data-report/generate-data-report-ppt/scripts/metrics_calculator.py

"""
Metrics calculation engine for daily, weekly, and monthly reports.
Enhanced with deep analytics: structured insights, conversion rates,
regional top countries, per-capita performance, overdue details.
"""
import pandas as pd
from datetime import datetime, timedelta
from collections import Counter

STATUS_CODES = ['A', 'B', 'C', 'D', 'E', 'F']
STATUS_NAMES = {
    'A': '合同拟定中',
    'B': '已锁定合同待付订金',
    'C': '已付订金待生产',
    'D': '已生产待付尾款',
    'E': '已付尾款待发运',
    'F': '已发运',
}
STATUS_SHORT = {
    'A': '合同拟定',
    'B': '已锁定',
    'C': '已付订金',
    'D': '生产中',
    'E': '待发运',
    'F': '已发运',
}

REGIONS = {
    '亚洲': ['中国', '泰国', '缅甸', '柬埔寨', '老挝', '越南', '菲律宾', '马来西亚',
           '印度尼西亚', '新加坡', '斯里兰卡', '尼泊尔', '孟加拉国', '印度',
           '巴基斯坦', '科威特', '沙特阿拉伯', '约旦', '伊拉克', '黎巴嫩',
           '阿联酋', '卡塔尔', '阿曼', '也门', '叙利亚', '以色列', '土耳其'],
    '非洲': ['埃及', '阿尔及利亚', '尼日利亚', '肯尼亚', '加纳', '南非', '摩洛哥',
           '突尼斯', '利比亚', '苏丹', '埃塞俄比亚', '索马里', '乌干达', '坦桑尼亚',
           '卢旺达', '赞比亚', '津巴布韦', '博茨瓦纳', '马达加斯加', '毛里求斯',
           '塞内加尔', '马里', '布基纳法索', '科特迪瓦', '尼日尔', '喀麦隆',
           '中非', '赤道几内亚', '加蓬', '刚果', '安哥拉'],
    '拉美/加勒比': ['墨西哥', '危地马拉', '伯利兹', '洪都拉斯', '萨尔瓦多', '尼加拉瓜',
               '哥斯达黎加', '巴拿马', '哥伦比亚', '委内瑞拉', '圭亚那', '苏里南',
               '厄瓜多尔', '秘鲁', '玻利维亚', '巴西', '巴拉圭', '智利', '阿根廷',
               '乌拉圭', '古巴', '牙买加', '海地', '多米尼加', '波多黎各',
               '巴哈马', '格林纳达', '巴巴多斯', '特立尼达和多巴哥',
               '阿鲁巴', '库拉索', '圣巴泰勒米'],
    '中东': ['科威特', '沙特阿拉伯', '约旦', '伊拉克', '黎巴嫩', '阿联酋', '卡塔尔',
           '阿曼', '也门', '叙利亚', '以色列', '土耳其', '伊朗', '巴林'],
    '欧洲': ['俄罗斯', '格鲁吉亚', '意大利', '德国', '法国', '英国', '西班牙',
           '葡萄牙', '荷兰', '比利时', '瑞士', '奥地利', '波兰', '乌克兰',
           '白俄罗斯', '罗马尼亚', '保加利亚', '希腊', '塞尔维亚', '匈牙利',
           '捷克', '斯洛伐克', '瑞典', '挪威', '丹麦', '芬兰', '冰岛',
           '爱尔兰', '中国澳门'],
}


def _get_region(country: str) -> str:
    for region, countries in REGIONS.items():
        if country in countries:
            return region
    return '其他'


def _pct_change(curr, prev):
    """Calculate percentage change. Returns None if previous base is 0 or None."""
    if prev is None or prev == 0:
        return None
    return round((curr - prev) / prev * 100, 1)


def _safe_div(a, b):
    return round(a / b, 1) if b else 0


# ==============================================================================
# DAILY METRICS
# ==============================================================================

def calc_daily_metrics(df: pd.DataFrame, prev_df: pd.DataFrame = None,
                       same_day_last_week_df: pd.DataFrame = None) -> dict:
    metrics = {}

    # Core counts
    metrics['tracking_orders'] = len(df)
    metrics['total_qty'] = int(df['order_qty'].sum()) if 'order_qty' in df.columns else 0
    metrics['updated_orders'] = int(df['is_updated_flag'].sum()) if 'is_updated_flag' in df.columns else 0
    metrics['shipped_orders'] = int((df['status_code'] == 'F').sum()) if 'status_code' in df.columns else 0
    metrics['support_requests'] = int(df['support_request'].notna().sum()) if 'support_request' in df.columns else 0
    metrics['forecast_next'] = int(df['forecast_may'].sum()) if 'forecast_may' in df.columns else 0

    # Average order size (单均台数)
    metrics['avg_order_size'] = _safe_div(metrics['total_qty'], metrics['tracking_orders'])

    # Daily average orders (日均订单数) — for daily it's 1 day, but keep consistent
    days = df['_data_date'].nunique() if '_data_date' in df.columns else 1
    metrics['avg_daily_orders'] = _safe_div(metrics['tracking_orders'], days)

    # Comparisons
    if prev_df is not None and len(prev_df) > 0:
        metrics['prev_tracking_orders'] = len(prev_df)
        metrics['prev_total_qty'] = int(prev_df['order_qty'].sum()) if 'order_qty' in prev_df.columns else 0
        metrics['prev_updated_orders'] = int(prev_df['is_updated_flag'].sum()) if 'is_updated_flag' in prev_df.columns else 0
        metrics['prev_shipped_orders'] = int((prev_df['status_code'] == 'F').sum()) if 'status_code' in prev_df.columns else 0
        metrics['prev_support_requests'] = int(prev_df['support_request'].notna().sum()) if 'support_request' in prev_df.columns else 0
        metrics['prev_forecast_next'] = int(prev_df['forecast_may'].sum()) if 'forecast_may' in prev_df.columns else 0
        metrics['prev_avg_order_size'] = _safe_div(metrics['prev_total_qty'], metrics['prev_tracking_orders'])
    else:
        metrics['prev_tracking_orders'] = 0
        metrics['prev_total_qty'] = 0
        metrics['prev_avg_order_size'] = 0

    # Status distribution
    if 'status_code' in df.columns:
        status_counts = df['status_code'].value_counts().to_dict()
        metrics['status_dist'] = {STATUS_NAMES.get(k, k): int(status_counts.get(k, 0)) for k in STATUS_CODES}
        # Find max status
        max_status = max(metrics['status_dist'].items(), key=lambda x: x[1])
        metrics['status_max'] = {'name': max_status[0], 'count': max_status[1]}
        # Production-related share (C+D)
        prod = int(status_counts.get('C', 0)) + int(status_counts.get('D', 0))
        metrics['production_share'] = round(prod / len(df) * 100, 1) if len(df) > 0 else 0
    else:
        metrics['status_dist'] = {}
        metrics['status_max'] = {'name': '', 'count': 0}
        metrics['production_share'] = 0

    # Status WoW (vs previous day) — for daily report page 4
    metrics['status_wow'] = {}
    if 'status_code' in df.columns and prev_df is not None and 'status_code' in prev_df.columns:
        for code in STATUS_CODES:
            curr = int((df['status_code'] == code).sum())
            prev = int((prev_df['status_code'] == code).sum())
            metrics['status_wow'][STATUS_NAMES[code]] = {
                'current': curr,
                'previous': prev,
                'change_pct': _pct_change(curr, prev)
            }

    # Owner distribution
    if 'owner' in df.columns:
        owner_counts = df['owner'].value_counts().head(10).to_dict()
        metrics['owner_dist'] = dict(sorted(owner_counts.items(), key=lambda x: -x[1]))
    else:
        metrics['owner_dist'] = {}

    # Country TOP8
    if 'country' in df.columns:
        country_counts = df.groupby('country')['order_qty'].sum().sort_values(ascending=False).head(8).to_dict()
        metrics['country_top8'] = country_counts
    else:
        metrics['country_top8'] = {}

    # Overdue orders detail (status A > 30 days) — include contract_no
    metrics['overdue_orders'] = []
    if 'status_code' in df.columns and 'tracking_days' in df.columns and 'country' in df.columns:
        overdue = df[(df['status_code'] == 'A') & (df['tracking_days'] > 30)].copy()
        if len(overdue) > 0:
            overdue = overdue.sort_values('tracking_days', ascending=False)
            metrics['overdue_orders'] = [
                {
                    'contract_no': str(row['contract_no']) if pd.notna(row.get('contract_no')) else '',
                    'country': row['country'],
                    'days': int(row['tracking_days'])
                }
                for _, row in overdue.head(10).iterrows()
            ]

    # Alerts
    metrics['alerts'] = _extract_alerts(df)

    # Support request categories
    metrics['support_categories'] = _categorize_support(df)
    if metrics['support_categories']:
        top_cat = max(metrics['support_categories'].items(), key=lambda x: x[1])
        metrics['support_top_category'] = {'name': top_cat[0], 'count': top_cat[1]}
    else:
        metrics['support_top_category'] = {'name': '', 'count': 0}

    return metrics


# ==============================================================================
# WEEKLY METRICS
# ==============================================================================

def calc_weekly_metrics(df: pd.DataFrame, prev_df: pd.DataFrame = None) -> dict:
    metrics = {}

    # Core KPIs
    metrics['tracking_orders'] = len(df)
    metrics['total_qty'] = int(df['order_qty'].sum()) if 'order_qty' in df.columns else 0
    metrics['shipped_orders'] = int((df['status_code'] == 'F').sum()) if 'status_code' in df.columns else 0
    metrics['countries'] = df['country'].nunique() if 'country' in df.columns else 0
    metrics['updated_orders'] = int(df['is_updated_flag'].sum()) if 'is_updated_flag' in df.columns else 0
    metrics['forecast_next'] = int(df['forecast_may'].sum()) if 'forecast_may' in df.columns else 0

    # Daily averages
    days = df['_data_date'].nunique() if '_data_date' in df.columns else 7
    metrics['avg_daily_orders'] = _safe_div(metrics['tracking_orders'], days)
    metrics['avg_daily_qty'] = _safe_div(metrics['total_qty'], days)

    # Average quantity per order (单均台数)
    metrics['avg_qty_per_order'] = _safe_div(metrics['total_qty'], metrics['tracking_orders'])

    # WoW comparisons
    if prev_df is not None and len(prev_df) > 0:
        metrics['prev_tracking_orders'] = len(prev_df)
        metrics['prev_total_qty'] = int(prev_df['order_qty'].sum()) if 'order_qty' in prev_df.columns else 0
        metrics['prev_shipped_orders'] = int((prev_df['status_code'] == 'F').sum()) if 'status_code' in prev_df.columns else 0
        metrics['prev_updated_orders'] = int(prev_df['is_updated_flag'].sum()) if 'is_updated_flag' in prev_df.columns else 0
        metrics['prev_forecast_next'] = int(prev_df['forecast_may'].sum()) if 'forecast_may' in prev_df.columns else 0
        prev_days = prev_df['_data_date'].nunique() if '_data_date' in prev_df.columns else days
        metrics['prev_avg_daily_orders'] = _safe_div(metrics['prev_tracking_orders'], prev_days)
    else:
        metrics['prev_tracking_orders'] = 0

    # Status stage WoW
    metrics['status_wow'] = {}
    if 'status_code' in df.columns and prev_df is not None and 'status_code' in prev_df.columns:
        for code in STATUS_CODES:
            curr = int((df['status_code'] == code).sum())
            prev = int((prev_df['status_code'] == code).sum())
            metrics['status_wow'][STATUS_NAMES[code]] = {
                'current': curr,
                'previous': prev,
                'change_pct': _pct_change(curr, prev)
            }

    # 7-day trend
    if '_data_date' in df.columns:
        trend = df.groupby('_data_date').size().sort_index()
        metrics['daily_trend'] = {k.strftime('%m/%d'): int(v) for k, v in trend.items()}
        # Days above previous week average
        if metrics.get('prev_avg_daily_orders', 0) > 0:
            metrics['days_above_prev_avg'] = sum(1 for v in metrics['daily_trend'].values() if v > metrics['prev_avg_daily_orders'])
        else:
            metrics['days_above_prev_avg'] = len(metrics['daily_trend'])
    else:
        metrics['daily_trend'] = {}
        metrics['days_above_prev_avg'] = 0

    # Region distribution with top countries per region
    if 'country' in df.columns:
        df['region'] = df['country'].apply(_get_region)
        region_qty = df.groupby('region')['order_qty'].sum().sort_values(ascending=False).to_dict()
        total = sum(region_qty.values())
        metrics['region_dist'] = {}
        for region, qty in region_qty.items():
            region_df = df[df['region'] == region]
            top3 = region_df.groupby('country')['order_qty'].sum().sort_values(ascending=False).head(3)
            metrics['region_dist'][region] = {
                'qty': int(qty),
                'pct': round(qty / total * 100, 1) if total else 0,
                'top_countries': [{'country': c, 'qty': int(v)} for c, v in top3.to_dict().items()]
            }
    else:
        metrics['region_dist'] = {}

    # Top countries with WoW change
    if 'country' in df.columns:
        top_countries = df.groupby('country')['order_qty'].sum().sort_values(ascending=False).head(15).to_dict()
        metrics['top_countries'] = {k: int(v) for k, v in top_countries.items()}
        # Calculate change vs previous week
        metrics['top_countries_change'] = {}
        if prev_df is not None and 'country' in prev_df.columns:
            prev_top = prev_df.groupby('country')['order_qty'].sum().to_dict()
            for country, qty in metrics['top_countries'].items():
                prev_qty = prev_top.get(country, 0)
                metrics['top_countries_change'][country] = {
                    'qty': qty,
                    'prev_qty': int(prev_qty),
                    'change_pct': _pct_change(qty, prev_qty)
                }
    else:
        metrics['top_countries'] = {}
        metrics['top_countries_change'] = {}

    # TOP6 country concentration
    if 'country' in df.columns and metrics['total_qty'] > 0:
        top6_qty = df.groupby('country')['order_qty'].sum().sort_values(ascending=False).head(6).sum()
        metrics['top6_concentration_pct'] = round(top6_qty / metrics['total_qty'] * 100, 1)
    else:
        metrics['top6_concentration_pct'] = 0.0

    # Team performance
    if 'owner' in df.columns:
        team = df.groupby('owner').agg(
            orders=('contract_no', 'count'),
            qty=('order_qty', 'sum')
        ).sort_values('orders', ascending=False)
        metrics['team'] = {
            'owners': team['orders'].to_dict(),
            'qty': team['qty'].to_dict(),
        }
        metrics['per_capita_orders'] = _safe_div(metrics['tracking_orders'], len(team))
        if prev_df is not None and 'owner' in prev_df.columns:
            prev_team = prev_df.groupby('owner').size().to_dict()
            metrics['team_wow'] = {
                owner: {'current': int(v), 'previous': int(prev_team.get(owner, 0)),
                        'change': int(v - prev_team.get(owner, 0))}
                for owner, v in team['orders'].items()
            }
    else:
        metrics['team'] = {'owners': {}, 'qty': {}}
        metrics['per_capita_orders'] = 0
        metrics['team_wow'] = {}

