Trading-Bot/download_data.py

# download_data.py
"""
Script para descargar datos históricos de múltiples símbolos y timeframes
"""
import os
from dotenv import load_dotenv
from datetime import datetime
from pathlib import Path
from src.utils.logger import log
from src.data.fetcher import DataFetcher
from src.data.processor import DataProcessor
from src.data.storage import StorageManager

def setup_environment():
    """Carga variables de entorno"""
    env_path = Path(__file__).parent / 'config' / 'secrets.env'
    load_dotenv(dotenv_path=env_path)
    log.success("✓ Variables de entorno cargadas")

def download_multiple_symbols():
    """
    Descarga datos históricos para múltiples símbolos y timeframes
    """
    log.info("="*70)
    log.info("📥 DESCARGA MASIVA DE DATOS HISTÓRICOS")
    log.info("="*70)
    
    # Configuración
    setup_environment()
    
    exchange_name = os.getenv('EXCHANGE_NAME', 'binance')
    
    # Símbolos a descargar (puedes añadir más)
    symbols = [
        'BTC/USDT',
        'ETH/USDT',
        'BNB/USDT',
        'SOL/USDT',
        'XRP/USDT',
    ]
    
    # Timeframes a descargar
    timeframes = [
        '1h',   # 1 hora
        '4h',   # 4 horas
        '1d',   # 1 día
    ]
    
    # Días históricos
    days_back = 120  # 4 meses
    
    log.info(f"\n📊 Configuración:")
    log.info(f"   Exchange: {exchange_name}")
    log.info(f"   Símbolos: {len(symbols)} → {symbols}")
    log.info(f"   Timeframes: {timeframes}")
    log.info(f"   Días históricos: {days_back}")
    log.info(f"   Total descargas: {len(symbols) * len(timeframes)}")
    
    # Confirmar
    print("\n" + "="*70)
    response = input("¿Continuar con la descarga? (s/n): ").lower()
    if response != 's':
        log.warning("Descarga cancelada por el usuario")
        return
    
    # Inicializar componentes
    log.info("\n🔧 Inicializando componentes...")
    
    fetcher = DataFetcher(
        exchange_name=exchange_name,
        api_key=os.getenv('API_KEY') if os.getenv('API_KEY') else None,
        api_secret=os.getenv('API_SECRET') if os.getenv('API_SECRET') else None
    )
    
    processor = DataProcessor()
    
    storage = StorageManager(
        db_host=os.getenv('DB_HOST'),
        db_port=int(os.getenv('DB_PORT', 5432)),
        db_name=os.getenv('DB_NAME'),
        db_user=os.getenv('DB_USER'),
        db_password=os.getenv('DB_PASSWORD'),
        redis_host=os.getenv('REDIS_HOST', 'localhost'),
        redis_port=int(os.getenv('REDIS_PORT', 6379))
    )
    
    # Estadísticas
    total_downloads = len(symbols) * len(timeframes)
    current_download = 0
    successful = 0
    failed = 0
    total_records = 0
    
    # Descargar datos
    log.info("\n" + "="*70)
    log.info("🚀 INICIANDO DESCARGAS...")
    log.info("="*70 + "\n")
    
    for symbol in symbols:
        for timeframe in timeframes:
            current_download += 1
            
            log.info(f"\n{'='*70}")
            log.info(f"📥 [{current_download}/{total_downloads}] {symbol} @ {timeframe}")
            log.info(f"{'='*70}")
            
            try:
                # Verificar si ya existen datos
                last_timestamp = storage.get_latest_timestamp(symbol, timeframe)
                
                if last_timestamp:
                    log.info(f"⚠️  Ya existen datos hasta: {last_timestamp}")
                    log.info(f"   Se descargarán solo datos nuevos...")
                
                # Descargar datos
                df = fetcher.fetch_historical(
                    symbol=symbol,
                    timeframe=timeframe,
                    days=days_back
                )
                
                if df.empty:
                    log.warning(f"❌ No se obtuvieron datos para {symbol} @ {timeframe}")
                    failed += 1
                    continue
                
                # Procesar
                log.info(f"🧹 Procesando datos...")
                df_clean = processor.clean_data(df)
                df_clean = processor.calculate_returns(df_clean)
                
                # Guardar
                log.info(f"💾 Guardando en base de datos...")
                records_saved = storage.save_ohlcv(df_clean)
                
                total_records += records_saved
                successful += 1
                
                log.success(f"✅ {symbol} @ {timeframe} → {records_saved} registros guardados")
                
            except Exception as e:
                log.error(f"❌ Error descargando {symbol} @ {timeframe}: {e}")
                failed += 1
                continue
    
    # Resumen final
    log.info("\n" + "="*70)
    log.info("📊 RESUMEN DE DESCARGA")
    log.info("="*70)
    log.info(f"✅ Exitosas: {successful}/{total_downloads}")
    log.info(f"❌ Fallidas: {failed}/{total_downloads}")
    log.info(f"📦 Total registros guardados: {total_records:,}")
    
    # Ver datos disponibles
    log.info("\n📋 Datos disponibles en base de datos:")
    available_data = storage.get_available_data()
    print(available_data.to_string(index=False))
    
    # Cleanup
    storage.close()
    
    log.info("\n" + "="*70)
    log.success("✅ DESCARGA COMPLETADA")
    log.info("="*70)

if __name__ == "__main__":
    try:
        download_multiple_symbols()
    except KeyboardInterrupt:
        log.warning("\n⚠️  Descarga interrumpida por el usuario")
    except Exception as e:
        log.error(f"\n❌ Error fatal: {e}")
        raise