app.py

# ==============================================================================
# 🚀 app.py (V61.4 - GEM-Architect: Full UI Integrity Restored)
# ==============================================================================

import os
import sys
import traceback
import asyncio
import gc
import time
import json
import logging 
from datetime import datetime, timedelta
from contextlib import asynccontextmanager, redirect_stdout, redirect_stderr
from io import StringIO
from typing import List, Dict, Any, Optional

from fastapi import FastAPI, HTTPException, BackgroundTasks
import gradio as gr
import pandas as pd
import plotly.graph_objects as go 

# ------------------------------------------------------------------------------
# Logging Setup
# ------------------------------------------------------------------------------
logging.basicConfig(
    level=logging.INFO,
    format="[%(levelname)s] %(message)s", 
    handlers=[logging.StreamHandler(sys.stdout)]
)
logger = logging.getLogger("TitanCore")

# ------------------------------------------------------------------------------
# Imports
# ------------------------------------------------------------------------------
try:
    from r2 import R2Service, INITIAL_CAPITAL 
    from ml_engine.data_manager import DataManager
    from ml_engine.processor import MLProcessor, SystemLimits 
    from whale_monitor.core import EnhancedWhaleMonitor
    from whale_monitor.rpc_manager import AdaptiveRpcManager
    from sentiment_news import NewsFetcher
    from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
    from learning_hub.adaptive_hub import AdaptiveHub
    from trade_manager import TradeManager
    from periodic_tuner import AutoTunerScheduler
    
    try:
        from backtest_engine import run_strategic_optimization_task
        BACKTEST_AVAILABLE = True
    except ImportError:
        BACKTEST_AVAILABLE = False
        
except ImportError as e:
    logger.critical(f"❌ [FATAL ERROR] Failed to import core modules: {e}")
    traceback.print_exc()
    sys.exit(1)

# ------------------------------------------------------------------------------
# Global Context
# ------------------------------------------------------------------------------
r2: R2Service = None
data_manager: DataManager = None
ml_processor: MLProcessor = None
adaptive_hub: AdaptiveHub = None  
trade_manager: TradeManager = None
whale_monitor: EnhancedWhaleMonitor = None
news_fetcher: NewsFetcher = None
senti_analyzer: SentimentIntensityAnalyzer = None
sys_state: 'SystemState' = None
scheduler: AutoTunerScheduler = None

# ------------------------------------------------------------------------------
# State Management
# ------------------------------------------------------------------------------
class SystemState:
    def __init__(self):
        self.ready = False
        self.cycle_running = False
        self.training_running = False 
        self.auto_pilot = True 
        
        self.last_cycle_time: datetime = None
        self.last_cycle_error = None
        self.app_start_time = datetime.now()
        
        self.last_cycle_logs = "System Initializing..."
        self.training_status_msg = "Adaptive Mode: Active" 
        
        self.scan_interval = 60 

    def set_ready(self): 
        self.ready = True
        self.last_cycle_logs = "✅ System Ready. Cybernetic Loop ON."
        logger.info("✅ System State set to READY.")
    
    def set_cycle_start(self): 
        self.cycle_running = True
        self.last_cycle_logs = "🌀 [Cycle START] Scanning Markets..."
        logger.info("🌀 Cycle STARTED.")

    def set_cycle_end(self, error=None, logs=None):
        self.cycle_running = False
        self.last_cycle_time = datetime.now()
        self.last_cycle_error = str(error) if error else None
        
        if logs: 
            self.last_cycle_logs = logs
        elif error:
            self.last_cycle_logs = f"❌ [Cycle ERROR] {error}"
            logger.error(f"Cycle Error: {error}")
        else:
            self.last_cycle_logs = f"✅ [Cycle END] Finished successfully."
            logger.info("✅ Cycle ENDED.")

sys_state = SystemState()

# ------------------------------------------------------------------------------
# Utilities
# ------------------------------------------------------------------------------
def calculate_duration_str(timestamp_str):
    if not timestamp_str: return "--:--:--"
    try:
        if isinstance(timestamp_str, str):
            start_time = datetime.fromisoformat(timestamp_str)
        else: start_time = timestamp_str
        
        diff = datetime.now() - start_time
        total_seconds = int(diff.total_seconds())
        
        days = total_seconds // 86400
        hours = (total_seconds % 86400) // 3600
        minutes = (total_seconds % 3600) // 60
        seconds = total_seconds % 60
        
        if days > 0: return f"{days}d {hours:02}:{minutes:02}:{seconds:02}"
        return f"{hours:02}:{minutes:02}:{seconds:02}"
    except: return "--:--:--"

def format_pnl_split(profit, loss):
    """تنسيق الربحية كجزأين (ربح / خسارة) بألوان"""
    return f"<span style='color:#00ff00'>+${profit:,.0f}</span> / <span style='color:#ff0000'>-${abs(loss):,.0f}</span>"

