Update utils.py

utils.py

@@ -390,17 +390,17 @@ def document_storage_chroma(splits):
     embedding_fn = HuggingFaceEmbeddings(model_name=EMBEDDING_MODELL, model_kwargs={"device": "cpu"}, encode_kwargs={'normalize_embeddings': False})
 
     # Vectorstore initialisieren und Dokumente hinzufügen
-    vectorstore = Chroma.from_documents(documents=splits, embedding=embedding_fn) #, persist_directory = PATH_WORK + CHROMA_DIR)
+    vectorstore = Chroma.from_documents(documents=splits, embedding=embedding_fn) 
     
     return vectorstore
 
 ########################################################
 #Vektorstore speichern - bzw. laden
-def save_vectorstore(vectorstore, filename="vectorstore.pkl"):
+def save_splits_and_metadata(splits, filename="splits_and_metadata.pkl"):
     with open(filename, "wb") as f:
-        pickle.dump(vectorstore, f)
+        pickle.dump(splits, f)
 
-def load_vectorstore(filename="vectorstore.pkl"):
+def load_splits_and_metadata(filename="splits_and_metadata.pkl"):
     if os.path.exists(filename):
         with open(filename, "rb") as f:
             return pickle.load(f)
@@ -413,11 +413,19 @@ def create_vectorstore():
     if PREPROCESSED_SPLITS:
         # Vektordatenbank zu den Splits erstellen
         vektordatenbank = document_storage_chroma(PREPROCESSED_SPLITS)
-        # Speichern des Vektorstores
-        save_vectorstore(vektordatenbank)
+        # Speichern der Splits und Metadaten
+        save_splits_and_metadata((PREPROCESSED_SPLITS, SPLIT_TO_ORIGINAL_MAPPING))
         return vektordatenbank
     else:
         return None
+        
+#Laden des Vektorstores - aus den gespeicherten splits
+def load_vectorstore():
+    splits_and_metadata = load_splits_and_metadata()
+    if splits_and_metadata is not None:
+        PREPROCESSED_SPLITS, SPLIT_TO_ORIGINAL_MAPPING = splits_and_metadata
+        return document_storage_chroma(PREPROCESSED_SPLITS)
+    return None
 
 # Beispiel-Upload-Funktion
 def upload_file_to_huggingface(file_path, upload_path):