Update app.py

app.py

@@ -98,15 +98,22 @@ def embedder(uploaded_file_path):
         )
         documents = text_splitter.split_documents(raw_documents)
 
-        # --- START: CORRECTED CODE ---
-        # 1. Sanitize the base filename to conform to Pinecone's rules.
-        sanitized_filename = original_filename.replace('.pdf', '').lower()
-        # 2. Replace all invalid characters (anything not a-z, 0-9, or -) with a hyphen.
-        sanitized_filename = re.sub(r'[^a-z0-9-]', '-', sanitized_filename)
-
-        # 3. Use the sanitized filename to generate the vector IDs.
+        # --- INIZIO: CODICE CORRETTO ---
+        # 1. Rimuove l'estensione, gli spazi iniziali/finali e converte in minuscolo.
+        sanitized_filename = original_filename.replace('.pdf', '').strip().lower()
+        
+        # 2. Sostituisce qualsiasi carattere non alfanumerico con un trattino.
+        sanitized_filename = re.sub(r'[^a-z0-9]', '-', sanitized_filename)
+        
+        # 3. Sostituisce trattini multipli consecutivi con un singolo trattino.
+        sanitized_filename = re.sub(r'-+', '-', sanitized_filename)
+        
+        # 4. (Opzionale ma consigliato) Rimuove eventuali trattini all'inizio o alla fine.
+        sanitized_filename = sanitized_filename.strip('-')
+
+        # 5. Usa il nome sanificato per generare gli ID.
         uuids = [f"{sanitized_filename}-{i}" for i in range(len(documents))]
-        # --- END: CORRECTED CODE ---
+        # --- FINE: CODICE CORRETTO ---
 
         batch_size = 100
         for i in range(0, len(documents), batch_size):
@@ -120,7 +127,6 @@ def embedder(uploaded_file_path):
         return f"Unable to create embeddings: {str(e)}", get_files_df()
 
 
-
 # --- Gradio Interface Functions ---
 def get_files_df():
     files = get_stored_files()