# ------------------------------------------------------------------------------
# Auto-Pilot Daemon
# ------------------------------------------------------------------------------
async def auto_pilot_loop():
    logger.info("🤖 [Auto-Pilot] Daemon started.")
    while True:
        try:
            await asyncio.sleep(5) 
            if not sys_state.ready: continue
            
            if adaptive_hub and int(time.time()) % 60 == 0:
                sys_state.training_status_msg = adaptive_hub.get_status()

            if trade_manager and len(trade_manager.open_positions) > 0:
                wd_status = await trade_manager.ensure_active_guardians()
                if "No active" not in wd_status:
                    if not sys_state.cycle_running:
                        sys_state.last_cycle_logs = trade_manager.latest_guardian_log
                continue 

            if sys_state.auto_pilot and not sys_state.cycle_running and not sys_state.training_running:
                if sys_state.last_cycle_time:
                    elapsed = (datetime.now() - sys_state.last_cycle_time).total_seconds()
                    if elapsed < sys_state.scan_interval: 
                        continue

                logger.info("🤖 [Auto-Pilot] Triggering scan...")
                asyncio.create_task(run_unified_cycle())
                await asyncio.sleep(5) 
                    
        except Exception as e:
            logger.error(f"⚠️ [Auto-Pilot Error] {e}")
            await asyncio.sleep(30)

# ------------------------------------------------------------------------------
# Lifespan
# ------------------------------------------------------------------------------
@asynccontextmanager
async def lifespan(app: FastAPI):
    global r2, data_manager, ml_processor, adaptive_hub, trade_manager, whale_monitor, news_fetcher, senti_analyzer, sys_state, scheduler
    
    logger.info("\n🚀 [System] Startup Sequence (Titan V61.4 - Restoration)...")
    try:
        r2 = R2Service()
        data_manager = DataManager(contracts_db={}, whale_monitor=None, r2_service=r2)
        await data_manager.initialize() 
        await data_manager.load_contracts_from_r2()
        
        whale_monitor = EnhancedWhaleMonitor(contracts_db=data_manager.get_contracts_db(), r2_service=r2)
        rpc_mgr = AdaptiveRpcManager(data_manager.http_client)
        whale_monitor.set_rpc_manager(rpc_mgr)
        
        news_fetcher = NewsFetcher()
        senti_analyzer = SentimentIntensityAnalyzer()
        data_manager.whale_monitor = whale_monitor
        
        adaptive_hub = AdaptiveHub(r2_service=r2)
        await adaptive_hub.initialize()

        ml_processor = MLProcessor(data_manager=data_manager)
        await ml_processor.initialize()
        
        trade_manager = TradeManager(r2_service=r2, data_manager=data_manager, processor=ml_processor)
        trade_manager.learning_hub = adaptive_hub 
        
        await trade_manager.initialize_sentry_exchanges()
        await trade_manager.start_sentry_loops()
        
        scheduler = AutoTunerScheduler(trade_manager)
        asyncio.create_task(scheduler.start_loop())
        logger.info("🕰️ [Scheduler] Auto-Tuner Background Task Started.")
        
        sys_state.set_ready()
        asyncio.create_task(auto_pilot_loop())
        logger.info("✅ [System READY] All modules operational.")
        yield
        
    except Exception as e:
        logger.critical(f"❌ [FATAL STARTUP ERROR] {e}")
        traceback.print_exc()
    finally:
        sys_state.ready = False
        if trade_manager: await trade_manager.stop_sentry_loops()
        if data_manager: await data_manager.close()
        if whale_monitor and whale_monitor.rpc_manager: await whale_monitor.rpc_manager.close()
        logger.info("✅ [System] Shutdown Complete.")

# ------------------------------------------------------------------------------
# Helper Tasks
# ------------------------------------------------------------------------------
async def _analyze_symbol_task(candidate_data: Dict[str, Any]) -> Dict[str, Any]:
    try:
        symbol = candidate_data['symbol']
        required_tfs = ["5m", "15m", "1h", "4h"]
        
        data_tasks = [data_manager.get_latest_ohlcv(symbol, tf, limit=300) for tf in required_tfs]
        all_data = await asyncio.gather(*data_tasks)
        
        ohlcv_data = {}
        for tf, data in zip(required_tfs, all_data):
            if data and len(data) > 0: ohlcv_data[tf] = data
            
        if '1h' not in ohlcv_data or '5m' not in ohlcv_data: 
            return None
        
        current_price = await data_manager.get_latest_price_async(symbol)
        
        raw_data = {
            'symbol': symbol, 
            'ohlcv': ohlcv_data, 
            'current_price': current_price, 
            'timestamp': time.time(),
            'dynamic_limits': candidate_data.get('dynamic_limits', {}),
            'asset_regime': candidate_data.get('asset_regime', 'UNKNOWN'),
            'strategy_tag': candidate_data.get('strategy_tag', 'NONE'),
            'strategy_type': candidate_data.get('strategy_type', 'NORMAL'), 
            'l1_score': candidate_data.get('l1_sort_score', 0)
        }
        
        res = await ml_processor.process_compound_signal(raw_data)
        if not res: return None
        
        res['strategy_type'] = candidate_data.get('strategy_type', 'NORMAL')
        return res
    except Exception: return None

# ------------------------------------------------------------------------------
# Unified Cycle
# ------------------------------------------------------------------------------
async def run_unified_cycle():
    log_buffer = StringIO()
    def log_and_print(message):
        logger.info(message)
        log_buffer.write(message + '\n')

    if sys_state.cycle_running or sys_state.training_running: return
    if not sys_state.ready: return

    sys_state.set_cycle_start()
    
    try:
        await trade_manager.sync_internal_state_with_r2()
        
        if len(trade_manager.open_positions) > 0:
            log_and_print(f"ℹ️ [Cycle] Active Positions: {len(trade_manager.open_positions)}")
            for sym, tr in trade_manager.open_positions.items():
                curr_p = await data_manager.get_latest_price_async(sym)
                entry_p = float(tr.get('entry_price', 0))
                pnl = ((curr_p - entry_p) / entry_p) * 100 if entry_p > 0 else 0
                log_and_print(f"   🔒 {sym}: {pnl:+.2f}%")

        log_and_print(f"   [1/5] 🔍 L1 Screening (Bottom/Momentum)...")
        candidates = await data_manager.layer1_rapid_screening(adaptive_hub_ref=adaptive_hub) 
        if not candidates:
            log_and_print("⚠️ No valid candidates found (Quality Filter).")
            sys_state.set_cycle_end(logs=log_buffer.getvalue())
            return
            
        log_and_print(f"   [2/5] 🧠 L2 Deep Analysis ({len(candidates)} items)...")
        tasks = [_analyze_symbol_task(c) for c in candidates]
        results = await asyncio.gather(*tasks)
        valid_l2 = [res for res in results if res is not None]
        
        semi_finalists = sorted(valid_l2, key=lambda x: x.get('enhanced_final_score', 0.0), reverse=True)[:10]
        if not semi_finalists:
            log_and_print("⚠️ No valid L2 candidates.")
            sys_state.set_cycle_end(logs=log_buffer.getvalue())
            return

        log_and_print(f"   [3/5] 📡 L3 Deep Dive (Whales & News) for TOP {len(semi_finalists)}...")
        final_candidates = []

        for sig in semi_finalists:
            symbol = sig['symbol']
            l2_score = sig.get('enhanced_final_score', 0.0)
            
            whale_points = 0.0
            try:
                if whale_monitor:
                    w_data = await whale_monitor.get_symbol_whale_activity(symbol, known_price=sig.get('current_price', 0))
                    if w_data and w_data.get('data_available', False) and 'trading_signal' in w_data:
                        signal = w_data['trading_signal']
                        action = signal.get('action', 'HOLD')
                        confidence = float(signal.get('confidence', 0.5))
                        dynamic_impact = SystemLimits.L3_WHALE_IMPACT_MAX * confidence 
                        if action == 'BUY': whale_points = dynamic_impact 
                        elif action == 'SELL': whale_points = -dynamic_impact
            except Exception: pass

            news_points = 0.0
            try:
                if news_fetcher and senti_analyzer:
                    n_data = await news_fetcher.get_news(symbol)
                    summary_text = n_data.get('summary', '')
                    if "No specific news" not in summary_text:
                        sent = senti_analyzer.polarity_scores(summary_text)
                        compound_score = sent['compound']
                        news_points = compound_score * SystemLimits.L3_NEWS_IMPACT_MAX
            except Exception: pass

            mc_a_points = 0.0
            try:
                raw_mc_a = await ml_processor.run_advanced_monte_carlo(symbol, '1h')
                mc_a_points = max(-SystemLimits.L3_MC_ADVANCED_MAX, min(SystemLimits.L3_MC_ADVANCED_MAX, raw_mc_a))
            except Exception: pass

            final_score = l2_score + whale_points + news_points + mc_a_points
            
            sig['whale_score'] = whale_points
            sig['news_score'] = news_points
            sig['mc_advanced_score'] = mc_a_points
            sig['final_total_score'] = final_score 
            final_candidates.append(sig)

        final_candidates.sort(key=lambda x: x['final_total_score'], reverse=True)

        approved_signals = []
        
        header = (f"{'SYM':<9} | {'TYPE':<10} | {'L2':<5} | {'TITAN':<5} | {'PATT':<5} | "
                  f"{'WHALE':<6} | {'MC(A)':<6} | {'FINAL':<6} | {'ORACLE':<6} | {'STATUS'}")
        
        log_and_print("-" * 115)
        log_and_print(header)
        log_and_print("-" * 115)
        
        for sig in final_candidates:
            symbol = sig['symbol']
            
            decision = await ml_processor.consult_oracle(sig)
            
            action = decision.get('action', 'WAIT')
            oracle_conf = decision.get('confidence', 0.0)
            target_class = decision.get('target_class', '')
            
            status_str = "WAIT 🔴"
            if action == 'WATCH' or action == 'BUY':
                status_str = f"✅ {target_class}"
                sig.update(decision)
                approved_signals.append(sig)

