so OCR the file and create an archive with any
- # test located via OCR
+ # text located via OCR
import ocrmypdf
from ocrmypdf import InputFileError, EncryptedPdfError