app.py

# ==============================================================================
# Aura Mind Glow - Main Application (Refactored)
# ==============================================================================
"""
This script launches the Aura Mind Glow application, now with multiple modes.
"""

# --- Step 0: Essential Imports ---
import gradio as gr
from PIL import Image
import os
import warnings
import socket
import tempfile
import json # New import

# Suppress potential warnings for a cleaner console
warnings.filterwarnings("ignore")
os.environ["TORCH_COMPILE_DISABLE"] = "1"  # Ensure torch compile is off

# --- Step 1: Import Core Components from Modules ---
from vision_model import load_vision_model
from knowledge_base import KnowledgeBase
from agent_setup import initialize_adk
from google.genai import types
from story_generator import create_story_prompt_from_pdf, generate_video_from_prompt
from langchain_huggingface import HuggingFaceEndpoint
from bigquery_uploader import upload_diagnosis_to_bigquery

print("✅ All libraries imported successfully.")

# --- Step 2: Global Initialization ---
# This expensive setup runs only ONCE when the application starts.

print("Performing initial setup...")
VISION_MODEL, PROCESSOR = load_vision_model()
KB = KnowledgeBase()
RETRIEVER = KB # The retriever is now the KB itself

# Initialize ADK components for Connected Mode
adk_components = initialize_adk(VISION_MODEL, PROCESSOR, RETRIEVER)
ADK_RUNNER = adk_components["runner"] if adk_components else None
DIAGNOSIS_TOOL = adk_components["diagnosis_tool"] if adk_components else None
REMEDY_TOOL = adk_components["remedy_tool"] if adk_components else None
SESSION_SERVICE = adk_components["session_service"] if adk_components else None

# Initialize a separate LLM for the Story Generator
STORY_LLM = None
if os.environ.get("HF_TOKEN"):
    try:
        STORY_LLM = HuggingFaceEndpoint(
            repo_id="HuggingFaceH4/zephyr-7b-beta",
            huggingfacehub_api_token=os.environ.get("HF_TOKEN"),
            max_new_tokens=150,
            temperature=0.4,
        )
        print("✅ Story Generator LLM initialized successfully.")
    except Exception as e:
        print(f"❌ Could not initialize Story Generator LLM: {e}")
else:
    print("❌ HF_TOKEN not found. Story Generator Mode will be disabled.")


# --- Step 3: Define Gradio UIs ---

def create_field_mode_ui():
    """Creates the Gradio UI for the offline Field Mode."""
    def get_diagnosis_and_remedy(uploaded_image: Image.Image) -> str:
        if uploaded_image is None:
            return "Please upload an image of a maize plant first."
        if RETRIEVER is None:
            raise gr.Error("Knowledge base is not loaded. Cannot find remedy. Check logs.")

        temp_file_path = None
        try:
            with tempfile.NamedTemporaryFile(delete=False, suffix=".png") as temp_file:
                uploaded_image.save(temp_file.name)
                temp_file_path = temp_file.name

            diagnosis = DIAGNOSIS_TOOL(temp_file_path)
            print(f"Diagnosis received: {diagnosis}")

            if "Could not parse" in diagnosis:
                return f"Sorry, I couldn't identify the condition from the image. Raw output: {diagnosis}"

            search_query = diagnosis
            report_title = diagnosis

            results = KB.search(search_query)
            
            if not results:
                remedy = "No remedy found in the local knowledge base."
            else:
                remedy = results[0]['content']

            final_response = f"""
            ## Diagnosis Report

            **Condition Identified:**
            ### {report_title}

            ---

            ## Suggested Remedy

            {remedy}
            """
            print("Workflow complete. Returning response.")

            # Prepare data for BigQuery upload
            diagnosis_data = {
                "ai_diagnosis": report_title,
                "recommended_action": remedy,
                "confidence_score": None, # Placeholder, as confidence score is not calculated here
                "farmer_id": "unknown", # Placeholder
                "gps_latitude": None, # Placeholder
                "gps_longitude": None, # Placeholder
                "crop_type": "Maize", # Assuming maize for this app
                "crop_variety": None, # Placeholder
                "farmer_feedback": None, # Placeholder
                "treatment_applied": None, # Placeholder
                "outcome_image_id": None, # Placeholder
            }
            upload_diagnosis_to_bigquery(diagnosis_data)

            return final_response

        except Exception as e:
            print(f"An error occurred during the analysis workflow: {e}")
            raise gr.Error(f"An unexpected error occurred: {e}")
        finally:
            if temp_file_path and os.path.exists(temp_file_path):
                os.remove(temp_file_path)