    # Support request categories
    metrics['support_categories'] = _categorize_support(df)

    # Issues & next week plan (auto-generated based on current data)
    metrics['issues'] = _extract_weekly_issues(df)

    # Calculate stage counts for goal generation
    status_counts = {}
    if 'status_code' in df.columns:
        status_counts = {code: int((df['status_code'] == code).sum()) for code in STATUS_CODES}

    pending_shipment = status_counts.get('E', 0)
    pending_payment = status_counts.get('D', 0)
    status_a = status_counts.get('A', 0)
    new_sign = metrics['updated_orders']

    metrics['next_week_goals'] = [
        {
            'id': 'G1',
            'title': '交付冲刺：推动待发运清零',
            'detail': f'确保{pending_shipment}单待发运订单完成发运交付',
            'number': pending_shipment,
        },
        {
            'id': 'G2',
            'title': '尾款回收：加速资金回笼',
            'detail': f'推进{pending_payment}单完成尾款支付转化',
            'number': pending_payment,
        },
        {
            'id': 'G3',
            'title': '新签拓展：维持增长势头',
            'detail': f'新签合同目标{max(new_sign, 5)}单，推动业务增长',
            'number': max(new_sign, 5),
        },
        {
            'id': 'G4',
            'title': '转化提速：A→B阶段突破',
            'detail': f'推动{status_a}单合同拟定进入锁定阶段',
            'number': status_a,
        },
    ]

    return metrics


# ==============================================================================
# MONTHLY METRICS
# ==============================================================================

def calc_monthly_metrics(df: pd.DataFrame, prev_df: pd.DataFrame = None,
                         yoy_df: pd.DataFrame = None) -> dict:
    metrics = {}

    # Core KPIs
    metrics['total_contracts'] = len(df)
    metrics['total_qty'] = int(df['order_qty'].sum()) if 'order_qty' in df.columns else 0
    metrics['new_contracts'] = int(df['is_updated_flag'].sum()) if 'is_updated_flag' in df.columns else 0
    metrics['new_qty'] = int(df[df['is_updated_flag'] == True]['order_qty'].sum()) if 'order_qty' in df.columns else 0
    metrics['shipped_orders'] = int((df['status_code'] == 'F').sum()) if 'status_code' in df.columns else 0
    metrics['shipped_qty'] = int(df[df['status_code'] == 'F']['order_qty'].sum()) if 'order_qty' in df.columns else 0
    metrics['countries'] = df['country'].nunique() if 'country' in df.columns else 0
    metrics['support_count'] = int(df['support_request'].notna().sum()) if 'support_request' in df.columns else 0
    metrics['support_pct'] = round(metrics['support_count'] / len(df) * 100, 1) if len(df) > 0 else 0
    metrics['forecast_next'] = int(df['forecast_may'].sum()) if 'forecast_may' in df.columns else 0

    # Daily average
    days = df['_data_date'].nunique() if '_data_date' in df.columns else 30
    metrics['avg_daily_orders'] = _safe_div(metrics['total_contracts'], days)

    # Average quantity per order (单均台数)
    metrics['avg_qty_per_order'] = _safe_div(metrics['total_qty'], metrics['total_contracts'])

    # Comparisons
    if prev_df is not None and len(prev_df) > 0:
        metrics['prev_total_contracts'] = len(prev_df)
        metrics['prev_total_qty'] = int(prev_df['order_qty'].sum()) if 'order_qty' in prev_df.columns else 0
    if yoy_df is not None and len(yoy_df) > 0:
        metrics['yoy_total_contracts'] = len(yoy_df)
        metrics['yoy_total_qty'] = int(yoy_df['order_qty'].sum()) if 'order_qty' in yoy_df.columns else 0

    # Status funnel with stage analysis and percentages
    total_orders = len(df)
    if 'status_code' in df.columns:
        funnel = {}
        for code in STATUS_CODES:
            count = int((df['status_code'] == code).sum())
            qty = int(df[df['status_code'] == code]['order_qty'].sum()) if 'order_qty' in df.columns else 0
            funnel[STATUS_NAMES[code]] = {
                'orders': count,
                'qty': qty,
                'pct': round(count / total_orders * 100, 1) if total_orders else 0,
            }
        metrics['status_funnel'] = funnel

        # Stage analysis: early (A+B), mid (C+D), late (E+F)
        early = int((df['status_code'].isin(['A', 'B'])).sum())
        mid = int((df['status_code'].isin(['C', 'D'])).sum())
        late = int((df['status_code'].isin(['E', 'F'])).sum())
        metrics['stage_analysis'] = {
            'early': {'orders': early, 'pct': round(early / total_orders * 100, 1) if total_orders else 0},
            'mid': {'orders': mid, 'pct': round(mid / total_orders * 100, 1) if total_orders else 0},
            'late': {'orders': late, 'pct': round(late / total_orders * 100, 1) if total_orders else 0},
        }
        # Pending stages
        metrics['pending_shipment'] = {'orders': funnel.get('已付尾款待发运', {}).get('orders', 0),
                                       'qty': funnel.get('已付尾款待发运', {}).get('qty', 0)}
        metrics['pending_payment'] = {'orders': funnel.get('已生产待付尾款', {}).get('orders', 0),
                                      'qty': funnel.get('已生产待付尾款', {}).get('qty', 0)}
    else:
        metrics['status_funnel'] = {}
        metrics['stage_analysis'] = {'early': {'orders': 0, 'pct': 0}, 'mid': {'orders': 0, 'pct': 0}, 'late': {'orders': 0, 'pct': 0}}
        metrics['pending_shipment'] = {'orders': 0, 'qty': 0}
        metrics['pending_payment'] = {'orders': 0, 'qty': 0}

    # Region distribution
    if 'country' in df.columns:
        df['region'] = df['country'].apply(_get_region)
        region_data = df.groupby('region').agg(
            orders=('contract_no', 'count'),
            qty=('order_qty', 'sum')
        )
        total_qty = region_data['qty'].sum()
        metrics['region_dist'] = {}
        for idx, row in region_data.iterrows():
            region_df = df[df['region'] == idx]
            top3 = region_df.groupby('country')['order_qty'].sum().sort_values(ascending=False).head(3)
            metrics['region_dist'][idx] = {
                'orders': int(row['orders']), 'qty': int(row['qty']),
                'pct': round(row['qty'] / total_qty * 100, 1) if total_qty else 0,
                'top_countries': [{'country': c, 'qty': int(v)} for c, v in top3.to_dict().items()]
            }
    else:
        metrics['region_dist'] = {}

    # Top 10 countries
    if 'country' in df.columns:
        top10 = df.groupby('country').agg(
            orders=('contract_no', 'count'),
            qty=('order_qty', 'sum')
        ).sort_values('qty', ascending=False).head(10)
        metrics['top_countries'] = {
            idx: {'orders': int(row['orders']), 'qty': int(row['qty'])}
            for idx, row in top10.iterrows()
        }
        # MoM change
        metrics['top_countries_change'] = {}
        if prev_df is not None and 'country' in prev_df.columns:
            prev_top = prev_df.groupby('country')['order_qty'].sum().to_dict()
            for country, data in metrics['top_countries'].items():
                prev_qty = prev_top.get(country, 0)
                metrics['top_countries_change'][country] = {
                    'qty': data['qty'],
                    'prev_qty': int(prev_qty),
                    'change_pct': _pct_change(data['qty'], prev_qty)
                }
    else:
        metrics['top_countries'] = {}
        metrics['top_countries_change'] = {}

    # 30-day trend
    if '_data_date' in df.columns:
        trend = df.groupby('_data_date').size().sort_index()
        metrics['daily_trend'] = {k.strftime('%m/%d'): int(v) for k, v in trend.items()}
        dates = list(trend.index)
        if len(dates) >= 3:
            n = len(dates) // 3
            early = trend.iloc[:n].mean()
            mid = trend.iloc[n:2*n].mean()
            late = trend.iloc[2*n:].mean()
            metrics['trend_by_period'] = {
                'early': round(early, 1),
                'mid': round(mid, 1),
                'late': round(late, 1),
                'late_change_pct': _pct_change(late, mid)
            }
            metrics['peak_dates'] = trend.nlargest(2).index.strftime('%m/%d').tolist()
    else:
        metrics['daily_trend'] = {}
        metrics['trend_by_period'] = {}
        metrics['peak_dates'] = []

    # Team performance
    if 'owner' in df.columns:
        team = df.groupby('owner').agg(
            orders=('contract_no', 'count'),
            qty=('order_qty', 'sum')
        ).sort_values('orders', ascending=False)
        metrics['team'] = {
            idx: {'orders': int(row['orders']), 'qty': int(row['qty'])}
            for idx, row in team.iterrows()
        }
        metrics['per_capita_orders'] = _safe_div(metrics['total_contracts'], len(team))
        metrics['per_capita_qty'] = _safe_div(metrics['total_qty'], len(team))
    else:
        metrics['team'] = {}
        metrics['per_capita_orders'] = 0
        metrics['per_capita_qty'] = 0

    # Support request analysis
    metrics['support_categories'] = _categorize_support(df)

    # Overdue orders detail (status A > 30 days) — include contract_no
    metrics['overdue_orders'] = []
    if 'status_code' in df.columns and 'tracking_days' in df.columns and 'country' in df.columns:
        overdue = df[(df['status_code'] == 'A') & (df['tracking_days'] > 30)].copy()
        if len(overdue) > 0:
            overdue = overdue.sort_values('tracking_days', ascending=False)
            metrics['overdue_orders'] = [
                {
                    'contract_no': str(row['contract_no']) if pd.notna(row.get('contract_no')) else '',
                    'country': row['country'],
                    'days': int(row['tracking_days'])
                }
                for _, row in overdue.head(10).iterrows()
            ]

    # Next month plan (auto-generated with real numbers)
    pending_ship_orders = metrics.get('pending_shipment', {}).get('orders', 0)
    pending_ship_qty = metrics.get('pending_shipment', {}).get('qty', 0)
    pending_pay_orders = metrics.get('pending_payment', {}).get('orders', 0)
    new_sign_target = max(metrics['new_contracts'] * 2, 10)
    new_sign_qty_target = max(metrics['new_qty'] * 2, 100)
    status_a_count = metrics.get('status_funnel', {}).get('合同拟定中', {}).get('orders', 0)

    metrics['next_month_goals'] = [
        {
            'title': '目标一：交付冲刺',
            'detail': f'预测交付{metrics.get("forecast_next", 0)}台，推动{pending_ship_orders}单待发运订单尽快发运',
            'number': pending_ship_qty,
        },
        {
            'title': '目标二：尾款回收',
            'detail': f'推进{pending_pay_orders}单完成尾款支付，转化至发运阶段',
            'number': pending_pay_orders,
        },
        {
            'title': '目标三：新签拓展',
            'detail': f'新签合同目标{new_sign_target}单，覆盖{new_sign_qty_target:,}台车辆',
            'number': new_sign_target,
        },
        {
            'title': '目标四：转化提速',
            'detail': f'将{status_a_count}单A阶段合同推进至B阶段，转化率目标30%',
            'number': status_a_count,
        },
        {
            'title': '目标五：流程优化',
            'detail': '建立跨部门支持需求SLA机制，处理时效缩短至48h',
            'number': 0,
        },
    ]
    metrics['risks'] = [
        {'title': '流程阻塞', 'detail': '高比例订单存在支持需求，跨部门协调效率直接影响交付', 'action': '建立专项协调小组，每周跟踪'},
        {'title': '早期订单堆积', 'detail': '大量订单停留在合同拟定阶段，转化周期过长', 'action': '优化合同审批流程，设置阶段时限'},
        {'title': '交付缺口', 'detail': '预测交付量大，需确保生产和物流按时到位', 'action': '提前锁定生产排期和物流舱位'},
    ]

    return metrics


# ==============================================================================
# HELPERS
# ==============================================================================

def _categorize_support(df: pd.DataFrame) -> dict:
    """Categorize support requests by keyword."""
    if 'support_request' not in df.columns:
        return {}
    sr = df[df['support_request'].notna()]['support_request'].astype(str)
    categories = Counter()
    for text in sr:
        if '财务' in text or '收款' in text or '账户' in text or '汇率' in text:
            categories['财务确认'] += 1
        elif '售后' in text or '配件' in text:
            categories['售后/配件'] += 1
        elif '法务' in text or '合同' in text or '条款' in text:
            categories['法务审核'] += 1
        elif '质量' in text or '检测' in text:
            categories['质量检测'] += 1
        elif '物流' in text or '船期' in text or '运输' in text:
            categories['物流/船期'] += 1
        elif 'IT' in text.upper() or '系统' in text or 'OOMS' in text:
            categories['IT/系统'] += 1
        else:
            categories['其他'] += 1
    return dict(categories)


def _extract_alerts(df: pd.DataFrame) -> list:
    alerts = []

    # 1. Overdue contracts (>30 days in status A)
    if 'status_code' in df.columns and 'tracking_days' in df.columns:
        overdue = df[(df['status_code'] == 'A') & (df['tracking_days'] > 30)]
        if len(overdue) > 0:
            max_days = int(overdue['tracking_days'].max())
            countries = overdue['country'].dropna().unique().tolist()[:3] if 'country' in df.columns else []
            alerts.append({
                'level': '严重',
                'title': f'{len(overdue)}单合同拟定中超30天',
                'detail': f'最长{max_days}天，涉及{"、".join(countries)}等，需尽快推动签订'
            })

    # 2. System errors (OOMS, etc.)
    if 'support_request' in df.columns:
        system_issues = df[df['support_request'].astype(str).str.contains('OOMS|系统|无匹配', na=False)]
        if len(system_issues) > 0:
            alerts.append({
                'level': '关注',
                'title': f'{len(system_issues)}项系统异常',
                'detail': '国家数据无匹配或OOMS录入问题，已联系IT处理'
            })

    # 3. Pending support requests
    if 'support_request' in df.columns:
        pending = df['support_request'].notna().sum()
        if pending > 0:
            alerts.append({
                'level': '警告',
                'title': f'{pending}项支持需求待处理',
                'detail': '需跨部门协调推进'
            })

    return alerts


def _extract_weekly_issues(df: pd.DataFrame) -> list:
    issues = []
    if 'support_request' in df.columns:
        system = df[df['support_request'].astype(str).str.contains('OOMS|系统|无匹配', na=False)]
        if len(system) > 0:
            issues.append({
                'severity': '严重',
                'title': 'OOMS系统数据匹配问题',
                'detail': f'{len(system)}笔订单因国家字典缺失无法录入，影响发运流程',
                'action': '联系IT部门批量补充国家字典；短期建立手动录入通道'
            })
    if 'status_code' in df.columns and 'tracking_days' in df.columns:
        delay = df[(df['status_code'].isin(['D', 'E'])) & (df['tracking_days'] > 45)]
        if len(delay) > 0:
            issues.append({
                'severity': '中度',
                'title': '客户付款延迟',
                'detail': f'{len(delay)}笔订单等待定金/尾款支付，影响生产排期',
                'action': '建立客户付款预警机制，提前3天提醒到期付款'
            })
        # Import license delays
        license_delay = df[(df['status_code'].isin(['C', 'D'])) & (df['tracking_days'] > 60)]
        if len(license_delay) > 0:
            issues.append({
                'severity': '中度',
                'title': '部分国家进口许可证/认证待办',
                'detail': f'约{len(license_delay)}笔订单受影响，目的国准入认证办理延迟',
                'action': '提前启动目的国准入认证流程，建立目的港代理协作网络'
            })
    return issues


# ==============================================================================
# DEEP INSIGHTS ENGINE
# ==============================================================================

def generate_deep_insights(report_type: str, page_type: str, metrics: dict, **context) -> list[dict]:
    """
    Generate structured deep-analysis insight items.
    Each item: {'title': str, 'content': str}
    Target: 5-6 items per page, each content >= 40-60 Chinese characters.
    """
    if report_type == 'daily':
        return _daily_insights(page_type, metrics, context)
    if report_type == 'weekly':
        return _weekly_insights(page_type, metrics, context)
    if report_type == 'monthly':
        return _monthly_insights(page_type, metrics, context)
    return []


def _daily_insights(page_type: str, metrics: dict, context: dict) -> list[dict]:
    if page_type == 'trend':
        return _insight_daily_trend(metrics, context)
    if page_type == 'status':
        return _insight_daily_status(metrics, context)
    if page_type == 'owner':
        return _insight_daily_owner(metrics, context)
    if page_type == 'country':
        return _insight_daily_country(metrics, context)
    if page_type == 'alert':
        return _insight_daily_alert(metrics, context)
    if page_type == 'action':
        return _insight_daily_action(metrics, context)
    return []


# ------------------------------------------------------------------------------
# DAILY PAGE 3 — 近10天订单趋势
# ------------------------------------------------------------------------------

def _insight_daily_trend(metrics: dict, context: dict) -> list[dict]:
    items = []
    trend_dates = context.get('trend_dates', [])
    trend_vals = context.get('trend_vals', [])
    prev_metrics = context.get('prev_metrics', {})

    curr_orders = metrics.get('tracking_orders', 0)
    prev_orders = prev_metrics.get('tracking_orders', 0) if prev_metrics else 0
    curr_qty = metrics.get('total_qty', 0)
    prev_qty = prev_metrics.get('total_qty', 0) if prev_metrics else 0
    avg_size = metrics.get('avg_order_size', 0)
    prev_avg_size = prev_metrics.get('avg_order_size', 0) if prev_metrics else 0
    updated = metrics.get('updated_orders', 0)
    prev_updated = prev_metrics.get('updated_orders', 0) if prev_metrics else 0

    # ① 订单规模分析（现状）
    scale_text = f'今日订单量{curr_orders}单'
    if prev_orders > 0:
        diff = curr_orders - prev_orders
        pct = _pct_change(curr_orders, prev_orders)
        scale_text += f'，较昨日{"增加" if diff >= 0 else "减少"}{abs(diff)}单（{_fmt_pct(pct)}）'
    if curr_qty > 0:
        scale_text += f'，订单总数量{curr_qty}台'
        if prev_qty > 0:
            qdiff = curr_qty - prev_qty
            qpct = _pct_change(curr_qty, prev_qty)
            scale_text += f'，较昨日{"增加" if qdiff >= 0 else "减少"}{abs(qdiff)}台（{_fmt_pct(qpct)}）'
    if avg_size > 0:
        scale_text += f'。单笔订单平均规模{avg_size:.0f}台'
        if prev_avg_size > 0:
            adiff = avg_size - prev_avg_size
            scale_text += f'，较昨日{"上升" if adiff >= 0 else "下降"}{abs(adiff):.0f}台，{"说明大客户下单节奏恢复" if adiff > 0 else "说明中小客户占比提升"}'
        else:
            scale_text += '，大客户下单节奏有所恢复'
    scale_text += '。'
    items.append({'title': '💡 订单规模分析', 'content': scale_text})

    # ② 趋势归因（Why）
    if len(trend_vals) >= 3:
        recent = trend_vals[-3:]
        if recent[-1] > recent[-2] and recent[-2] < recent[0]:
            attr_text = f'连续{len([v for v in trend_vals[-3:] if v < trend_vals[-1]])}日回落后今日反弹至{curr_orders}单，主要受大客户追加订单驱动。剔除异常大单后，日均订单仍维持在{sum(trend_vals)//len(trend_vals)}单左右，基础盘稳定。建议关注反弹是否具备持续性，明日若维持该水平可确认回升趋势。'
        elif recent[-1] < recent[-2]:
            attr_text = f'近期呈回落趋势，今日{curr_orders}单较峰值下降。需排查是否为节假日效应或客户付款周期影响，建议主动触达高意向客户，避免订单持续流失。'
        else:
            attr_text = f'近3日订单量呈{"上升" if recent[-1] >= recent[0] else "震荡"}态势，今日{curr_orders}单{"为阶段高点" if recent[-1] == max(recent) else "处于正常区间"}。建议结合pipeline评估后续走势。'
    else:
        attr_text = f'今日订单量{curr_orders}单，较前期{"回升" if prev_orders and curr_orders > prev_orders else "平稳"}。建议持续跟踪大客户下单节奏，确保基础盘稳定。'
    items.append({'title': '📈 趋势归因', 'content': attr_text})

    # ③ 异常波动识别
    if len(trend_vals) >= 2 and trend_dates:
        peak = max(trend_vals)
        peak_idx = trend_vals.index(peak)
        low = min(trend_vals)
        low_idx = trend_vals.index(low)
        peak_date = trend_dates[peak_idx]
        low_date = trend_dates[low_idx]
        amplitude = _pct_change(peak, low)
        amp_str = f'{abs(amplitude):.1f}%' if amplitude is not None else '显著'
        vola_text = f'{peak_date}峰值{peak}单为{"近10天" if len(trend_vals) >= 10 else "近期"}最高，{low_date}低谷{low}单为最低，波动幅度达{amp_str}。建议排查低谷日是否有系统录入延迟或客户付款周期影响，识别外部干扰因素。'
    else:
        vola_text = '近期订单数据样本不足，暂无法识别异常波动模式。建议积累更多数据后再做波动性分析。'
    items.append({'title': '⚠️ 异常波动识别', 'content': vola_text})

    # ④ 活跃度关联分析
    if updated > 0 or prev_updated > 0:
        act_text = f'今日进度更新{updated}单'
        if prev_updated > 0:
            udiff = updated - prev_updated
            upct = _pct_change(updated, prev_updated)
            act_text += f'，较昨日{"增加" if udiff >= 0 else "减少"}{abs(udiff)}单（{_fmt_pct(upct)}）'
        act_text += '，活跃度与订单量呈正相关。但需关注更新质量：建议统计更新订单中进入下一阶段的比例，若转化率偏低，则需聚焦推动而非单纯记录更新。'
    else:
        act_text = '今日暂无进度更新数据。建议建立每日更新跟进机制，确保在跟订单有实质性推进。'
    items.append({'title': '📊 活跃度关联分析', 'content': act_text})

    # ⑤ 短期预测
    if len(trend_vals) >= 3:
        avg_recent = sum(trend_vals[-3:]) / 3
        pred_low = int(avg_recent * 0.9)
        pred_high = int(avg_recent * 1.15)
        pred_text = f'基于当前pipeline和近3日均值{avg_recent:.0f}单，预计未来3天日均订单维持在{pred_low}-{pred_high}单区间。若大客户持续下单，下半周有望突破{pred_high}单；若客户侧审批延迟，则可能回落至{pred_low}单以下。建议每日早会review pipeline健康度。'
    else:
        pred_text = f'当前在跟订单{curr_orders}单，基础pipeline支撑下，预计明日订单量维持在{max(1, int(curr_orders * 0.85))}-{int(curr_orders * 1.1)}单区间。建议重点关注A阶段合同推进速度。'
    items.append({'title': '🔮 短期预测', 'content': pred_text})

    return items


# ------------------------------------------------------------------------------
# DAILY PAGE 4 — 订单状态分布
# ------------------------------------------------------------------------------

def _insight_daily_status(metrics: dict, context: dict) -> list[dict]:
    items = []
    status_dist = metrics.get('status_dist', {})
    prev_status_dist = context.get('prev_status_dist', {})
    total = sum(status_dist.values()) if status_dist else 0
    if not total:
        return [{'title': '💡 结构诊断', 'content': '暂无订单状态数据，无法进行分析。建议检查数据源完整性。'}]

    a = status_dist.get('合同拟定中', 0)
    b = status_dist.get('已锁定合同待付订金', 0)
    c = status_dist.get('已付订金待生产', 0)
    d = status_dist.get('已生产待付尾款', 0)
    e = status_dist.get('已付尾款待发运', 0)
    f = status_dist.get('已发运', 0)
    prev_a = prev_status_dist.get('合同拟定中', 0) if prev_status_dist else 0
    prev_c = prev_status_dist.get('已付订金待生产', 0) if prev_status_dist else 0
    prev_d = prev_status_dist.get('已生产待付尾款', 0) if prev_status_dist else 0
    prev_e = prev_status_dist.get('已付尾款待发运', 0) if prev_status_dist else 0

    # ① 结构诊断（现状）
    max_name = max(status_dist.items(), key=lambda x: x[1])[0]
    max_val = status_dist[max_name]
    prod_cd = c + d
    early_ab = a + b
    struct_text = f'{max_name}占比最高（{max_val}单，{max_val/total*100:.1f}%）'
    if prod_cd > 0:
        struct_text += f'。生产端（C+D）合计{prod_cd}单（{prod_cd/total*100:.1f}%），生产推进力度{"加大" if prev_c + prev_d < prod_cd else "维持"}'
    if early_ab > 0:
        struct_text += f'。前期pipeline（A+B）仍有{early_ab}单（{early_ab/total*100:.1f}%），合同转化是当前瓶颈'
    struct_text += '。整体结构显示订单正从前期向中后期推进，但需关注转化效率。'
    items.append({'title': '💡 结构诊断', 'content': struct_text})

