app.py

import os
import gradio as gr
from gradio import ChatMessage
from typing import Iterator
import google.generativeai as genai
import time # Import time module for potential debugging/delay

# get Gemini API Key from the environ variable
GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
genai.configure(api_key=GEMINI_API_KEY)

# we will be using the Gemini 2.0 Flash model with Thinking capabilities
model = genai.GenerativeModel("gemini-2.0-flash-thinking-exp-1219")


def format_chat_history(messages: list) -> list:
    """
    Formats the chat history into a structure Gemini can understand
    """
    formatted_history = []
    for message in messages:
        # Skip thinking messages (messages with metadata)
        if not (message.get("role") == "assistant" and "metadata" in message):
            formatted_history.append({
                "role": "user" if message.get("role") == "user" else "assistant",
                "parts": [message.get("content", "")]
            })
    return formatted_history

def stream_gemini_response(message_input: str|gr.File, messages: list) -> Iterator[list]:
    """
    Streams thoughts and response with conversation history support, handling text or file input.
    """
    user_message = ""
    input_file = None

    if isinstance(message_input, str):
        user_message = message_input
        print(f"\n=== New Request (Text) ===")
        print(f"User message: {user_message}")
        if not user_message.strip(): # Robust check: if text message is empty or whitespace
            messages.append(ChatMessage(role="assistant", content="Please provide a non-empty text message or upload a file.")) # More specific message
            yield messages
            return

    elif isinstance(message_input, gr.File): #gr.File directly should be used with newer gradio versions (v4+)
        input_file = message_input.name # Access the temporary file path
        file_type = message_input.original_name.split('.')[-1].lower() #Get original filename's extension
        print(f"\n=== New Request (File) ===")
        print(f"File uploaded: {input_file}, type: {file_type}")

        try:
            with open(input_file, "rb") as f: #Open file in binary mode for universal handling
                file_data = f.read()

            if file_type in ['png', 'jpg', 'jpeg', 'gif']: #Example Image Types - expand as needed
                user_message = {"inline_data": {"mime_type": f"image/{file_type}", "data": file_data}} #Prepare image part for Gemini
            elif file_type == 'csv':
                user_message = {"inline_data": {"mime_type": "text/csv", "data": file_data}} #Prepare csv part

        except Exception as e:
            print(f"Error reading file: {e}")
            messages.append(ChatMessage(role="assistant", content=f"Error reading file: {e}"))
            yield messages
            return
    else:
        messages.append(ChatMessage(role="assistant", content="Sorry, I cannot understand this input format. Please use text or upload a valid file.")) # More informative error
        yield messages
        return


    try:
        # Format chat history for Gemini
        chat_history = format_chat_history(messages)

        # Initialize Gemini chat
        chat = model.start_chat(history=chat_history)
        response = chat.send_message(user_message, stream=True) #Send the message part as is

        # Initialize buffers and flags - same as before
        thought_buffer = ""
        response_buffer = ""
        thinking_complete = False

        # Add initial thinking message - same as before
        messages.append(
            ChatMessage(
                role="assistant",
                content="",
                metadata={"title": "⚙️ Thinking: *The thoughts produced by the model are experimental"}
            )
        )

        for chunk in response: #streaming logic - same as before
            parts = chunk.candidates[0].content.parts
            current_chunk = parts[0].text

            if len(parts) == 2 and not thinking_complete:
                # Complete thought and start response
                thought_buffer += current_chunk
                print(f"\n=== Complete Thought ===\n{thought_buffer}")

                messages[-1] = ChatMessage(
                    role="assistant",
                    content=thought_buffer,
                    metadata={"title": "⚙️ Thinking: *The thoughts produced by the model are experimental"}
                )
                yield messages

                # Start response
                response_buffer = parts[1].text
                print(f"\n=== Starting Response ===\n{response_buffer}")

                messages.append(
                    ChatMessage(
                        role="assistant",
                        content=response_buffer
                    )
                )
                thinking_complete = True

            elif thinking_complete:
                # Stream response
                response_buffer += current_chunk
                print(f"\n=== Response Chunk ===\n{current_chunk}")

                messages[-1] = ChatMessage(
                    role="assistant",
                    content=response_buffer
                )

            else:
                # Stream thinking
                thought_buffer += current_chunk
                print(f"\n=== Thinking Chunk ===\n{current_chunk}")

                messages[-1] = ChatMessage(
                    role="assistant",
                    content=thought_buffer,
                    metadata={"title": "⚙️ Thinking: *The thoughts produced by the model are experimental"}
                )
            #time.sleep(0.05) #Optional: Uncomment this line to add a slight delay for debugging/visualization of streaming. Remove for final version

            yield messages

        print(f"\n=== Final Response ===\n{response_buffer}")

    except Exception as e:
        print(f"\n=== Error ===\n{str(e)}")
        messages.append(
            ChatMessage(
                role="assistant",
                content=f"I apologize, but I encountered an error: {str(e)}"
            )
        )
        yield messages

def user_message(message_text, file_upload, history: list) -> tuple[str, None, list]:
    """Adds user message to chat history"""
    msg = message_text if message_text else file_upload
    history.append(ChatMessage(role="user", content=msg if isinstance(msg, str) else msg.name)) #Store message or filename in history.
    return "", None, history #clear both input fields

# Create the Gradio interface
with gr.Blocks(theme=gr.themes.Soft(primary_hue="teal", secondary_hue="slate", neutral_hue="neutral")) as demo:
    gr.Markdown("# Gemini 2.0 Flash 'Thinking' Chatbot 💭")

    chatbot = gr.Chatbot(
        type="messages",
        label="Gemini2.0 'Thinking' Chatbot (Streaming Output)", #Label now indicates streaming
        render_markdown=True,
        scale=1,
        avatar_images=(None,"https://lh3.googleusercontent.com/oxz0sUBF0iYoN4VvhqWTmux-cxfD1rxuYkuFEfm1SFaseXEsjjE4Je_C_V3UQPuJ87sImQK3HfQ3RXiaRnQetjaZbjJJUkiPL5jFJ1WRl5FKJZYibUA=w214-h214-n-nu")
    )

    with gr.Row(equal_height=True):
        input_box = gr.Textbox(
            lines=1,
            label="Chat Message",
            placeholder="Type your message here...",
            scale=3
        )
        file_upload = gr.File(label="Upload File", file_types=["image", ".csv"], scale=2) # Allow image and CSV files

        clear_button = gr.Button("Clear Chat", scale=1)

    # Add example prompts
    example_prompts = [
        ["Write a short poem about the sunset."],
        ["Explain the theory of relativity in simple terms."],
        ["If a train leaves Chicago at 6am traveling at 60mph, and another train leaves New York at 8am traveling at 80mph, at what time will they meet?"],
        ["Summarize the plot of Hamlet."],
        ["Write a haiku about a cat."]
    ]

    gr.Examples(
        examples=example_prompts,
        inputs=[input_box],
        label="Examples: Get Gemini to show its thinking process with these prompts!",
        examples_per_page=5 # Adjust as needed
    )


    # Set up event handlers
    msg_store = gr.State("")  # Store for preserving user message


    input_box.submit(
        user_message,
        inputs=[input_box, file_upload, chatbot],
        outputs=[input_box, file_upload, chatbot],
        queue=False
    ).then(
        stream_gemini_response,
        inputs=[input_box, chatbot], # Input either from text box or file, logic inside stream_gemini_response
        outputs=chatbot
    )

    file_upload.upload(
        user_message,
        inputs=[input_box, file_upload, chatbot], # even textbox is input here so clearing both will work
        outputs=[input_box, file_upload, chatbot],
        queue=False
    ).then(
        stream_gemini_response,
        inputs=[file_upload, chatbot], # Input is now the uploaded file.
        outputs=chatbot
    )


    clear_button.click(
        lambda: ([], "", ""),
        outputs=[chatbot, input_box, msg_store],
        queue=False
    )

    gr.Markdown(  # Description moved to the bottom
        """
        <br><br><br>  <!-- Add some vertical space -->
        ---
        ### About this Chatbot
        This chatbot demonstrates the experimental 'thinking' capability of the **Gemini 2.0 Flash** model.
        You can observe the model's thought process as it generates responses, displayed with the "⚙️ Thinking" prefix.

        **Try out the example prompts below to see Gemini in action!**

        **Key Features:**
        *   Powered by Google's **Gemini 2.0 Flash** model.
        *   Shows the model's **thoughts** before the final answer (experimental feature).
        *   Supports **conversation history** for multi-turn chats.
        *   Supports **Image and CSV file uploads** for analysis.
        *   Uses **streaming** for a more interactive experience.
        **Instructions:**
        1.  Type your message in the input box or Upload a file below.
        2.  Press Enter/Submit or Upload to send.
        3.  Observe the chatbot's "Thinking" process followed by the final response.
        4.  Use the "Clear Chat" button to start a new conversation.

        *Please note*: The 'thinking' feature is experimental and the quality of thoughts may vary. File analysis capabilities may be limited depending on the model's experimental features.
        """
    )


# Launch the interface
if __name__ == "__main__":
    demo.launch(debug=True)