From 6965165c76e501c09366850fb67f03d56724282f Mon Sep 17 00:00:00 2001
From: Trenton H <797416+stumpylog@users.noreply.github.com>
Date: Sun, 14 Jul 2024 16:35:16 -0700
Subject: [PATCH] Removes Turkish from the NLTK languages (#7246)
---
src/paperless/settings.py | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/src/paperless/settings.py b/src/paperless/settings.py
index fe7caebb8..000904aef 100644
--- a/src/paperless/settings.py
+++ b/src/paperless/settings.py
@@ -1126,6 +1126,10 @@ def _get_nltk_language_setting(ocr_lang: str) -> Optional[str]:
NLTK Languages:
- https://www.nltk.org/api/nltk.stem.snowball.html#nltk.stem.snowball.SnowballStemmer
+ - https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/packages/tokenizers/punkt.zip
+ - https://raw.githubusercontent.com/nltk/nltk_data/gh-pages/packages/corpora/stopwords.zip
+
+ The common intersection between all languages in those 3 is handled here
"""
ocr_lang = ocr_lang.split("+")[0]
@@ -1142,7 +1146,6 @@ def _get_nltk_language_setting(ocr_lang: str) -> Optional[str]:
"rus": "russian",
"spa": "spanish",
"swe": "swedish",
- "tur": "turkish",
}
return iso_code_to_nltk.get(ocr_lang)