app.py

import gradio as gr
import shutil
import os
import subprocess
from src.mcp.tools import letter_counter
from src.mcp import video_tools
from src.llm import llm
from gradio.oauth import OAuthToken

def is_ffmpeg_installed():
    """Checks if FFmpeg is installed and accessible in the system's PATH."""
    return shutil.which("ffmpeg") is not None

def update_ffmpeg_status():
    """Creates a colored status message indicating if FFmpeg is installed."""
    if is_ffmpeg_installed():
        return "<p style='color:green; font-weight:bold;'>ffmpeg is installed</p>", True
    return "<p style='color:red; font-weight:bold;'>ffmpeg is not installed</p>", False

def handle_video_upload(file_obj, ffmpeg_installed):
    """
    Validates an uploaded MP4 file, saves it to a temporary directory,
    and controls the visibility of UI components based on validation success.
    """
    if not ffmpeg_installed:
        return None, "Cannot process video: FFmpeg is not installed.", None, gr.update(visible=False)
    if not file_obj:
        return None, "Please upload a file.", None, gr.update(visible=False)

    tmp_dir = "tmp"
    if not os.path.exists(tmp_dir):
        os.makedirs(tmp_dir)

    # Basic check for file extension
    if not file_obj.name.lower().endswith('.mp4'):
        return None, "File must be an .mp4 file.", None, gr.update(visible=False)

    # More robust check using ffprobe
    try:
        subprocess.run(
            ["ffprobe", "-v", "error", "-show_format", "-show_streams", file_obj.name],
            capture_output=True, text=True, check=True
        )
    except (subprocess.CalledProcessError, FileNotFoundError):
        return None, "Could not validate file. Ensure it's a valid MP4.", None, gr.update(visible=False)

    file_path = os.path.join(tmp_dir, os.path.basename(file_obj.name))
    shutil.copy(file_obj.name, file_path)
    
    return file_path, "File uploaded successfully!", file_path, gr.update(visible=True)

def clear_previous_outputs():
    """Clears all video-related outputs to ensure a clean state for new uploads."""
    return (
        None, # video_output
        "",   # upload_status_text
        None, # uploaded_video_path_state
        gr.update(visible=False), # video_tools_group
        None, # first_frame_img
        None, # last_frame_img
        None, # llm_media_output
        "",   # video_command_status
        "",   # llm_debug_output
        "",   # llm_raw_response (now a textbox, clear with empty string)
    )

# --- Gradio UI ---
with gr.Blocks() as demo:
    # --- State Variables ---
    ffmpeg_installed_state = gr.State(False)
    uploaded_video_path_state = gr.State("")

    with gr.Tabs():
        # --- Setup & Video Tab ---
        with gr.Tab("Setup & Video"):
            gr.Markdown("## System Status")
            with gr.Row():
                check_ffmpeg_btn = gr.Button("Check FFmpeg Status")
                status_text = gr.Markdown("Status will be checked on load.")

            gr.Markdown("---")
            gr.Markdown("## Video Tools")
            with gr.Row():
                with gr.Column(scale=1):
                    gr.Markdown("### Upload Video")
                    file_input = gr.File(label="Upload MP4", file_types=[".mp4"])
                    video_output = gr.Video(label="Preview", interactive=False, height="50vh")
                    upload_status_text = gr.Textbox(label="Upload Status", interactive=False)
                with gr.Column(scale=1):
                    # This empty column will take up the other 50% of the space
                    pass
        
        with gr.Tab("Debug"):
            with gr.Column(scale=2, visible=False) as video_tools_group:
                gr.Markdown("### Manual Frame Extraction")
                with gr.Row():
                    get_first_frame_btn = gr.Button("Get First Frame")
                    get_last_frame_btn = gr.Button("Get Last Frame")
                with gr.Row():
                    first_frame_img = gr.Image(label="First Frame", type="filepath", interactive=False)
                    last_frame_img = gr.Image(label="Last Frame", type="filepath", interactive=False)

        with gr.Tab("LLM Video Commands"):
            gr.Markdown("## Test MCP Tool Calls with an LLM")
            with gr.Row():
                with gr.Column(scale=1):
                    gr.Markdown("### Video Command Center")
                    video_prompt_textbox = gr.Textbox(label="Video Command Prompt", placeholder="e.g., 'Get me the first frame of the video'")
                    process_prompt_btn = gr.Button("Run Command")
                    video_command_status = gr.Textbox(label="LLM Status", interactive=False)

                    with gr.Accordion("Debug Info", open=False):
                        llm_debug_output = gr.Textbox(label="Debug Info", lines=10, interactive=False)
                        llm_raw_response = gr.Textbox(label="Raw LLM Response", lines=10, interactive=True)

                with gr.Column(scale=1):
                    gr.Markdown("### LLM Result")
                    llm_media_output = gr.Image(label="Output", type="filepath", interactive=False)

        # --- LLM Configuration Tab ---
        with gr.Tab("LLM Configuration"):
            gr.Markdown("## Configure LLM")
            llm_provider = gr.Radio(
                ["Ollama", "Hugging Face"], 
                value="Ollama", 
                label="LLM Provider",
                info="Select the LLM provider to use."
            )

