import gradio as gr
import pandas as pd
import yfinance as yf
from utils import (
    calculate_technical_indicators,
    generate_trading_signals,
    get_fundamental_data,
    predict_prices,
    create_price_chart,
    create_technical_chart,
    create_prediction_chart,
    calculate_advanced_risk_metrics 
)
import warnings
import numpy as np 

warnings.filterwarnings("ignore")


def analyze_stock(symbol, prediction_days=30):
    try:
        if not symbol.strip():
            raise ValueError("Please enter a valid stock symbol.")

        if not symbol.endswith(".JK"):
            symbol = symbol.upper() + ".JK"

        stock = yf.Ticker(symbol)
        data = stock.history(period="1y", interval="1d") 

        if data.empty:
            raise ValueError("No price data available for this stock.")

        indicators = calculate_technical_indicators(data)
        signals = generate_trading_signals(data, indicators)
        fundamental_info = get_fundamental_data(stock)
        
        risk_metrics = calculate_advanced_risk_metrics(data.copy())

        predictions = predict_prices(data, prediction_days=prediction_days)

        fig_price = create_price_chart(data, indicators)
        fig_technical = create_technical_chart(data, indicators)
        fig_prediction = create_prediction_chart(data, predictions)

        # kalkulasi TP1, TP2, SL yang diperbarui berdasarkan quantiles/range prediksi
        last_price = data['Close'].iloc[-1]
        
        # Dapatkan array prediksi dengan fallback ke array yang berisi harga terakhir
        q05 = predictions.get('values', np.array([last_price]))
        q01 = predictions.get('q01', np.array([last_price * 0.95]))
        q09 = predictions.get('q09', np.array([last_price * 1.05]))
        
        # Robust max/min calculation
        q05_max = np.max(q05) if q05.size > 0 else last_price
        q09_max = np.max(q09) if q09.size > 0 else last_price * 1.05
        q01_min = np.min(q01) if q01.size > 0 else last_price * 0.95
        
        # Target/SL calculation
        tp1 = (last_price + q05_max) / 2
        tp2 = q09_max
        sl = q01_min
        
        if tp1 > tp2:
            tp1, tp2 = tp2, tp1 
        if sl > last_price:
            sl = last_price * 0.95 # Fallback

        predictions["tp1"] = tp1
        predictions["tp2"] = tp2
        predictions["sl"] = sl

        return fundamental_info, indicators, signals, risk_metrics, fig_price, fig_technical, fig_prediction, predictions

    except Exception as e:
        print(f"Error analyzing {symbol}: {e}")
        
        try:
            stock = yf.Ticker(symbol)
            data = stock.history(period="1d", interval="1d")
            last_price = data['Close'].iloc[-1] if not data.empty else 0
        except:
            last_price = 0

        default_tp1 = last_price * 1.01 
        default_tp2 = last_price * 1.02 
        default_sl = last_price * 0.95

        empty_predictions = {
            "high_30d": 0, "low_30d": 0, "change_pct": 0,
            "summary": f"Prediction unavailable. Model error: {e}",
            # Menggunakan list kosong sebagai fallback jika error terjadi
            "q01": [], "q09": [], 
            "tp1": default_tp1, "tp2": default_tp2, "sl": default_sl,
        }
        empty_risk = {"error": "Prediction Model Failed to Load/Run. See console for details."}
        
        # Mengembalikan None untuk output plot Gradio untuk membersihkan plot
        return {}, {}, {}, empty_risk, None, None, None, empty_predictions


def update_analysis(symbol, prediction_days):
    (
        fundamental_info,
        indicators,
        signals,
        risk_metrics,
        fig_price,
        fig_technical,
        fig_prediction,
        predictions,
    ) = analyze_stock(symbol, prediction_days)

    # Cek apakah ada plot yang None (berarti ada error)
    if fig_price is None: 
        error_msg = f"Unable to run AI prediction or fetch data for {symbol.upper()}. Check the model logs for details."
        tp_sl_info = f"<b>TP1:</b> Rp{predictions.get('tp1', 0):,.2f}<br><b>TP2:</b> Rp{predictions.get('tp2', 0):,.2f}<br><b>Stop Loss:</b> Rp{predictions.get('sl', 0):,.2f}<br><br><b>Model Insight:</b><br>{predictions.get('summary', 'Data fetching or model execution failed. Cannot proceed with analysis.')}"
        
        return (
            f"""<div style="color: red; padding: 10px; border: 1px solid red; border-radius: 5px;">{error_msg}</div><br>{tp_sl_info}""",
            None, # fig_price
            None, # fig_technical
            None, # fig_prediction
        )

    # --- AI FORECAST ---
    q01_values = predictions.get('q01', [])
    q09_values = predictions.get('q09', [])

    # FIX: Robust check untuk array NumPy/list kosong
    # Konversi ke NumPy array dan periksa ukuran (.size > 0) untuk menghindari ValueError
    q01_arr = np.array(q01_values)
    q09_arr = np.array(q09_values)

    band_min = float(np.min(q01_arr)) if q01_arr.size > 0 else 0
    band_max = float(np.max(q09_arr)) if q09_arr.size > 0 else 0

