From 968f846b88cbde635887edec97a069f25d4a62f6 Mon Sep 17 00:00:00 2001 From: mart-r Date: Wed, 27 Aug 2025 14:27:03 +0100 Subject: [PATCH 1/2] CU-869a95nu1: Fix spacy eokznier old path cleanup issue --- medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py b/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py index 71634b24..ab57de01 100644 --- a/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py +++ b/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py @@ -106,7 +106,7 @@ def save_internals_to(self, folder_path: str) -> str: folder_path, f"{TOKENIZER_PREFIX}{self._spacy_model_name}") if os.path.exists(subfolder): # NOTE: always overwrite - shutil.rmtree(folder_path) + shutil.rmtree(subfolder) logger.debug("Saving spacy model to '%s'", subfolder) cur_path = self._nlp._path if cur_path is None: From 94fe65b08f5ea0a4159ef989b9aff7b655132348 Mon Sep 17 00:00:00 2001 From: mart-r Date: Wed, 27 Aug 2025 14:28:49 +0100 Subject: [PATCH 2/2] CU-869a95nu1: Simplify spacy model saving somewhat --- medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py b/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py index ab57de01..b6fe35b7 100644 --- a/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py +++ b/medcat-v2/medcat/tokenizing/spacy_impl/tokenizers.py @@ -108,10 +108,7 @@ def save_internals_to(self, folder_path: str) -> str: # NOTE: always overwrite shutil.rmtree(subfolder) logger.debug("Saving spacy model to '%s'", subfolder) - cur_path = self._nlp._path - if cur_path is None: - raise ValueError(f"Unable to save spacy: {self._nlp}") - shutil.copytree(cur_path, subfolder) + self._nlp.to_disk(subfolder) return subfolder def load_internals_from(self, folder_path: str) -> bool: