paradizelost/paperless-ngx
1
0
Fork 0
mirror of https://github.com/paperless-ngx/paperless-ngx.git synced 2025-04-09 09:58:20 -05:00
Code Issues Packages Projects Releases Wiki Activity

Don't use the sidecar file when redoing the OCR, it only contains new text

Browse Source
This commit is contained in:
Trenton H 2022-11-21 14:45:20 -08:00
parent 54f20b381e
commit b897d6de2e
1 changed files with 12 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
src/paperless_tesseract/parsers.py
View File

@ -95,7 +95,13 @@ class RasterisedDocumentParser(DocumentParser):
return None return None
def extract_text(self, sidecar_file, pdf_file): 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: with open(sidecar_file) as f:
text = f.read() text = f.read()
@ -142,7 +148,7 @@ class RasterisedDocumentParser(DocumentParser):
"input_file": input_file, "input_file": input_file,
"output_file": output_file, "output_file": output_file,
# need to use threads, since this will be run in daemonized # need to use threads, since this will be run in daemonized
# processes by django-q. # processes via the task library.
"use_threads": True, "use_threads": True,
"jobs": settings.THREADS_PER_WORKER, "jobs": settings.THREADS_PER_WORKER,
"language": settings.OCR_LANGUAGE, "language": settings.OCR_LANGUAGE,
@ -165,9 +171,11 @@ class RasterisedDocumentParser(DocumentParser):
if settings.OCR_MODE == "redo": if settings.OCR_MODE == "redo":
ocrmypdf_args["clean"] = True ocrmypdf_args["clean"] = True
else: else:
# --clean-final is not compatible with --redo-ocr
ocrmypdf_args["clean_final"] = True 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 ocrmypdf_args["deskew"] = True
if settings.OCR_ROTATE_PAGES: 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 # 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 # file created, so OCR the file and create an archive with any
# test located via OCR # text located via OCR
import ocrmypdf import ocrmypdf
from ocrmypdf import InputFileError, EncryptedPdfError from ocrmypdf import InputFileError, EncryptedPdfError