    # --- FUNDAMENTALS ---
    fundamentals = f"""
    <h4>COMPANY FUNDAMENTALS</h4>
    <b>Name:</b> {fundamental_info.get('name', 'N/A')} ({symbol.upper()})<br>
    <b>Current Price:</b> Rp{fundamental_info.get('current_price', 0):,.2f}<br>
    <b>Market Cap:</b> {fundamental_info.get('market_cap', 0):,}<br>
    <b>P/E Ratio:</b> {fundamental_info.get('pe_ratio', 0):.2f}<br>
    <b>Dividend Yield:</b> {fundamental_info.get('dividend_yield', 0):.2f}%<br>
    <b>Volume:</b> {fundamental_info.get('volume', 0):,}<br>
    """

    # --- TECHNICAL SIGNAL ---
    details_list = "".join(
        [f"<li>{line.strip()}</li>" for line in signals.get("details", "").split("\n") if line.strip()]
    )

    trading_signal = f"""
    <h4>TECHNICAL SIGNAL SUMMARY</h4>
    <b>Overall Trend:</b> {signals.get('overall', 'N/A')}<br>
    <b>Signal Strength:</b> {signals.get('strength', 0):.2f}%<br>
    <b>Support:</b> Rp{signals.get('support', 0):,.2f}<br>
    <b>Resistance:</b> Rp{signals.get('resistance', 0):,.2f}<br>
    <b>Stop Loss:</b> Rp{signals.get('stop_loss', 0):,.2f}<br><br>
    <b>Detailed Signals:</b>
    <ul style="margin-top: 8px; padding-left: 20px; line-height: 1.6;">
    {details_list}
    </ul>
    """
    
    # --- RISK METRICS ---
    risk_details = ""
    if "error" in risk_metrics:
        risk_details = f"<b style='color: red;'>{risk_metrics['error']}</b>"
    else:
        for key, value in risk_metrics.items():
            risk_details += f"<b>{key.replace('_', ' ')}:</b> {value}<br>"
            
    risk_report = f"""
    <h4>ADVANCED RISK METRICS (1Y HISTORICAL)</h4>
    {risk_details}
    """
    
    # --- AI FORECAST (Lanjutan) ---
    prediction = f"""
    <h4>{prediction_days}-DAY AI FORECAST (CHRONOS-2 + COVARIATES)</h4>
    <b>Predicted Median High:</b> Rp{predictions.get('high_30d', 0):,.2f}<br>
    <b>Predicted Median Low:</b> Rp{predictions.get('low_30d', 0):,.2f}<br>
    <b>Expected Change:</b> {predictions.get('change_pct', 0):.2f}%<br>
    ---
    <h4>RISK ANALYSIS (90% CONFIDENCE)</h4>
    <b>Min. Possible Price (Q0.1):</b> Rp{band_min:,.2f}<br>
    <b>Max. Possible Price (Q0.9):</b> Rp{band_max:,.2f}<br>
    ---
    <b>TP1 (Conservative Target):</b> Rp{predictions.get('tp1', 0):,.2f}<br>
    <b>TP2 (Aggressive Target):</b> Rp{predictions.get('tp2', 0):,.2f}<br>
    <b>Stop Loss (Q0.1 based):</b> Rp{predictions.get('sl', 0):,.2f}<br><br>
    <b>Model Insight:</b><br>{predictions.get('summary', 'No analysis available')}
    """

    # Menggunakan tata letak 4 kolom/panel
    return (
        f"""
        <div style="display: grid; grid-template-columns: repeat(4, 1fr); gap: 16px;">
            <div style="border: 1px solid #ccc; padding: 10px; border-radius: 5px; height: 100%;">{fundamentals}</div>
            <div style="border: 1px solid #ccc; padding: 10px; border-radius: 5px; height: 100%;">{trading_signal}</div>
            <div style="border: 1px solid #ccc; padding: 10px; border-radius: 5px; height: 100%;">{risk_report}</div>
            <div style="border: 1px solid #ccc; padding: 10px; border-radius: 5px; height: 100%;">{prediction}</div>
        </div>
        """,
        fig_price,
        fig_technical,
        fig_prediction,
    )


# --- Gradio Interface ---
with gr.Blocks(
    title="REXPRO FINANCIAL AI DASHBOARD"
) as app:
    gr.Markdown("# REXPRO FINANCIAL AI DASHBOARD")
    gr.Markdown(
        "Comprehensive stock analytics powered by **AI forecasting, advanced risk metrics, and future technical analysis.**"
    )

    with gr.Row():
        symbol = gr.Textbox(
            label="STOCK SYMBOL (IDX)",
            value="BBCA",
            placeholder="Example: BBCA, TLKM, ADRO, BMRI",
            interactive=True,
        )
        prediction_days = gr.Slider(
            label="FORECAST PERIOD (DAYS)",
            minimum=5,
            maximum=60,
            step=5,
            value=30,
            interactive=True,
        )
        analyze_button = gr.Button("RUN ADVANCED ANALYSIS")

    gr.Markdown("---")
    report_section = gr.HTML() 
    gr.Markdown("---")

    with gr.Tab("MARKET CHARTS"):
        with gr.Row():
            price_chart = gr.Plot(label="PRICE & MOVING AVERAGES")
            technical_chart = gr.Plot(label="TECHNICAL INDICATORS OVERVIEW")
        gr.Markdown("---")
        prediction_chart = gr.Plot(label="AI FORECAST & FUTURE TECHNICAL PROJECTION")

    analyze_button.click(
        fn=update_analysis,
        inputs=[symbol, prediction_days],
        outputs=[report_section, price_chart, technical_chart, prediction_chart],
    )

if __name__ == "__main__":
    app.launch(server_name="0.0.0.0", server_port=7860, ssr_mode=True)