            l2_hybrid = sig.get('enhanced_final_score', 0.0)
            titan_d = sig.get('titan_score', 0.0)
            patt_d = sig.get('patterns_score', 0.0)
            whale_d = sig.get('whale_score', 0.0)
            mca_d = sig.get('mc_advanced_score', 0.0)
            final_d = sig.get('final_total_score', 0.0)
            strat_type = sig.get('strategy_type', 'N/A')

            log_and_print(
                f"{symbol:<9} | "
                f"{strat_type:<10} | "
                f"{l2_hybrid:.2f}  | "
                f"{titan_d:.2f}  | "
                f"{patt_d:.2f}  | "
                f"{whale_d:+.2f}  | " 
                f"{mca_d:+.2f}  | " 
                f"{final_d:.2f}   | " 
                f"{oracle_conf:.2f}   | "
                f"{status_str}"
            )

        if approved_signals:
            log_and_print("-" * 115)
            log_and_print(f"   [4/5] 🎯 L4 Sniper -> 🏛️ Governance -> 💰 Portfolio ({len(approved_signals)} candidates)...")
            tm_log_buffer = StringIO()
            
            with redirect_stdout(tm_log_buffer), redirect_stderr(tm_log_buffer):
                await trade_manager.select_and_execute_best_signal(approved_signals)
            
            tm_logs = tm_log_buffer.getvalue()
            for line in tm_logs.splitlines():
                if line.strip(): log_and_print(line.strip())
        else:
            log_and_print("   -> 🛑 No candidates approved by Oracle for Sniper check.")

        gc.collect()
        sys_state.set_cycle_end(logs=log_buffer.getvalue())

    except Exception as e:
        logger.error(f"❌ [Cycle ERROR] {e}")
        traceback.print_exc()
        sys_state.set_cycle_end(error=e, logs=log_buffer.getvalue())

# ------------------------------------------------------------------------------
# Handlers
# ------------------------------------------------------------------------------
async def trigger_training_cycle():
    if adaptive_hub: return f"🤖 Adaptive System: {adaptive_hub.get_status()}"
    return "⚠️ System not ready."

async def trigger_strategic_backtest():
    if not BACKTEST_AVAILABLE: return "⚠️ Backtest Engine not found."
    if trade_manager and len(trade_manager.open_positions) > 0: return "⛔ Active trades exist."
    if sys_state.training_running: return "⚠️ Running."
        
    async def _run_bg_task():
        sys_state.training_running = True
        sys_state.training_status_msg = "🧪 Strategic Backtest Running..."
        try:
            logger.info("🧪 [Manual] Starting Strategic Backtest...")
            await run_strategic_optimization_task()
            if adaptive_hub: await adaptive_hub.initialize()
            logger.info("✅ [Manual] Backtest Complete.")
        except Exception as e: logger.error(f"❌ Backtest Failed: {e}")
        finally:
            sys_state.training_running = False
            sys_state.training_status_msg = adaptive_hub.get_status() if adaptive_hub else "Ready"

    asyncio.create_task(_run_bg_task())
    return "🧪 Strategic Backtest Started."

async def manual_close_current_trade():
    if not trade_manager.open_positions: return "⚠️ No trade."
    symbol = list(trade_manager.open_positions.keys())[0]
    await trade_manager.force_exit_by_manager(symbol, reason="MANUAL_UI")
    return f"✅ Closed {symbol}."

async def reset_history_handler():
    if trade_manager.open_positions: return "⚠️ Close active trades first."
    current_state = await r2.get_portfolio_state_async()
    preserved_capital = current_state.get('current_capital_usd', INITIAL_CAPITAL)
    await r2.reset_all_stats_async()
    if trade_manager and trade_manager.smart_portfolio:
         sp = trade_manager.smart_portfolio
         sp.state["current_capital"] = preserved_capital 
         sp.state["session_start_balance"] = preserved_capital
         sp.state["allocated_capital_usd"] = 0.0
         sp.state["daily_net_pnl"] = 0.0
         sp.state["is_trading_halted"] = False
         await sp._save_state_to_r2()
    return f"✅ History Cleared. Capital Preserved at ${preserved_capital:.2f}"

async def reset_capital_handler():
    if trade_manager.open_positions: return "⚠️ Close active trades first."
    if trade_manager and trade_manager.smart_portfolio:
         sp = trade_manager.smart_portfolio
         sp.state["current_capital"] = INITIAL_CAPITAL 
         sp.state["session_start_balance"] = INITIAL_CAPITAL
         sp.state["allocated_capital_usd"] = 0.0
         sp.state["daily_net_pnl"] = 0.0
         sp.state["is_trading_halted"] = False
         await sp._save_state_to_r2()
    return f"✅ Capital Reset to ${INITIAL_CAPITAL} (History Kept)"

async def reset_diagnostics_handler():
    await r2.reset_diagnostic_stats_async()
    return "✅ Diagnostic Matrix Reset."

async def reset_guardians_handler():
    await r2.reset_guardian_stats_async()
    if trade_manager: trade_manager.ai_stats = await r2.get_guardian_stats_async()
    return "✅ Guardian Stats Reset."

async def toggle_auto_pilot(enable):
    sys_state.auto_pilot = enable
    return f"Auto-Pilot: {enable}"

async def run_cycle_from_gradio():
    if sys_state.cycle_running: return "Busy."
    asyncio.create_task(run_unified_cycle())
    return "🚀 Launched."