    # ... (the rest of your Gradio Interface definition remains the same)
    css = """
    footer {visibility: hidden !important;}
    .gradio-container {font-family: 'IBM Plex Sans', sans-serif;}
    """

    return gr.Interface(
        fn=get_diagnosis_and_remedy,
        inputs=gr.Image(type="pil", label="Upload Maize Plant Image", sources=["upload", "webcam"]),
        outputs=gr.Markdown(label="Diagnosis and Remedy Report", value="The report will appear here..."),
        title="🌽 Aura Mind Glow: Field Mode (Offline)",
        description="**A 100% Offline-Capable Farming Assistant.** Upload an image of a maize plant. The AI will diagnose its condition and retrieve a treatment plan from its local knowledge base.",
        article="<p style='text-align: center;'>Built with Unsloth, LangChain, and Gradio. Version 2.1</p>",
        allow_flagging="never",
        theme=gr.themes.Soft(primary_hue="teal", secondary_hue="orange"),
        css=css
    )

def create_connected_mode_ui():
    """Creates the Gradio UI for the online Connected Mode."""
    # ... (This function remains unchanged) ...
    with gr.Blocks(theme=gr.themes.Soft(primary_hue="green", secondary_hue="lime")) as demo:
        gr.Markdown("# 🌽 Aura Mind Glow: Connected Mode 🤖")
        gr.Markdown("I am an AI farming assistant. Upload an image and ask for a diagnosis and remedy.")

        chatbot = gr.Chatbot(height=600)
        msg = gr.MultimodalTextbox(file_types=["image"], label="Ask a question and/or upload an image...")

        async def respond(chat_input, history):
            user_id = "default_user"  # In a real app, this would be unique per user
            if not SESSION_SERVICE or not ADK_RUNNER:
                history.append((chat_input["text"], "Connected mode is not available. Check logs."))
                yield history, gr.MultimodalTextbox(value=None)
                return

            session = await SESSION_SERVICE.create_session(user_id=user_id, app_name="AuraMindGlow")

            files = chat_input.get("files", [])
            text = chat_input.get("text", "")

            if not files:
                history.append([text, "Please upload an image for diagnosis."])
                yield history, gr.MultimodalTextbox(value=None)
                return

            # Create the prompt for the ADK agent
            with open(files[0], 'rb') as f:
                image_data = f.read()
            image_part = types.Part(
                inline_data=types.Blob(
                    mime_type='image/png',
                    data=image_data
                )
            )
            text_part = types.Part(text=text or "Diagnose this plant and provide a remedy.")
            prompt = types.Content(parts=[image_part, text_part], role="user")

            # Stream the response from the agent
            bot_message = ""
            history.append([(files[0],), bot_message])

            try:
                async for event in ADK_RUNNER.run_async(
                    session_id=session.id, user_id=user_id, new_message=prompt
                ):
                    if event.is_llm_response_chunk() and event.content.parts:
                        bot_message += event.content.parts[0].text
                        history[-1] = ((files[0],), bot_message)
                        yield history, gr.MultimodalTextbox(value=None)
                    elif event.is_final_response() and event.content.parts:
                        bot_message = event.content.parts[0].text
                        history[-1] = ((files[0],), bot_message)
                        yield history, gr.MultimodalTextbox(value=None)

            except Exception as e:
                print(f"Error during agent execution: {e}")
                history[-1] = ((files[0],), f"An error occurred: {e}")
                yield history, gr.MultimodalTextbox(value=None)


        msg.submit(respond, [msg, chatbot], [chatbot, msg])

    return demo

def create_document_analysis_ui():
    """Creates the Gradio UI for the Document Analysis Mode."""
    with gr.Blocks(theme=gr.themes.Soft(primary_hue="purple", secondary_hue="pink")) as demo:
        gr.Markdown("# 🌽 Aura Mind Glow: Document Analysis Mode 📄")
        gr.Markdown("Upload a PDF or a spreadsheet and ask questions about its content.")

