Update utils.py

utils.py

@@ -321,7 +321,11 @@ def load_word_with_metadata(file_path):
 ################################################
 #Vektorstore
 ################################################
-
+#liste aller files in einem dir...
+def list_files_in_hf_repo(repo_id):
+    repo_info = api.list_repo_files(repo_id=repo_id, repo_type=REPO_TYPE)
+    return repo_info
+    
 ################################################
 # Document Splitting
 ################################################    
@@ -353,21 +357,25 @@ def document_loading_splitting():
 
 
     # Verzeichnis für heruntergeladene Dateien
-    download_dir = "downloaded_files"
-    os.makedirs(download_dir, exist_ok=True)
+    #download_dir = CHROMA_PDF
+    #os.makedirs(download_dir, exist_ok=True)
 
     # Dateien im Hugging Face Space auflisten
     files_in_repo = list_files_in_hf_repo(STORAGE_REPO_ID)
 
-    # Dateien aus dem Hugging Face Space herunterladen
+    # Dateien aus dem Hugging Face Space mit der STORAGE_REPO_ID herunterladen
     for file_name in files_in_repo:
-        if file_name.endswith('.pdf') or file_name.endswith('.docx'):
-            local_file_path = os.path.join(download_dir, os.path.basename(file_name))
+        if file_name.endswith('.pdf'):
+            local_file_path = os.path.join(CHROMA_PDF, os.path.basename(file_name))
             download_file_from_hf(file_name, local_file_path)
-
+        if file_name.endswith('.docx'):
+            local_file_path = os.path.join(CHROMA_WORD, os.path.basename(file_name))
+            download_file_from_hf(file_name, local_file_path)
+        print("file_name..................."+str(file_name))
+        print("local_file_path..................."+str(local_file_path)) 
     # Erstellen von DirectoryLoader für jeden Dateityp
-    pdf_loader = create_directory_loader('.pdf', download_dir)
-    word_loader = create_directory_loader('.word', download_dir)
+    pdf_loader = create_directory_loader('.pdf', CHROMA_PDF)
+    word_loader = create_directory_loader('.word', CHROMA_WORD)
 
 
 
@@ -531,9 +539,6 @@ def download_file_from_hf(file_name, save_path):
         file.write(response.content)
     return save_path
 
-def list_files_in_hf_repo(repo_id):
-    repo_info = api.list_repo_files(repo_id=repo_id, repo_type=REPO_TYPE)
-    return repo_info