# ------------------------------------------------------------------------------
# UI Updates
# ------------------------------------------------------------------------------
async def check_live_pnl_and_status(selected_view="Dual-Core (Hybrid)"):
    empty_chart = go.Figure()
    empty_chart.update_layout(template="plotly_dark", paper_bgcolor="#0b0f19", plot_bgcolor="#0b0f19", xaxis={'visible':False}, yaxis={'visible':False})
    wl_df_empty = pd.DataFrame(columns=["Coin", "Score"])
    diag_df_empty = pd.DataFrame(columns=["Model", "Wins", "Losses", "PnL (USD)"])
    type_df_empty = pd.DataFrame(columns=["Coin Type", "Wins", "Losses", "Profitability"])
    
    if not sys_state.ready:
        return "Initializing...", "...", empty_chart, "0.0", "0.0", "0.0", "0.0", "0.0%", wl_df_empty, diag_df_empty, type_df_empty, "Loading...", "Loading...", "Loading..."

    try:
        sp = trade_manager.smart_portfolio
        equity = sp.state.get('current_capital', 10.0) 
        allocated = sp.state.get('allocated_capital_usd', 0.0)
        free_cap = max(0.0, equity - allocated)
        daily_pnl = sp.state.get('daily_net_pnl', 0.0)
        is_halted = sp.state.get('is_trading_halted', False)
        
        symbol = None; entry_p = 0.0; tp_p = 0.0; sl_p = 0.0; curr_p = 0.0; pnl_pct = 0.0; pnl_val_unrealized = 0.0
        active_trade_info = ""
        trade_dur_str = "--:--:--"

        if trade_manager.open_positions:
            symbol = list(trade_manager.open_positions.keys())[0]
            trade = trade_manager.open_positions[symbol]
            entry_p = float(trade.get('entry_price', 0.0))
            tp_p = float(trade.get('tp_price', 0.0))
            sl_p = float(trade.get('sl_price', 0.0))
            trade_dur_str = calculate_duration_str(trade.get('entry_time'))
            
            decision_data = trade.get('decision_data', {})
            gov_grade = decision_data.get('governance_grade', 'N/A')
            gov_score = decision_data.get('governance_score', 0.0)
            sys_conf = decision_data.get('system_confidence', 0.0)
            strat_type = trade.get('strategy_type', 'NORMAL')
            
            grade_color = "#ccc"
            if gov_grade == "ULTRA": grade_color = "#ff00ff" 
            elif gov_grade == "STRONG": grade_color = "#00ff00" 
            elif gov_grade == "NORMAL": grade_color = "#00e5ff" 
            elif gov_grade == "WEAK": grade_color = "#ffff00" 
            elif gov_grade == "REJECT": grade_color = "#ff0000" 

            curr_p = await data_manager.get_latest_price_async(symbol)
            if curr_p > 0 and entry_p > 0:
                pnl_pct = ((curr_p - entry_p) / entry_p) * 100
                size = float(trade.get('entry_capital', 0.0))
                pnl_val_unrealized = size * (pnl_pct / 100)
            
            active_trade_info = f"""
             <div style='display: flex; justify-content: space-between; font-size: 12px; color: #ccc; margin-top:5px; border-top: 1px solid #333; padding-top: 5px;'>
                <span>⏱️ Time:</span> <span style='color: #ffff00;'>{trade_dur_str}</span>
             </div>
             <div style='display: flex; justify-content: space-between; font-size: 12px; color: #ccc; margin-top:5px;'>
                <span>🏛️ Grade:</span> <span style='color: {grade_color}; font-weight:bold;'>{gov_grade} ({gov_score:.1f})</span>
             </div>
             <div style='display: flex; justify-content: space-between; font-size: 12px; color: #ccc; margin-top:5px;'>
                <span>🏷️ Type:</span> <span style='color: #orange;'>{strat_type}</span>
             </div>
            """

        virtual_equity = equity + pnl_val_unrealized
        active_pnl_color = "#00ff00" if pnl_val_unrealized >= 0 else "#ff0000"
        portfolio = await r2.get_portfolio_state_async()
        total_t = portfolio.get('total_trades', 0)
        wins = portfolio.get('winning_trades', 0)
        losses = portfolio.get('losing_trades', 0)
        if losses == 0 and total_t > 0: losses = total_t - wins
        tot_prof = portfolio.get('total_profit_usd', 0.0)
        tot_loss = portfolio.get('total_loss_usd', 0.0)
        net_prof = tot_prof - tot_loss
        win_rate = (wins / total_t * 100) if total_t > 0 else 0.0
        halt_status = "<span style='color:red; font-weight:bold;'>HALTED</span>" if is_halted else "<span style='color:#00ff00;'>ACTIVE</span>"

        wallet_md = f"""
        <div style='background-color: #1a1a1a; padding: 15px; border-radius: 8px; border: 1px solid #333; text-align:center;'>
            <h3 style='margin:0; color:#888; font-size:14px;'>💼 Institutional Portfolio</h3>
            <div style='font-size: 24px; font-weight: bold; color: white; margin: 5px 0 0 0;'>${virtual_equity:,.2f}</div>
            <div style='font-size: 14px; color: {active_pnl_color}; margin-bottom: 5px;'>({pnl_val_unrealized:+,.2f} USD)</div>
            