        with gr.Row():
            with gr.Column(scale=1):
                doc_input = gr.File(label="Upload Document", file_types=[".pdf", ".csv"])
                query_input = gr.Textbox(label="Ask a question about the document")
                submit_btn = gr.Button("Analyze and Query")
            with gr.Column(scale=2):
                answer_output = gr.Textbox(label="Answer", interactive=False, lines=10)
                status_output = gr.Textbox(label="Status", interactive=False, lines=3)

        def analysis_process(doc, query):
            if doc is None:
                yield "Please upload a document to begin.", ""
                return

            file_path = doc.name
            file_ext = os.path.splitext(file_path)[1].lower()

            if file_ext == ".pdf":
                yield "Analyzing PDF...", ""
                chain, vector_store = analyze_pdf(file_path)
                if chain and vector_store:
                    yield "PDF analyzed successfully. Now querying...", ""
                    answer = query_pdf(chain, vector_store, query)
                    yield answer, "Query complete."
                else:
                    yield "Failed to analyze PDF.", "Error"
            elif file_ext == ".csv":
                yield "Analyzing spreadsheet...", ""
                agent = analyze_spreadsheet(file_path)
                if agent:
                    yield "Spreadsheet analyzed successfully. Now querying...", ""
                    answer = query_spreadsheet(agent, query)
                    yield answer, "Query complete."
                else:
                    yield "Failed to analyze spreadsheet.", "Error"
            else:
                yield "Unsupported file type. Please upload a PDF or a CSV file.", "Error"

        submit_btn.click(
            analysis_process,
            inputs=[doc_input, query_input],
            outputs=[answer_output, status_output]
        )
    return demo


def create_story_mode_ui():
    """Creates the Gradio UI for the Farmer's Story Mode."""
    with gr.Blocks(theme=gr.themes.Soft(primary_hue="blue", secondary_hue="yellow")) as demo:
        gr.Markdown("# 🌽 Aura Mind Glow: Farmer's Story Mode 🎬")
        gr.Markdown("Create a short video story from your farm documents. Upload a PDF, describe the mood, and let the AI create a visual story.")

        with gr.Row():
            with gr.Column(scale=1):
                pdf_input = gr.File(label="Upload PDF", file_types=[".pdf"])
                image_input = gr.Image(type="filepath", label="Optional: Upload a Starting Image")
                user_prompt_input = gr.Textbox(label="Describe the video's tone or theme", placeholder="e.g., hopeful, a look back at a tough season, etc.")
                submit_btn = gr.Button("Generate Video Story")
            with gr.Column(scale=2):
                video_output = gr.Video(label="Generated Video Story")
                status_output = gr.Textbox(label="Status", interactive=False, lines=3)

        def story_generation_process(pdf, image, user_prompt):
            if pdf is None:
                yield None, "Please upload a PDF document to begin."
                return

            yield None, "Step 1: Reading PDF and generating creative prompt..."

            creative_prompt = create_story_prompt_from_pdf(pdf.name, user_prompt, STORY_LLM)

            if "Error" in creative_prompt:
                yield None, creative_prompt
                return

            yield None, f"Step 2: Generating video with prompt: '{creative_prompt[:100]}...' (This may take several minutes)"

            video_path = generate_video_from_prompt(creative_prompt, image)

            if "Error" in video_path:
                yield None, video_path
                return

            yield video_path, "Video generation complete!"

        submit_btn.click(
            story_generation_process,
            inputs=[pdf_input, image_input, user_prompt_input],
            outputs=[video_output, status_output]
        )
    return demo

def create_settings_ui():
    """Creates the Gradio UI for the Settings tab."""
    with gr.Blocks(theme=gr.themes.Soft(primary_hue="gray", secondary_hue="blue")) as demo:
        gr.Markdown("# ⚙️ Settings & Data Management")
        gr.Markdown("Manage application settings and data synchronization.")

        with gr.Row():
            with gr.Column():
                sync_btn = gr.Button("☁️ Sync Local Data to BigQuery Cloud")
                status_output = gr.Textbox(label="Sync Status", interactive=False, lines=5)

        def sync_data_to_cloud():
            yield "Local data sync is no longer required as diagnoses are uploaded directly to BigQuery."

        sync_btn.click(
            sync_data_to_cloud,
            inputs=[],
            outputs=[status_output]
        )
    return demo

def create_kb_management_ui():
    """Creates the Gradio UI for managing the knowledge base."""
    with gr.Blocks(theme=gr.themes.Soft(primary_hue="indigo", secondary_hue="purple")) as demo:
        gr.Markdown("# 📚 Knowledge Base Management")
        gr.Markdown("Manage the local, encrypted knowledge base.")

