From 90ba6a12c7d87c6335f8467df8104d50f6e423c3 Mon Sep 17 00:00:00 2001
From: Vit Novotny <witiko@mail.muni.cz>
Date: Wed, 24 Aug 2022 21:05:51 +0200
Subject: [PATCH] Do not copy /pero-ocr/*.txt to /output of PERO OCR unless the
page is single-column
---
scripts/combine_tesseract_with_pero_ocr_docker.py | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/scripts/combine_tesseract_with_pero_ocr_docker.py b/scripts/combine_tesseract_with_pero_ocr_docker.py
index 5d373042..a530d7d2 100644
--- a/scripts/combine_tesseract_with_pero_ocr_docker.py
+++ b/scripts/combine_tesseract_with_pero_ocr_docker.py
@@ -31,7 +31,7 @@ def main() -> None:
(OUTPUT_PATH / tesseract_filename.stem).with_suffix(f'.{suffix}'),
)
- for suffix in ('xml', 'txt'):
+ for suffix in ('xml', ):
shutil.copy(
(PERO_OCR_PATH / pero_ocr_filename.stem).with_suffix(f'.{suffix}'),
(OUTPUT_PATH / pero_ocr_filename.stem).with_suffix(f'.{suffix}'),
--
GitLab