            <table style='width:100%; font-size:12px; margin-top:5px; color:#ccc;'>
                <tr><td>Allocated:</td><td style='text-align:right; color:#ffa500;'>${allocated:.2f}</td></tr>
                <tr><td>Free Cap:</td><td style='text-align:right; color:#00ff00;'>${free_cap:.2f}</td></tr>
                <tr><td>Daily PnL:</td><td style='text-align:right; color:{"#00ff00" if daily_pnl>=0 else "#ff0000"};'>${daily_pnl:+.2f}</td></tr>
            </table>
            
            <hr style='border-color:#444; margin: 10px 0;'>
            
            <div style='display: flex; justify-content: space-between; font-size: 12px; color: #ccc;'>
                <span>🦅 Mood:</span> <span style='color: white;'>{sp.market_trend}</span>
            </div>
            <div style='display: flex; justify-content: space-between; font-size: 12px; color: #ccc; margin-top:5px;'>
                <span>🛡️ Status:</span> {halt_status}
            </div>
            {active_trade_info}
        </div>
        """

        # Guardian Stats
        key_map = {
            "Dual-Core (Hybrid)": "hybrid", 
            "Hydra: Crash (Panic)": "crash", 
            "Hydra: Giveback (Profit)": "giveback", 
            "Hydra: Stagnation (Time)": "stagnation"
        }
        target_key = key_map.get(selected_view, "hybrid")
        stats_data = trade_manager.ai_stats.get(target_key, {"total":0, "good":0, "saved":0.0, "missed":0.0})
        
        tot_ds = stats_data['total']
        ds_acc = (stats_data['good'] / tot_ds * 100) if tot_ds > 0 else 0.0
        
        # ✅ RESTORED MISSED ROW
        history_md = f"""
        <div style='background-color: #1a1a1a; padding: 10px; border-radius: 8px; border: 1px solid #333; font-size: 12px;'>
            <h3 style='margin:0 0 5px 0; color:#888; font-size:14px;'>📊 Performance</h3>
            <table style='width:100%; color:white;'>
                <tr><td>Trades:</td><td style='text-align:right;'>{total_t}</td></tr>
                <tr><td>Win Rate:</td><td style='text-align:right; color:{"#00ff00" if win_rate>=50 else "#ff0000"};'>{win_rate:.1f}%</td></tr>
                <tr><td>Wins:</td><td style='text-align:right; color:#00ff00;'>{wins} (+${tot_prof:,.2f})</td></tr>
                <tr><td>Losses:</td><td style='text-align:right; color:#ff0000;'>{losses} (-${tot_loss:,.2f})</td></tr>
                <tr><td style='border-top:1px solid #444;'>Net:</td><td style='border-top:1px solid #444; text-align:right; color:{"#00ff00" if net_prof>=0 else "#ff0000"};'>${net_prof:,.2f}</td></tr>
            </table>
            <hr style='border-color:#444; margin: 8px 0;'>
            <h3 style='margin:0 0 5px 0; color: #00e5ff; font-size:14px;'>🛡️ Guard IQ ({target_key})</h3>
            <table style='width:100%; color:white;'>
                <tr><td>Interventions:</td><td style='text-align:right;'>{tot_ds}</td></tr>
                <tr><td>Accuracy:</td><td style='text-align:right; color:#00e5ff;'>{ds_acc:.1f}%</td></tr>
                <tr><td>Saved:</td><td style='text-align:right; color:#00ff00;'>${stats_data['saved']:.2f}</td></tr>
                <tr><td>Missed:</td><td style='text-align:right; color:#ff0000;'>${stats_data['missed']:.2f}</td></tr>
            </table>
        </div>
        """

        fast_learn_prog = "0/100"
        if adaptive_hub:
            if hasattr(adaptive_hub, 'get_learning_progress'):
                fast_learn_prog = adaptive_hub.get_learning_progress()
        
        metrics = scheduler.get_status_metrics() if scheduler else {}
        sch_w_time = metrics.get("weekly_timer", "Wait")
        sch_m_time = metrics.get("monthly_timer", "Wait")

