Update app.py

app.py

@@ -6,7 +6,7 @@ from llama_index.llms.huggingface_api import HuggingFaceInferenceAPI
 import os
 from dotenv import load_dotenv
 import gradio as gr
-import markdowm as md
+import markdown as md
 import base64
 
 # Load environment variables
@@ -26,36 +26,21 @@ embed_models = [
     "BAAI/bge-large-en"
 ]
 
-# Global variables
+# Global state
 selected_llm_model_name = llm_models[0]
 selected_embed_model_name = embed_models[0]
 vector_index = None
 
-# Initialize the parser
+# Parser setup
 parser = LlamaParse(api_key=os.getenv("LLAMA_INDEX_API"), result_type='markdown')
-file_extractor = {
-    '.pdf': parser,
-    '.docx': parser,
-    '.doc': parser,
-    '.txt': parser,
-    '.csv': parser,
-    '.xlsx': parser,
-    '.pptx': parser,
-    '.html': parser,
-    '.jpg': parser,
-    '.jpeg': parser,
-    '.png': parser,
-    '.webp': parser,
-    '.svg': parser,
-}
+file_extractor = {ext: parser for ext in ['.pdf', '.docx', '.doc', '.txt', '.csv', '.xlsx', '.pptx', '.html', '.jpg', '.jpeg', '.png', '.webp', '.svg']}
 
 def load_files(file_path: str, embed_model_name: str):
+    global vector_index
     try:
-        global vector_index
         document = SimpleDirectoryReader(input_files=[file_path], file_extractor=file_extractor).load_data()
         embed_model = HuggingFaceEmbedding(model_name=embed_model_name)
         vector_index = VectorStoreIndex.from_documents(document, embed_model=embed_model)
-        print(f"Parsing done for {file_path}")
         filename = os.path.basename(file_path)
         return f"Ready to give response on {filename}"
     except Exception as e:
@@ -80,31 +65,46 @@ def respond(message, history):
         )
         if vector_index is not None:
             query_engine = vector_index.as_query_engine(llm=llm)
-            bot_message = query_engine.query(message)
-            print(f"\n{datetime.now()}:{selected_llm_model_name}:: {message} --> {str(bot_message)}\n")
-            return f"{selected_llm_model_name}:\n{str(bot_message)}"
+            bot_message = str(query_engine.query(message))
+            history.append((message, bot_message))
+            print(f"\n{datetime.now()}:{selected_llm_model_name}:: {message} --> {bot_message}\n")
+            return bot_message, history
         else:
-            return "Please upload a file."
+            return "Please upload a file first.", history
     except Exception as e:
-        return f"An error occurred: {e}"
+        return f"An error occurred: {e}", history
 
 def encode_image(image_path):
     with open(image_path, "rb") as image_file:
         return base64.b64encode(image_file.read()).decode('utf-8')
 
+# Encoded logos
 github_logo_encoded = encode_image("Images/github-logo.png")
 linkedin_logo_encoded = encode_image("Images/linkedin-logo.png")
 website_logo_encoded = encode_image("Images/ai-logo.png")
 
+# Markdown placeholders
+description = "### Welcome to **DocBot** - Ask Questions Based on Your Uploaded Documents"
+guide = "> Step 1: Upload\n> Step 2: Select Embedding\n> Step 3: Select LLM\n> Step 4: Ask Questions"
+footer = """
+<center>
+    <a href="https://github.com" target="_blank"><img src="data:image/png;base64,{}" height="30"/></a>&nbsp;
+    <a href="https://linkedin.com" target="_blank"><img src="data:image/png;base64,{}" height="30"/></a>&nbsp;
+    <a href="https://yourwebsite.com" target="_blank"><img src="data:image/png;base64,{}" height="30"/></a>
+</center>
+""".format(github_logo_encoded, linkedin_logo_encoded, website_logo_encoded)
+
+# Gradio UI
 with gr.Blocks(theme=gr.themes.Soft(font=[gr.themes.GoogleFont("Roboto Mono")]), css='footer {visibility: hidden}') as demo:
     gr.Markdown("# DocBot")
     with gr.Tabs():
         with gr.TabItem("Intro"):
-            gr.Markdown(md.description)
+            gr.Markdown(description)
 
         with gr.TabItem("DocBot"):
             with gr.Accordion("=== IMPORTANT: READ ME FIRST ===", open=False):
-                guid = gr.Markdown(md.guide)
+                gr.Markdown(guide)
+
             with gr.Row():
                 with gr.Column(scale=1):
                     file_input = gr.File(file_count="single", type='filepath', label="Step-1: Upload document")
@@ -114,20 +114,24 @@ with gr.Blocks(theme=gr.themes.Soft(font=[gr.themes.GoogleFont("Roboto Mono")]),
                         clear = gr.ClearButton()
                     output = gr.Text(label='Vector Index')
                     llm_model_dropdown = gr.Dropdown(llm_models, label="Step-3: Select LLM", interactive=True)
-                    model_selected_output = gr.Text(label="Model selected")  # FIXED OUTPUT COMPONENT
+                    model_selected_output = gr.Text(label="Model selected")
+
                 with gr.Column(scale=3):
-                    gr.ChatInterface(
-                        fn=respond,
-                        chatbot=gr.Chatbot(height=500),
-                        theme="soft",
-                        textbox=gr.Textbox(placeholder="Step-4: Ask me questions on the uploaded document!", container=False)
-                    )
-    gr.HTML(md.footer.format(github_logo_encoded, linkedin_logo_encoded, website_logo_encoded))
-
-    # Event bindings
-    llm_model_dropdown.change(fn=set_llm_model, inputs=llm_model_dropdown, outputs=model_selected_output)
-    btn.click(fn=load_files, inputs=[file_input, embed_model_dropdown], outputs=output)
-    clear.click(lambda: [None] * 3, outputs=[file_input, embed_model_dropdown, output])
+                    chatbot_ui = gr.Chatbot(height=500)
+                    message = gr.Textbox(placeholder="Step-4: Ask me questions on the uploaded document!", container=False)
+                    submit_btn = gr.Button("Send")
+
+            # Bind logic
+            llm_model_dropdown.change(fn=set_llm_model, inputs=llm_model_dropdown, outputs=model_selected_output)
+            btn.click(fn=load_files, inputs=[file_input, embed_model_dropdown], outputs=output)
+            clear.click(lambda: [None] * 3, outputs=[file_input, embed_model_dropdown, output])
+
+            # Chat logic
+            state = gr.State([])
+            submit_btn.click(fn=respond, inputs=[message, state], outputs=[chatbot_ui, state])
+            message.submit(fn=respond, inputs=[message, state], outputs=[chatbot_ui, state])
+
+    gr.HTML(footer)
 
 if __name__ == "__main__":
-    demo.launch(share=True)
+    demo.launch(share=True)