"""
营销策略模拟模块
"""
import numpy as np
import pandas as pd
from .utils import ensure_native_type
from .lifecycle_analyzer_service import enhanced_data_preprocessing


def simulate_marketing_strategy(revenue_series, quantity_series, date_series, stages_map, stage_statistics, strategy_type, strategy_params):
    """模拟营销策略对SKU生命周期的影响"""
    # 转换为可修改的列表
    simulated_revenue = list(revenue_series)
    simulated_quantity = list(quantity_series)
    
    # 解析策略参数
    start_date = strategy_params.get('start_date')
    end_date = strategy_params.get('end_date')
    intensity = float(strategy_params.get('intensity', 0.5))
    price_change = float(strategy_params.get('price_change', 0))
    
    # 找到策略作用的日期范围索引
    start_idx = None
    end_idx = None
    
    for i, date_str in enumerate(date_series):
        if start_date and date_str == start_date:
            start_idx = i
        if end_date and date_str == end_date:
            end_idx = i
    
    # 如果未找到精确匹配，尝试模糊匹配
    if start_idx is None and start_date:
        for i, date_str in enumerate(date_series):
            if start_date in date_str:
                start_idx = i
                break
    
    if end_idx is None and end_date:
        for i, date_str in enumerate(date_series):
            if end_date in date_str:
                end_idx = i
                break
    
    # 默认范围
    if start_idx is None:
        start_idx = 0
    if end_idx is None:
        end_idx = len(simulated_revenue) - 1
    
    # 确保范围有效
    start_idx = max(0, min(start_idx, len(simulated_revenue) - 1))
    end_idx = max(start_idx, min(end_idx, len(simulated_revenue) - 1))
    
    # 根据策略类型应用不同的影响
    strategy_effects = {
        'promotion': {  # 促销活动
            'quantity_boost': 1.3 + (intensity * 0.5),
            'revenue_boost': 1.15 + (intensity * 0.25),
        },
        'advertising': {  # 广告宣传
            'quantity_boost': 1.2 + (intensity * 0.6),
            'revenue_boost': 1.25 + (intensity * 0.6),
        },
        'price_cut': {  # 降价
            'quantity_boost': 1.4 + (abs(price_change) / 100 * 1.5),
            'revenue_boost': 1.0 + (price_change / 100),
        },
        'price_increase': {  # 涨价
            'quantity_boost': 1.0 - (price_change / 100 * 0.3),
            'revenue_boost': 1.0 + (price_change / 100),
        },
    }
    
    effect = strategy_effects.get(strategy_type, {'quantity_boost': 1.0, 'revenue_boost': 1.0})
    
    # 应用策略影响（考虑渐进式影响和衰减）
    for i in range(start_idx, end_idx + 1):
        progress = (i - start_idx) / max(1, end_idx - start_idx)
        
        # 使用钟形曲线模拟影响强度
        impact_factor = np.exp(-((progress - 0.5) ** 2) / 0.1)
        
        # 应用影响
        quantity_multiplier = 1.0 + (effect['quantity_boost'] - 1.0) * impact_factor
        revenue_multiplier = 1.0 + (effect['revenue_boost'] - 1.0) * impact_factor
        
        simulated_quantity[i] *= quantity_multiplier
        simulated_revenue[i] *= revenue_multiplier
    
    # 计算策略影响的后续效应
    carryover_length = min(20, (end_idx - start_idx) // 2)
    for i in range(end_idx + 1, min(end_idx + 1 + carryover_length, len(simulated_revenue))):
        decay_factor = 1.0 - ((i - end_idx) / carryover_length)
        carryover_boost = 0.1 * intensity * decay_factor
        simulated_quantity[i] *= (1.0 + carryover_boost)
        simulated_revenue[i] *= (1.0 + carryover_boost * 0.5)
    
    # 平滑处理
    smoothed_simulated_revenue, smoothed_simulated_quantity = enhanced_data_preprocessing(
        pd.Series(simulated_revenue), pd.Series(simulated_quantity)
    )
    
    # 计算对比指标
    original_total_revenue = sum(revenue_series)
    original_total_quantity = sum(quantity_series)
    simulated_total_revenue = sum(simulated_revenue)
    simulated_total_quantity = sum(simulated_quantity)
    
    revenue_increase = simulated_total_revenue - original_total_revenue
    quantity_increase = simulated_total_quantity - original_total_quantity
    revenue_increase_pct = (revenue_increase / original_total_revenue * 100) if original_total_revenue > 0 else 0
    quantity_increase_pct = (quantity_increase / original_total_quantity * 100) if original_total_quantity > 0 else 0
    
    # 计算策略期间的增量
    strategy_period_original_revenue = sum(revenue_series[start_idx:end_idx + 1])
    strategy_period_simulated_revenue = sum(simulated_revenue[start_idx:end_idx + 1])
    strategy_period_revenue_increase = strategy_period_simulated_revenue - strategy_period_original_revenue
    strategy_period_revenue_increase_pct = (
        (strategy_period_revenue_increase / strategy_period_original_revenue * 100) 
        if strategy_period_original_revenue > 0 else 0
    )
    
    return {
        'simulated_revenue': [ensure_native_type(v, decimal_places=2) for v in simulated_revenue],
        'simulated_quantity': [ensure_native_type(v, decimal_places=2) for v in simulated_quantity],
        'smoothed_simulated_revenue': [ensure_native_type(v, decimal_places=2) for v in smoothed_simulated_revenue.tolist()],
        'smoothed_simulated_quantity': [ensure_native_type(v, decimal_places=2) for v in smoothed_simulated_quantity.tolist()],
        'comparison': {
            'original_total_revenue': ensure_native_type(original_total_revenue, decimal_places=2),
            'simulated_total_revenue': ensure_native_type(simulated_total_revenue, decimal_places=2),
            'revenue_increase': ensure_native_type(revenue_increase, decimal_places=2),
            'revenue_increase_pct': ensure_native_type(revenue_increase_pct, decimal_places=2),
            'original_total_quantity': ensure_native_type(original_total_quantity, decimal_places=2),
            'simulated_total_quantity': ensure_native_type(simulated_total_quantity, decimal_places=2),
            'quantity_increase': ensure_native_type(quantity_increase, decimal_places=2),
            'quantity_increase_pct': ensure_native_type(quantity_increase_pct, decimal_places=2),
            'strategy_period_revenue_increase': ensure_native_type(strategy_period_revenue_increase, decimal_places=2),
            'strategy_period_revenue_increase_pct': ensure_native_type(strategy_period_revenue_increase_pct, decimal_places=2),
        },
        'strategy_range': {
            'start_idx': start_idx,
            'end_idx': end_idx,
            'start_date': date_series[start_idx],
            'end_date': date_series[end_idx],
        }
    }