        with gr.Row():
            with gr.Column():
                gr.Markdown("### Rebuild Knowledge Base")
                rebuild_btn = gr.Button("Rebuild from Source Files")
                rebuild_status = gr.Textbox(label="Status", interactive=False)

            with gr.Column():
                gr.Markdown("### Add PDF to Knowledge Base")
                pdf_input = gr.File(label="Upload PDF", file_types=[".pdf"])
                ingest_btn = gr.Button("Ingest PDF")
                ingest_status = gr.Textbox(label="Status", interactive=False)

        def rebuild_kb():
            yield "Rebuilding knowledge base..."
            try:
                KB.rebuild_from_default_files() # Call the new method to rebuild from default files
                yield "Knowledge base rebuilt successfully."
            except Exception as e:
                yield f"Error rebuilding knowledge base: {e}"

        def ingest_pdf(pdf):
            if pdf is None:
                return "Please upload a PDF file."
            yield "Ingesting PDF..."
            try:
                KB.ingest_pdf(pdf.name, os.path.basename(pdf.name))
                yield f"Successfully ingested {os.path.basename(pdf.name)}."
            except Exception as e:
                yield f"Error ingesting PDF: {e}"

        rebuild_btn.click(rebuild_kb, outputs=[rebuild_status])
        ingest_btn.click(ingest_pdf, inputs=[pdf_input], outputs=[ingest_status])

    return demo

# --- Step 4: App Launcher ---

def check_internet_connection(host="8.8.8.8", port=53, timeout=3):
    """Check for internet connectivity."""
    try:
        socket.setdefaulttimeout(timeout)
        socket.socket(socket.AF_INET, socket.SOCK_STREAM).connect((host, port))
        return True
    except socket.error:
        return False


def _setup_gcp_credentials():
    """
    Sets up Google Cloud credentials from an environment variable.
    If GOOGLE_APPLICATION_CREDENTIALS contains a JSON string, it writes
    it to a temporary file and updates the environment variable to point
    to that file. This is useful for environments like Hugging Face Spaces
    where secrets are provided as string values.
    """
    gcp_credentials_json = os.environ.get("GOOGLE_APPLICATION_CREDENTIALS")
    if gcp_credentials_json and gcp_credentials_json.strip().startswith("{"):
        try:
            # Validate JSON
            credentials_dict = json.loads(gcp_credentials_json)
            
            # Create a temporary file to store the credentials
            temp_dir = tempfile.gettempdir()
            temp_file_path = os.path.join(temp_dir, "gcp_credentials.json")
            
            with open(temp_file_path, "w") as f:
                json.dump(credentials_dict, f)
            
            os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = temp_file_path
            print(f"✅ Google Cloud credentials successfully set from environment variable to: {temp_file_path}")
        except json.JSONDecodeError as e:
            print(f"❌ Error decoding GOOGLE_APPLICATION_CREDENTIALS JSON: {e}")
        except Exception as e:
            print(f"❌ An unexpected error occurred while setting up GCP credentials: {e}")
    else:
        print("ℹ️ GOOGLE_APPLICATION_CREDENTIALS environment variable not found or not a JSON string. Assuming credentials are set via other means (e.g., gcloud CLI, default ADC).")

if __name__ == "__main__":
    _setup_gcp_credentials()
    # No local database initialization needed

    field_mode_ui = create_field_mode_ui()
    interface_list = [field_mode_ui]
    tab_titles = ["Field Mode (Offline)"]

    # Conditionally add modes that require an internet connection
    if check_internet_connection():
        if ADK_RUNNER:
            connected_mode_ui = create_connected_mode_ui()
            interface_list.append(connected_mode_ui)
            tab_titles.append("Connected Mode")
        else:
            print("⚠️ Connected Mode disabled: ADK components not initialized.")

        if STORY_LLM:
            story_mode_ui = create_story_mode_ui()
            interface_list.append(story_mode_ui)
            tab_titles.append("Farmer's Story Mode")
        else:
            print("⚠️ Farmer's Story Mode disabled: Story LLM not initialized.")

        # Add the new Document Analysis UI
        document_analysis_ui = create_document_analysis_ui()
        interface_list.append(document_analysis_ui)
        tab_titles.append("Document Analysis")
        
        # Add the Settings UI
        settings_ui = create_settings_ui()
        interface_list.append(settings_ui)
        tab_titles.append("Settings")

        # Add the Knowledge Base Management UI
        kb_management_ui = create_kb_management_ui()
        interface_list.append(kb_management_ui)
        tab_titles.append("Knowledge Base")

    else:
        print("❌ No internet connection. Launching in Offline Mode only.")

    # Launch the appropriate UI
    if len(interface_list) > 1:
        ui = gr.TabbedInterface(interface_list, tab_titles)
    else:
        ui = field_mode_ui

    ui.launch(share=True, debug=True)