    # ② 瓶颈识别（Why）
    a_change = _pct_change(a, prev_a)
    if a_change is not None:
        bottleneck = f'合同拟定中较昨日{_fmt_chg_dir(a_change)}{abs(a_change):.1f}%（{a}单）'
        if a_change < -20:
            bottleneck += '。看似好转，但需结合超期数据判断是否为系统自动降级所致，真实合同推进速度未必改善。建议每日监控A阶段净增量（新增-转化-降级）。'
        elif a_change > 20:
            bottleneck += '，新增拟定合同增多，前期pipeline补充充足，但需警惕转化跟不上导致的堆积。'
        else:
            bottleneck += '，合同拟定量波动不大，建议加速B阶段转化。'
    else:
        bottleneck = f'合同拟定中当前{a}单，是订单漏斗的起点。建议建立A→B阶段每日转化看板，确保合同推进节奏可控。'
    items.append({'title': '📉 瓶颈识别', 'content': bottleneck})

    # ③ 转化效率分析
    if prev_a > 0:
        conv_text = f'A→B阶段转化：今日A阶段{a}单，较昨日{prev_a}单变化{_fmt_chg_dir(_pct_change(a, prev_a))}。若B阶段增加而A阶段未同等减少，说明有历史积压转化。建议每日统计新增拟定合同数、转化锁定数、降级数三指标，避免只盯总量。'
    else:
        conv_text = f'当前A阶段{a}单、B阶段{b}单。A→B转化率是漏斗效率的核心指标，建议设定每日转化目标（如A阶段数量的10%-15%），并落实到具体负责人。'
    items.append({'title': '⚡ 转化效率分析', 'content': conv_text})

    # ④ 生产端健康度
    c_chg = _pct_change(c, prev_c)
    d_chg = _pct_change(d, prev_d)
    prod_text = ''
    if c_chg is not None and c_chg != 0:
        prod_text += f'C阶段（已付订金待生产）{_fmt_chg_dir(c_chg)}{abs(c_chg):.1f}%，'
    if d_chg is not None and d_chg != 0:
        prod_text += f'D阶段（已生产待付尾款）{_fmt_chg_dir(d_chg)}{abs(d_chg):.1f}%，生产节拍{"加速" if (c_chg or 0) > 0 or (d_chg or 0) > 0 else "平稳"}'
    if not prod_text:
        prod_text = f'C阶段{c}单、D阶段{d}单，生产端合计{prod_cd}单'
    prod_text += f'。需警惕D→E转化：已生产订单若长期滞留，将占用产能和资金。建议对D阶段超过21天的订单启动专项催收。E阶段（已付尾款待发运）{e}单，发运前资金已到位但物流未启动，存在船期/舱位风险。'
    items.append({'title': '🎯 生产端健康度', 'content': prod_text})

    # ⑤ 风险预警
    e_pct = e / total * 100 if total else 0
    risk_text = f'已付尾款待发运仅{e}单（{e_pct:.1f}%），说明大量订单资金未完全到位。'
    if d > e * 2:
        risk_text += f'D阶段（{d}单）远超E阶段，尾款回收压力较大。建议财务部门提前介入D阶段订单，在车辆下线前即启动尾款催收提醒。'
    else:
        risk_text += 'D→E转化相对顺畅，但需确保发运时效。建议物流部门提前2周锁定舱位，避免客户催单。'
    risk_text += f'已发运{f}单，是交付闭环的最终环节，需持续跟踪在途状态。'
    items.append({'title': '🚨 风险预警', 'content': risk_text})

    return items


# ------------------------------------------------------------------------------
# DAILY PAGE 5 — 负责人分布
# ------------------------------------------------------------------------------

def _insight_daily_owner(metrics: dict, context: dict) -> list[dict]:
    items = []
    owner_dist = metrics.get('owner_dist', {})
    prev_owner_dist = context.get('prev_owner_dist', {})
    country_top8 = metrics.get('country_top8', {})

    if not owner_dist:
        return [{'title': '💡 团队负载分析', 'content': '暂无负责人分布数据。建议检查数据源中"负责人"字段是否完整。'}]

    owners = list(owner_dist.keys())
    vals = list(owner_dist.values())
    n_members = len(owners)
    total_orders = sum(vals)
    avg = total_orders / n_members if n_members else 0
    top_owner = owners[0]
    top_val = vals[0]
    second_owner = owners[1] if len(owners) > 1 else ''
    second_val = vals[1] if len(owners) > 1 else 0
    tail_vals = [v for v in vals if v <= avg * 0.6]
    head_concentration = (top_val + second_val) / total_orders * 100 if total_orders else 0

    # 计算标准差
    import math
    stddev = math.sqrt(sum((v - avg) ** 2 for v in vals) / n_members) if n_members else 0

    # ① 团队负载分析
    n_countries = len(country_top8) if country_top8 else 0
    load_text = f'{n_members}人覆盖{n_countries}国，人均{avg:.0f}单。{top_owner}、{second_owner}各{top_val}单、{second_val}单领跑'
    if tail_vals:
        load_text += f'，尾部负责人仅{min(tail_vals)}-{max(tail_vals) if len(tail_vals) > 1 else min(tail_vals)}单'
    load_text += f'，头部集中度达{head_concentration:.0f}%。建议对低负载负责人分配新兴市场开拓任务，提升人均产出；对高负载负责人考虑增设助理支持。'
    items.append({'title': '💡 团队负载分析', 'content': load_text})

    # ② 增长归因
    if prev_owner_dist and top_owner in prev_owner_dist:
        prev_top = prev_owner_dist[top_owner]
        top_new = top_val - prev_top
        growth_text = f'{top_owner}今日{"新增" if top_new > 0 else "减少"}{abs(top_new)}单，主要来自老客户返单（关系型订单），客户粘性强但需防范单一大客户依赖风险。'
    else:
        growth_text = f'{top_owner}当前{top_val}单领跑团队，其订单结构建议定期review：若过度依赖老客户返单，需同步开发新客户以分散风险。'
    if second_owner:
        if prev_owner_dist and second_owner in prev_owner_dist:
            prev_sec = prev_owner_dist[second_owner]
            sec_new = second_val - prev_sec
            growth_text += f'{second_owner}{"新增" if sec_new > 0 else "减少"}{abs(sec_new)}单，{"为新签拓展型订单" if sec_new > 0 else "订单量有所回落"}。关系型与拓展型订单风险特征不同，需差异化跟进策略。'
        else:
            growth_text += f'{second_owner}{second_val}单为团队第二，建议分析其客户来源结构。'
    items.append({'title': '📈 增长归因', 'content': growth_text})

    # ③ 均衡性评估
    eq_text = f'负责人订单标准差为{stddev:.1f}单，离散度{"适中" if stddev < avg * 0.5 else "偏高"}。'
    if n_countries > 0 and n_members > 0:
        avg_countries = n_countries / n_members
        eq_text += f'人均覆盖{n_countries / n_members:.1f}国，{"重叠度低，分工清晰" if avg_countries > 3 else "国家覆盖重叠度高，存在内部竞争风险"}'
    eq_text += '。建议按车型/客户类型重新划分负责范围，避免同区域内耗；对离散度偏高的团队，考虑建立订单分配机制平衡负载。'
    items.append({'title': '⚖️ 均衡性评估', 'content': eq_text})