        neural_md = f"""
        <div style='background-color: #1a1a1a; padding: 10px; border-radius: 8px; border: 1px solid #333; font-size: 12px; margin-top: 10px;'>
            <h3 style='margin:0; color:#00e5ff; font-size:14px;'>🧠 Neural Cycles</h3>
            <table style='width:100%; color:#ccc;'>
                <tr style='border-bottom: 1px solid #333;'>
                    <td style='padding:4px 0;'>⚡ Fast Learner:</td>
                    <td style='text-align:right; color:#ffff00; font-weight:bold;'>{fast_learn_prog}</td>
                </tr>
                <tr style='border-bottom: 1px solid #333;'>
                    <td style='padding:4px 0;'>📅 Weekly Tune:</td>
                    <td style='text-align:right; color:#fff;'>{sch_w_time}</td>
                </tr>
                <tr>
                    <td style='padding:4px 0;'>🗓️ Monthly Evo:</td>
                    <td style='text-align:right; color:#fff;'>{sch_m_time}</td>
                </tr>
            </table>
        </div>
        """
        
        # Diagnostic Matrix
        diag_data = await r2.get_diagnostic_stats_async()
        diag_list = []
        ordered_models = ["Titan", "Patterns", "Oracle", "Sniper", "MonteCarlo_L", "Governance"]
        
        for m in ordered_models:
            stats = diag_data.get(m, {"wins": 0, "losses": 0, "pnl": 0.0})
            pnl_val = stats['pnl']
            color = "#00ff00" if pnl_val >= 0 else "#ff0000"
            pnl_str = f"<span style='color: {color}; font-weight: bold;'>${pnl_val:+.2f}</span>"
            diag_list.append([m, stats['wins'], stats['losses'], pnl_str])
        
        diag_df = pd.DataFrame(diag_list, columns=["Model", "Wins", "Losses", "PnL (USD)"])

        # Type Stats DataFrame
        type_stats_list = []
        if trade_manager and hasattr(trade_manager, 'type_stats'):
            for t_name, t_data in trade_manager.type_stats.items():
                name_clean = t_name.replace("_", " ")
                wins = t_data.get('wins', 0)
                losses = t_data.get('losses', 0)
                prof = t_data.get('profit_usd', 0.0)
                loss_val = t_data.get('loss_usd', 0.0)
                profitability_html = format_pnl_split(prof, loss_val)
                type_stats_list.append([name_clean, wins, losses, profitability_html])
        
        type_df = pd.DataFrame(type_stats_list, columns=["Coin Type", "Wins", "Losses", "Profitability"])

        wl_data = [[k, f"{v.get('final_total_score',0):.2f}"] for k, v in trade_manager.watchlist.items()]
        wl_df = pd.DataFrame(wl_data, columns=["Coin", "Score"])

        status_txt = sys_state.last_cycle_logs
        status_line = f"Cycle: {'RUNNING' if sys_state.cycle_running else 'IDLE'} | Auto-Pilot: {'ON' if sys_state.auto_pilot else 'OFF'}"
        
        fig = empty_chart
        if symbol and curr_p > 0:
            ohlcv = await data_manager.get_latest_ohlcv(symbol, '5m', 120)
            if ohlcv:
                df = pd.DataFrame(ohlcv, columns=['timestamp', 'open', 'high', 'low', 'close', 'volume'])
                df['datetime'] = pd.to_datetime(df['timestamp'], unit='ms')
                fig = go.Figure(data=[go.Candlestick(
                    x=df['datetime'], open=df['open'], high=df['high'], low=df['low'], close=df['close'],
                    increasing_line_color='#00ff00', decreasing_line_color='#ff0000', name=symbol
                )])
                if entry_p > 0:
                    fig.add_hline(y=entry_p, line_dash="dash", line_color="white", annotation_text="ENTRY", annotation_position="top left")
                if tp_p > 0:
                    fig.add_hline(y=tp_p, line_color="#00ff00", line_width=2, annotation_text="TP", annotation_position="top left")
                if sl_p > 0:
                    fig.add_hline(y=sl_p, line_color="#ff0000", line_width=2, annotation_text="SL", annotation_position="bottom left")

                fig.update_layout(
                    template="plotly_dark", 
                    paper_bgcolor="#0b0f19", 
                    plot_bgcolor="#0b0f19",
                    margin=dict(l=0, r=40, t=30, b=0), 
                    height=400, 
                    xaxis_rangeslider_visible=False,
                    title=dict(text=f"{symbol} (Long) | PnL: {pnl_pct:+.2f}%", font=dict(color="white"))
                )

        train_status = sys_state.training_status_msg
        if sys_state.training_running: train_status = "🧪 Backtest Running..."

        return (status_txt, status_line, fig, f"{curr_p:.6f}", f"{entry_p:.6f}", f"{tp_p:.6f}", f"{sl_p:.6f}",
                f"{pnl_pct:+.2f}%", wl_df, diag_df, type_df, wallet_md, history_md, neural_md)

    except Exception:
        traceback.print_exc()
        return "Error", "Error", empty_chart, "0", "0", "0", "0", "0%", wl_df_empty, diag_df_empty, type_df_empty, "Err", "Err", "Err"

