Update app.py

app.py

@@ -30,9 +30,9 @@ st.set_page_config(page_title="Gemini RAG", layout="wide")
 api_key = 'AIzaSyCvXRggpO2yNwIpZmoMy_5Xhm2bDyD-pOo'
 
 
-# os.mkdir('faiss_index')
+#os.mkdir('faiss_index')
 
-# empty faise_index and chat_history.json
+# empty faiss_index and chat_history.json
 def delete_files_in_folder(folder_path):
     try:
         # Iterate over all the files in the folder
@@ -134,7 +134,49 @@ def user_input(user_question, api_key):
     #chat history
     update_chat_history(user_question, response["output_text"])
 
+''''''''''''''''''
 
+def clear_faiss_index(folder_path):
+    try:
+        if os.path.exists(folder_path):
+            for file_name in os.listdir(folder_path):
+                file_path = os.path.join(folder_path, file_name)
+                if os.path.isfile(file_path):
+                    os.remove(file_path)
+            st.write("Existing FAISS index files cleared successfully!")
+        else:
+            st.write("No existing FAISS index files found.")
+    except Exception as e:
+        st.error(f"An error occurred while clearing FAISS index files: {e}")
+# Function to process PDF files and recreate FAISS index
+
+
+def recreate_faiss_index(pdf_docs, chunk_size, chunk_overlap, api_key):
+    try:
+        # Clear existing FAISS index files
+        clear_faiss_index("faiss_index")
+
+        # Process PDF files and extract text
+        text = ""
+        for pdf in pdf_docs:
+            pdf_reader = PdfReader(pdf)
+            for page in pdf_reader.pages:
+                text += page.extract_text()
+
+        # Split text into chunks
+        text_splitter = RecursiveCharacterTextSplitter(chunk_size=chunk_size, chunk_overlap=chunk_overlap)
+        chunks = text_splitter.split_text(text)
+
+        # Generate embeddings for text chunks
+        embeddings = GoogleGenerativeAIEmbeddings(model="models/embedding-001", google_api_key=api_key)
+        vector_store = FAISS.from_texts(chunks, embedding=embeddings)
+
+        # Save FAISS index
+        vector_store.save_local("faiss_index")
+
+        st.success("FAISS index recreated successfully!")
+    except Exception as e:
+        st.error(f"An error occurred while recreating FAISS index: {e}")
 
 
 def main():
@@ -152,6 +194,8 @@ def main():
                                     accept_multiple_files=True, key="pdf_uploader")
         if st.button("Submit & Process", key="process_button") and api_key:
             with st.spinner("Processing..."):
+                recreate_faiss_index(pdf_docs, CH_size, CH_overlap, api_key)
+
                 raw_text = get_pdf_text(pdf_docs)
                 text_chunks = get_text_chunks(raw_text)
                 get_vector_store(text_chunks, api_key)
@@ -159,4 +203,5 @@ def main():
 
 
 if __name__ == "__main__":
-    main()
+    main()
+