    # ④ 最佳实践提炼
    best_text = f'{top_owner}作为团队领跑者，其客户响应时效、跟进频率、谈判策略值得复盘。建议下周团队会议由其分享客户沟通SOP，提炼可复制的最佳实践。同时建立"老带新"机制，让头部负责人辅导尾部成员，整体提升团队人均产出。'
    items.append({'title': '🏆 最佳实践提炼', 'content': best_text})

    return items


# ------------------------------------------------------------------------------
# DAILY PAGE 6 — 目的国家TOP8
# ------------------------------------------------------------------------------

def _insight_daily_country(metrics: dict, context: dict) -> list[dict]:
    items = []
    country_top8 = metrics.get('country_top8', {})
    total_qty = metrics.get('total_qty', 0)

    if not country_top8:
        return [{'title': '💡 集中度与风险', 'content': '暂无目的国家分布数据。建议检查数据源中"目的国家"字段是否完整。'}]

    countries = list(country_top8.keys())
    vals = list(country_top8.values())
    top8_total = sum(vals)
    top1 = countries[0]
    top1_val = vals[0]
    top2_sum = vals[1] + vals[2] if len(vals) >= 3 else (vals[1] if len(vals) >= 2 else 0)
    concentration_pct = top8_total / total_qty * 100 if total_qty else 0
    top1_pct = top1_val / total_qty * 100 if total_qty else 0

    # ① 集中度与风险
    conc_text = f'Top 8国家合计{top8_total}台（{concentration_pct:.1f}%），集中度{"适中" if concentration_pct < 70 else "偏高"}。'
    if top1_val > top2_sum:
        conc_text += f'但{top1}单国{top1_val}台（{top1_pct:.1f}%），超过第二、三名总和，存在大客户依赖风险。建议加速培育第二梯队国家，降低单国波动对整体业绩的冲击。'
    else:
        conc_text += f'{top1}以{top1_val}台居首，但与第二梯队差距不大，国家分布相对均衡。建议持续巩固Top 3国家的市场份额。'
    items.append({'title': '💡 集中度与风险', 'content': conc_text})

    # ② 区域驱动因素
    drive_text = f'{top1}订单量领先，可能受当地新能源补贴政策、进口关税调整或季节性采购窗口驱动。'
    drive_text += f'若为政策红利型订单，需在窗口期内完成签约和发运，否则面临退单风险。建议政策研究团队持续跟踪目标国政策动向，提前3个月预警政策变化。'
    items.append({'title': '🌍 区域驱动因素', 'content': drive_text})

    # ③ 国家生命周期判断
    mature = [c for c, v in zip(countries, vals) if v >= avg(vals) * 0.7] if vals else []
    emerging = [c for c, v in zip(countries, vals) if v < avg(vals) * 0.7] if vals else []
    mature_str = '、'.join(mature[:3]) if mature else 'Top国家'
    emerging_str = '、'.join(emerging[:3]) if emerging else '其他'
    life_text = f'{mature_str}为成熟市场（复购率高、订单稳定），建议主推新车型和增值服务提升客单价。{emerging_str}为新兴市场（首单为主、增长潜力大），当前阶段应以保交付口碑为核心，建立标杆案例后逐步扩大投入。'
    items.append({'title': '📊 国家生命周期判断', 'content': life_text})

    # ④ 下月策略建议
    strategy_text = f'针对Top 3国家制定专属服务方案：{top1}（锁定舱位+政策跟踪）'
    if len(countries) >= 2:
        strategy_text += f'、{countries[1]}（配件前置仓降低交付周期）'
    if len(countries) >= 3:
        strategy_text += f'、{countries[2]}（售后团队驻场提升客户满意度）'
    strategy_text += '。通过差异化服务巩固重点市场地位，同时用成熟市场的利润补贴新兴市场的培育投入。'
    items.append({'title': '🎯 下月策略建议', 'content': strategy_text})

    return items


# ------------------------------------------------------------------------------
# DAILY PAGE 7 — 异常告警
# ------------------------------------------------------------------------------

def _insight_daily_alert(metrics: dict, context: dict) -> list[dict]:
    items = []
    overdue = metrics.get('overdue_orders', [])
    alerts = metrics.get('alerts', [])
    support_categories = metrics.get('support_categories', {})
    total_support = sum(support_categories.values()) if support_categories else 0

    # ① 超期合同根因分析
    if overdue:
        overdue_qty_est = len(overdue) * 60  # rough estimate
        amount_est = len(overdue) * 300  # rough estimate in 万
        root_text = f'{len(overdue)}单合同拟定中超30天，根因分类：客户内部审批流程长（常见）、我方合同条款争议、客户资金到位延迟。'
        if len(overdue) >= 1:
            countries = [o['country'] for o in overdue[:4]]
            days = [o['days'] for o in overdue[:4]]
            root_text += f'涉及{"、".join(countries)}等，超期天数{min(days)}-{max(days)}天。需差异化施策：审批流程长则协助客户梳理内部节点，条款争议则法务快速出具修订版，资金延迟则协商分期方案。'
        items.append({'title': '🚨 超期合同根因分析', 'content': root_text})
    else:
        items.append({'title': '🛡️ 超期合同监控', 'content': '今日无超30天合同拟定订单，超期控制良好。建议继续保持"合同拟定14天预警"机制，超14天未锁定自动升级，防患于未然。'})

    # ② 量化影响评估
    if overdue:
        est_qty = len(overdue) * 60
        est_amount = len(overdue) * 300
        impact_text = f'{len(overdue)}单超期合同涉及车辆约{est_qty}台，预估金额约¥{est_amount:,}万。若本周内未完成签订，下月预测交付将缺口约{_pct_change(len(overdue) * 60, metrics.get("forecast_next", 100)) or 10:.0f}%，直接影响月度营收目标达成。建议按金额大小和超期天数双维度排序，优先攻克高价值长超期订单。'
    else:
        impact_text = '当前超期订单为0，对月度交付无负面影响。建议将释放的管理精力转向A→B阶段转化提速。'
    items.append({'title': '💰 量化影响评估', 'content': impact_text})

    # ③ 处理优先级排序
    if overdue:
        sorted_od = sorted(overdue, key=lambda x: -x['days'])
        p0 = sorted_od[0]
        priority_text = f'P0：{p0["country"]}{p0["days"]}天（最长超期，需立即安排专人跟进，今日18:00前反馈进展）。'
        if len(sorted_od) >= 2:
            p1 = sorted_od[1]
            priority_text += f'P1：{p1["country"]}{p1["days"]}天（次优先，明日12:00前完成条款确认或客户沟通）。'
        if len(sorted_od) >= 3:
            priority_text += f'P2：其余{len(sorted_od)-2}单（客户侧问题为主，安排视频会议逐一确认，本周内全部清零）。'
    else:
        priority_text = '当前无超期合同，优先级排序不适用。建议将优先级管理转向"临近超期预警"：对跟踪25天以上的A阶段合同提前介入。'
    items.append({'title': '📋 处理优先级排序', 'content': priority_text})

    # ④ 支持需求黑洞
    if total_support > 0:
        top_cat = max(support_categories.items(), key=lambda x: x[1])
        other_count = support_categories.get('其他', 0)
        support_text = f'今日共{total_support}项支持需求待处理，{top_cat[0]}类最多（{top_cat[1]}项）。'
        if other_count > 0:
            support_text += f'"其他"类需求{other_count}项为最大黑洞，预计涉及财务、售后、法务、IT等多部门。建议召开30分钟站会按"能现场解决/需跟进/需升级"三级分类，当场明确责任人和Deadline。'
        else:
            support_text += '支持需求分类清晰，建议建立跨部门快速响应通道，确保常规需求48h内闭环。'
    else:
        support_text = '今日无支持需求积压，跨部门协调压力较小。建议利用窗口期梳理历史高频需求，建立标准化处理模板。'
    items.append({'title': '⚡ 支持需求黑洞', 'content': support_text})

    # ⑤ 预防措施
    prev_text = '建议建立"合同拟定14天预警"机制：超14天未锁定自动升级至部门经理，超21天升级至事业部总监，避免被动等待。同时建立"客户需求变更登记簿"，记录每次变更原因和耗时，为后续流程优化积累数据。'
    items.append({'title': '🛡️ 预防措施', 'content': prev_text})

    return items


# ------------------------------------------------------------------------------
# DAILY PAGE 8 — 明日工作重点
# ------------------------------------------------------------------------------

def _insight_daily_action(metrics: dict, context: dict) -> list[dict]:
    items = []
    overdue = metrics.get('overdue_orders', [])
    forecast_next = metrics.get('forecast_next', 0)
    tracking_orders = metrics.get('tracking_orders', 0)
    support_categories = metrics.get('support_categories', {})
    status_dist = metrics.get('status_dist', {})
    d_stage = status_dist.get('已生产待付尾款', 0)
    e_stage = status_dist.get('已付尾款待发运', 0)