# ------------------------------------------------------------------------------
# Gradio UI Construction
# ------------------------------------------------------------------------------
def create_gradio_ui():
    custom_css = ".gradio-container {background:#0b0f19} .dataframe {background:#1a1a1a!important} .html-box {min-height:180px}"
    
    with gr.Blocks(title="Titan V61.4 (Full Neural Dashboard)") as demo:
        gr.HTML(f"<style>{custom_css}</style>")
        
        gr.Markdown("# 🚀 Titan V61.4 (Cybernetic: Dual-Type Engine)")
        
        with gr.Row():
            with gr.Column(scale=3):
                live_chart = gr.Plot(label="Chart")
                with gr.Row():
                    t_price = gr.Textbox(label="Price", interactive=False)
                    t_pnl = gr.Textbox(label="PnL %", interactive=False)
                with gr.Row():
                    t_entry = gr.Textbox(label="Entry", interactive=False)
                    t_tp = gr.Textbox(label="TP", interactive=False)
                    t_sl = gr.Textbox(label="SL", interactive=False)

            with gr.Column(scale=1):
                wallet_out = gr.HTML(label="Smart Wallet", elem_classes="html-box")
                neural_out = gr.HTML(label="Neural Cycles", elem_classes="html-box")
                
                # Type Stats Table
                gr.Markdown("### 💎 Opportunity Types")
                type_stats_out = gr.Dataframe(
                    headers=["Coin Type", "Wins", "Losses", "Profitability"],
                    datatype=["str", "number", "number", "html"],
                    interactive=False,
                    label="Type Performance"
                )

                gr.Markdown("### 🕵️ Diagnostic Matrix")
                diagnostic_out = gr.Dataframe(
                    headers=["Model", "Wins", "Losses", "PnL (USD)"],
                    datatype=["str", "number", "number", "html"], 
                    interactive=False,
                    label="Model Performance"
                )
                
                with gr.Row():
                    btn_reset_diag = gr.Button("🧹 Reset Matrix", size="sm", variant="secondary")
                    btn_reset_guard = gr.Button("🛡️ Reset Guardians", size="sm", variant="secondary")
                
                gr.HTML("<hr style='border-color:#444; margin: 10px 0;'>")
                
                stats_dd = gr.Dropdown([
                    "Dual-Core (Hybrid)", 
                    "Hydra: Crash (Panic)", 
                    "Hydra: Giveback (Profit)", 
                    "Hydra: Stagnation (Time)"
                ], value="Dual-Core (Hybrid)", label="View Guard Stats")
                history_out = gr.HTML(label="Stats", elem_classes="html-box")
                watchlist_out = gr.DataFrame(label="Watchlist")

        gr.HTML("<hr style='border-color:#333'>")
        
        with gr.Row():
            with gr.Column(scale=1):
                auto_pilot = gr.Checkbox(label="✈️ Auto-Pilot", value=True)
                with gr.Row():
                    btn_run = gr.Button("🚀 Scan", variant="primary")
                    btn_close = gr.Button("🚨 Close", variant="stop")
                with gr.Row():
                    btn_train = gr.Button("🤖 Status", variant="secondary")
                    btn_backtest = gr.Button("🧪 Run Strategic Backtest", variant="secondary")
                with gr.Row():
                    btn_history_reset = gr.Button("🗑️ Clear History", variant="secondary")
                    btn_cap_reset = gr.Button("💰 Reset Capital", variant="secondary")
                    
                status = gr.Markdown("Init...")
                alert = gr.Textbox(label="Alerts", interactive=False)
            
            with gr.Column(scale=3):
                logs = gr.Textbox(label="Logs", lines=14, autoscroll=True, elem_classes="log-box", type="text")
                gr.HTML("<style>.log-box textarea { font-family: 'Consolas', 'Monaco', monospace !important; font-size: 12px !important; white-space: pre !important; }</style>")

        btn_run.click(fn=run_cycle_from_gradio, outputs=alert)
        btn_close.click(fn=manual_close_current_trade, outputs=alert)
        btn_history_reset.click(fn=reset_history_handler, outputs=alert)
        btn_cap_reset.click(fn=reset_capital_handler, outputs=alert)
        btn_train.click(fn=trigger_training_cycle, outputs=alert)
        btn_backtest.click(fn=trigger_strategic_backtest, outputs=alert)
        auto_pilot.change(fn=toggle_auto_pilot, inputs=auto_pilot, outputs=alert)
        
        btn_reset_diag.click(fn=reset_diagnostics_handler, outputs=alert)
        btn_reset_guard.click(fn=reset_guardians_handler, outputs=alert)
        
        gr.Timer(3).tick(fn=check_live_pnl_and_status, inputs=stats_dd, 
                         outputs=[logs, status, live_chart, t_price, t_entry, t_tp, t_sl, t_pnl, watchlist_out, diagnostic_out, type_stats_out, wallet_out, history_out, neural_out])
    return demo

fast_api_server = FastAPI(lifespan=lifespan)
gradio_dashboard = create_gradio_ui()
app = gr.mount_gradio_app(app=fast_api_server, blocks=gradio_dashboard, path="/")

if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860)