Update app.py

app.py

@@ -10,6 +10,7 @@ from langchain_community.document_loaders import PyPDFLoader
 from langchain_text_splitters import RecursiveCharacterTextSplitter
 import io
 import pandas as pd
+import re # Make sure to add this import at the top of your file
 
 # Load environment variables from a .env file
 load_dotenv()
@@ -56,6 +57,7 @@ def delete_file_from_vectorstore(filename):
         return f"Error while deleting the file: {str(e)}", get_files_df()
 
 
+
 def embedder(uploaded_file_path):
     """Handles embedding of the uploaded PDF file."""
     if uploaded_file_path is None:
@@ -96,7 +98,15 @@ def embedder(uploaded_file_path):
         )
         documents = text_splitter.split_documents(raw_documents)
 
-        uuids = [f"{original_filename.replace('.pdf', '')}_{i+1}" for i in range(len(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.
+        uuids = [f"{sanitized_filename}-{i}" for i in range(len(documents))]
+        # --- END: CORRECTED CODE ---
 
         batch_size = 100
         for i in range(0, len(documents), batch_size):
@@ -110,6 +120,7 @@ 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()