            # Shared state for model name across providers
            model_name_state = gr.State("")

            with gr.Group(visible=True) as ollama_config:
                gr.Markdown("### Ollama Configuration")
                with gr.Row():
                    ollama_url_textbox = gr.Textbox(
                        placeholder="http://localhost:11434",
                        label="Ollama Endpoint URL",
                        interactive=True,
                        elem_id="ollama_url"
                    )
                    check_endpoint_btn = gr.Button("Check Endpoint")
                ollama_status_textbox = gr.Textbox(label="Status", interactive=False)
                
                with gr.Row():
                    ollama_model_dropdown = gr.Dropdown(
                        label="Select a Model",
                        interactive=True,
                        visible=False,
                        elem_id="ollama_model_dropdown"
                    )
                    set_preferred_btn = gr.Button("Set as Preferred", visible=False)

                preferred_llm_display = gr.Textbox(label="Preferred Model Status", interactive=False)
                
            with gr.Group(visible=False) as hf_config:
                gr.Markdown("### Hugging Face Configuration")
                gr.Markdown(f"**Model ID:** `{llm.HF_MODEL_ID}` (hardcoded).")
                gr.Markdown("To use this provider, please log in with your Hugging Face account. API calls will use your personal token.")
                login_button = gr.LoginButton()

        # Hidden tab for exposing functions as MCP tools
        with gr.Tab("MCP Tools", visible=False):
            gr.Markdown("## Tools for MCP Server")
            
            # Input for video path for MCP tools
            mcp_video_path_input = gr.Textbox(label="Video Path for MCP")
            
            # For letter_counter
            lc_word_input = gr.Textbox(label="Word")
            lc_letter_input = gr.Textbox(label="Letter")
            lc_output = gr.Number(label="Count")
            lc_btn = gr.Button("Count Letters")
            
            # For video tools
            mcp_get_first_frame_btn = gr.Button("MCP Get First Frame")
            mcp_get_last_frame_btn = gr.Button("MCP Get Last Frame")
            mcp_convert_to_gif_btn = gr.Button("MCP Convert to GIF")

    # --- Event Handlers ---

    # --- LLM Provider Change ---
    def update_provider_visibility(provider):
        if provider == "Ollama":
            return gr.update(visible=True), gr.update(visible=False)
        else: # Hugging Face
            return gr.update(visible=False), gr.update(visible=True)

    llm_provider.change(
        fn=update_provider_visibility,
        inputs=llm_provider,
        outputs=[ollama_config, hf_config],
        show_progress=False
    )
    
    # --- System Events ---
    demo.load(
        fn=update_ffmpeg_status,
        outputs=[status_text, ffmpeg_installed_state]
    ).then(
        fn=llm.load_settings,
        outputs=[
            llm_provider,
            ollama_url_textbox,
            model_name_state,
            preferred_llm_display,
        ]
    ).then(
        fn=update_provider_visibility,
        inputs=llm_provider,
        outputs=[ollama_config, hf_config],
        show_progress=False
    ).then(
        fn=llm.check_on_load,
        inputs=[ollama_url_textbox, model_name_state],
        outputs=[ollama_status_textbox, ollama_model_dropdown, set_preferred_btn, ollama_url_textbox, preferred_llm_display]
    )
    
    # --- FFmpeg Events ---
    check_ffmpeg_btn.click(
        fn=update_ffmpeg_status,
        outputs=[status_text, ffmpeg_installed_state]
    )

    file_input.upload(
        fn=clear_previous_outputs,
        outputs=[
            video_output,
            upload_status_text,
            uploaded_video_path_state,
            video_tools_group,
            first_frame_img,
            last_frame_img,
            llm_media_output,
            video_command_status,
            llm_debug_output,
            llm_raw_response
        ]
    ).then(
        fn=handle_video_upload,
        inputs=[file_input, ffmpeg_installed_state],
        outputs=[video_output, upload_status_text, uploaded_video_path_state, video_tools_group]
    )

    get_first_frame_btn.click(
        fn=video_tools.getFirstFrame,
        inputs=uploaded_video_path_state,
        outputs=first_frame_img
    )

    get_last_frame_btn.click(
        fn=video_tools.getLastFrame,
        inputs=uploaded_video_path_state,
        outputs=last_frame_img
    )

    # --- Ollama Events ---
    check_endpoint_btn.click(
        fn=llm.check_ollama_endpoint,
        inputs=[ollama_url_textbox, model_name_state],
        outputs=[ollama_status_textbox, ollama_model_dropdown, set_preferred_btn, ollama_url_textbox]
    )

    set_preferred_btn.click(
        fn=llm.set_preferred_model,
        inputs=[ollama_model_dropdown, ollama_url_textbox],
        outputs=[model_name_state, preferred_llm_display]
    )

    # --- LLM Command Events ---
    process_prompt_btn.click(
        fn=llm.dispatch_video_prompt,
        inputs=[
            llm_provider,
            video_prompt_textbox, 
            uploaded_video_path_state,
            ollama_url_textbox,
            model_name_state,
        ],
        outputs=[
            llm_media_output,
            llm_debug_output,
            video_command_status,
            llm_raw_response
        ]
    )

    # --- MCP Tool Events (Hidden) ---
    lc_btn.click(
        fn=letter_counter,
        inputs=[lc_word_input, lc_letter_input],
        outputs=lc_output,
        api_name="letter_counter"
    )
    
    mcp_get_first_frame_btn.click(
        fn=video_tools.getFirstFrame,
        inputs=[mcp_video_path_input],
        api_name="getFirstFrame"
    )

    mcp_get_last_frame_btn.click(
        fn=video_tools.getLastFrame,
        inputs=[mcp_video_path_input],
        api_name="getLastFrame"
    )

    mcp_convert_to_gif_btn.click(
        fn=video_tools.convert_mp4_to_gif,
        inputs=[mcp_video_path_input],
        api_name="convert_mp4_to_gif"
    )

if __name__ == "__main__":
    demo.launch(mcp_server=True)