Update app.py

app.py

@@ -12,6 +12,8 @@ from pptx import Presentation
 from io import BytesIO
 import shutil
 import logging
+import chromadb
+import tempfile
 
 # Set up logging
 logging.basicConfig(level=logging.INFO)
@@ -40,7 +42,7 @@ vector_store = None
 qa_chain = None
 chat_history = []
 memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
-PERSIST_DIRECTORY = "./chroma_db"
+PERSIST_DIRECTORY = tempfile.mkdtemp()  # Use temporary directory for ChromaDB
 
 # Custom PPTX loader
 class PPTXLoader:
@@ -94,7 +96,7 @@ def process_documents(files, chunk_size, chunk_overlap, embedding_model):
     if not files:
         return "Please upload at least one document.", None
 
-    # Clear existing vector store to avoid dimensionality mismatch
+    # Clear existing vector store
     if os.path.exists(PERSIST_DIRECTORY):
         try:
             shutil.rmtree(PERSIST_DIRECTORY)
@@ -102,6 +104,7 @@ def process_documents(files, chunk_size, chunk_overlap, embedding_model):
         except Exception as e:
             logger.error(f"Error clearing ChromaDB directory: {str(e)}")
             return f"Error clearing vector store: {str(e)}", None
+    os.makedirs(PERSIST_DIRECTORY, exist_ok=True)
 
     # Load documents
     documents = load_documents(files)
@@ -130,7 +133,15 @@ def process_documents(files, chunk_size, chunk_overlap, embedding_model):
 
     # Create vector store
     try:
-        vector_store = Chroma.from_documents(doc_splits, embeddings, persist_directory=PERSIST_DIRECTORY)
+        # Reset ChromaDB client to avoid tenant issues
+        client = chromadb.PersistentClient(path=P Persistent_DIRECTORY)
+        client.reset()
+        vector_store = Chroma.from_documents(
+            documents=doc_splits,
+            embedding=embeddings,
+            persist_directory=PERSIST_DIRECTORY,
+            collection_name="doctalk_collection"
+        )
         return f"Processed {len(documents)} documents into {len(doc_splits)} chunks.", None
     except Exception as e:
         logger.error(f"Error creating vector store: {str(e)}")
@@ -173,8 +184,8 @@ def answer_question(question, llm_model, embedding_model, temperature, chunk_siz
 
     try:
         response = qa_chain({"question": question})["answer"]
-        chat_history.append(("User", question))
-        chat_history.append(("Bot", response))
+        chat_history.append({"role": "user", "content": question})
+        chat_history.append({"role": "assistant", "content": response})
         logger.info(f"Answered question: {question}")
         return response, chat_history
     except Exception as e:
@@ -189,8 +200,10 @@ def export_chat():
         timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
         filename = f"chat_history_{timestamp}.txt"
         with open(filename, "w") as f:
-            for role, message in chat_history:
-                f.write(f"{role}: {message}\n\n")
+            for message in chat_history:
+                role = message["role"].capitalize()
+                content = message["content"]
+                f.write(f"{role}: {content}\n\n")
         logger.info(f"Exported chat history to {filename}.")
         return f"Chat history exported to {filename}.", filename
     except Exception as e:
@@ -207,6 +220,7 @@ def reset_app():
         memory = ConversationBufferMemory(memory_key="chat_history", return_messages=True)
         if os.path.exists(PERSIST_DIRECTORY):
             shutil.rmtree(PERSIST_DIRECTORY)
+            os.makedirs(PERSIST_DIRECTORY, exist_ok=True)
             logger.info("Cleared ChromaDB directory on reset.")
         logger.info("App reset successfully.")
         return "App reset successfully.", None
@@ -238,20 +252,22 @@ with gr.Blocks(theme=gr.themes.Soft(), title="DocTalk: Document Q&A Chatbot") as
     gr.Markdown("## Chat Interface")
     question = gr.Textbox(label="Ask a Question", placeholder="Type your question here...")
     answer = gr.Textbox(label="Answer", interactive=False)
-    chat_display = gr.Chatbot(label="Chat History")
+    chat_display = gr.Chatbot(label="Chat History", type="messages")
     export_button = gr.Button("Export Chat History")
     export_file = gr.File(label="Exported Chat File")
 
-    # Event handlers
+    # Event handlers with loading state
     process_button.click(
         fn=process_documents,
         inputs=[file_upload, chunk_size, chunk_overlap, embedding_model],
-        outputs=[status, chat_display]
+        outputs=[status, chat_display],
+        _js="() => { return { 'Process Documents': 'Processing...' } }"
     )
     init_button.click(
         fn=initialize_qa_chain,
         inputs=[llm_model, temperature],
-        outputs=[status, chat_display]
+        outputs=[status, chat_display],
+        _js="() => { return { 'Initialize QA Chain': 'Initializing...' } }"
     )
     question.submit(
         fn=answer_question,