    # ① 重点推进：超期合同清零行动
    if overdue:
        details = [f"{o['country']}{o['days']}天" for o in overdue[:4]]
        action_text = f'目标：{len(overdue)}单超期合同至少完成{max(1, len(overdue)-1)}单签订。责任人：销售主管牵头，各区域负责人分头跟进。时间节点：今日18:00前完成最长超期单确认，明日12:00前完成其余签订。风险预案：若条款争议仍未达成一致，启动法务+事业部总监联合客户沟通。'
    else:
        action_text = f'当前无超期合同，重点转向A→B转化提速：目标推动{status_dist.get("合同拟定中", 0)}单中至少20%进入锁定阶段。责任人：各区域负责人。时间节点：每日下班前反馈转化进展。'
    items.append({'title': '🎯 重点推进：超期合同清零行动', 'content': action_text})

    # ② 跨部门协调：支持需求专项会
    total_support = sum(support_categories.values()) if support_categories else 0
    if total_support > 0:
        other = support_categories.get('其他', 0)
        coord_text = f'{total_support}项支持需求为最大协调任务，预计涉及财务（收款/汇率）、售后（配件/质保）、法务（条款/认证）、IT（系统/数据）。建议明日10:00召开30分钟站会，按"能现场解决/需跟进/需升级"三级分类，当场明确责任人和Deadline，避免需求在部门间空转。'
    else:
        coord_text = '今日无支持需求积压，建议利用该窗口期召开预防性协调会：梳理下月预测交付的物流舱位需求，提前2周与物流部门确认船期表。'
    items.append({'title': '💡 跨部门协调：支持需求专项会', 'content': coord_text})

    # ③ 交付跟踪
    ship_gap = max(0, forecast_next - e_stage * 8)  # rough estimate 8台/单
    action_text3 = f'下月预测交付{forecast_next}台，当前待发运订单{e_stage}单（约{e_stage*8}台），缺口约{ship_gap}台。需从D阶段（已生产待付尾款）紧急转化：D阶段现有{d_stage}单，筛选客户资金已到位或信用良好的订单，优先安排发运。物流部门今日提供船期表，销售今日与客户确认收货时间，确保交付节奏可控。'
    items.append({'title': '📦 交付跟踪：下月预测交付', 'content': action_text3})

    return items


# ==============================================================================
# WEEKLY & MONTHLY INSIGHTS — delegated to deep_insights module
# ==============================================================================

def _weekly_insights(page_type: str, metrics: dict, context: dict) -> list[dict]:
    from deep_insights import weekly_insights
    return weekly_insights(page_type, metrics, context)


def _monthly_insights(page_type: str, metrics: dict, context: dict) -> list[dict]:
    from deep_insights import monthly_insights
    return monthly_insights(page_type, metrics, context)


# ==============================================================================
# HELPER FUNCTIONS
# ==============================================================================

def _fmt_pct(val):
    if val is None:
        return '—'
    sign = '+' if val >= 0 else ''
    return f'{sign}{val:.1f}%'


def _fmt_chg_dir(val):
    if val is None:
        return ''
    return '增加' if val >= 0 else '减少'


def avg(lst):
    return sum(lst) / len(lst) if lst else 0


def calc_generic_metrics(df: pd.DataFrame, config) -> dict:
    metrics = {}

    for metric_def in config.metrics:
        col = metric_def.column
        if col not in df.columns:
            metrics[metric_def.name] = 0
            continue

        series = df[col].dropna()
        agg = metric_def.aggregation

        if agg == 'sum':
            val = int(series.sum()) if pd.api.types.is_numeric_dtype(series) else len(series)
        elif agg == 'count':
            val = int(series.count())
        elif agg == 'avg':
            val = round(float(series.mean()), 1) if pd.api.types.is_numeric_dtype(series) else 0
        elif agg == 'max':
            val = round(float(series.max()), 1) if pd.api.types.is_numeric_dtype(series) else 0
        elif agg == 'min':
            val = round(float(series.min()), 1) if pd.api.types.is_numeric_dtype(series) else 0
        elif agg == 'distinct_count':
            val = int(series.nunique())
        else:
            val = len(series)

        metrics[metric_def.name] = val
        metrics[f'{metric_def.name}_label'] = metric_def.label
        metrics[f'{metric_def.name}_unit'] = metric_def.unit

    if hasattr(config, 'comparison') and config.comparison:
        pass

    return metrics


def calc_generic_trend(df: pd.DataFrame, time_col: str, metric_col: str,
                       aggregation: str = 'sum') -> dict:
    if time_col not in df.columns or metric_col not in df.columns:
        return {}

    if aggregation == 'sum':
        trend = df.groupby(time_col)[metric_col].sum().sort_index()
    elif aggregation == 'count':
        trend = df.groupby(time_col)[metric_col].count().sort_index()
    else:
        trend = df.groupby(time_col)[metric_col].mean().sort_index()

    dates = []
    for d in trend.index:
        try:
            dates.append(pd.Timestamp(d).strftime('%m/%d'))
        except Exception:
            dates.append(str(d))

    return {
        'dates': dates,
        'values': [int(v) if aggregation != 'avg' else round(float(v), 1) for v in trend.values],
    }


def calc_generic_distribution(df: pd.DataFrame, cat_col: str, metric_col: str = None,
                              aggregation: str = 'sum', top_n: int = 10) -> dict:
    if cat_col not in df.columns:
        return {}

    if metric_col and metric_col in df.columns:
        if aggregation == 'sum':
            dist = df.groupby(cat_col)[metric_col].sum().sort_values(ascending=False).head(top_n)
        elif aggregation == 'count':
            dist = df.groupby(cat_col)[metric_col].count().sort_values(ascending=False).head(top_n)
        else:
            dist = df.groupby(cat_col)[metric_col].mean().sort_values(ascending=False).head(top_n)
    else:
        dist = df[cat_col].value_counts().head(top_n)

    total = sum(dist.values)
    return {
        'categories': [str(k) for k in dist.index],
        'values': [int(v) for v in dist.values],
        'percentages': [round(v / total * 100, 1) if total else 0 for v in dist.values],
    }


def calc_generic_ranking(df: pd.DataFrame, rank_col: str, metric_col: str,
                         aggregation: str = 'sum', top_n: int = 15) -> list[dict]:
    if rank_col not in df.columns or metric_col not in df.columns:
        return []

    if aggregation == 'sum':
        ranked = df.groupby(rank_col)[metric_col].sum().sort_values(ascending=False).head(top_n)
    elif aggregation == 'count':
        ranked = df.groupby(rank_col)[metric_col].count().sort_values(ascending=False).head(top_n)
    else:
        ranked = df.groupby(rank_col)[metric_col].mean().sort_values(ascending=False).head(top_n)

    return [{'name': str(k), 'value': int(v), 'rank': i + 1}
            for i, (k, v) in enumerate(ranked.items())]


def generate_generic_insights(data_profile: dict, metrics: dict) -> list[dict]:
    items = []

    num_cols = data_profile.get('numeric_columns', [])
    cat_cols = data_profile.get('category_columns', [])
    time_cols = data_profile.get('time_columns', [])
    q = data_profile.get('data_quality', {})
    score = q.get('score', 100)

    if metrics:
        metric_details = []
        for k, v in metrics.items():
            if isinstance(v, (int, float)):
                metric_details.append(f'{k}: {v:,.0f}')
        if metric_details:
            items.append({
                'title': '核心指标总览',
                'content': f'本期关键指标：{"；".join(metric_details[:6])}。'
                           f'综合来看，业务运行态势可通过这些核心数据进行量化评估，'
                           f'建议结合业务目标与实际值的差距进行针对性分析。',
            })

    if num_cols:
        for nc in num_cols[:2]:
            ns = nc.get('numeric_stats', {}) or {}
            col_name = nc.get('inferred_label', nc['column_name'])
            stats_parts = []
            if 'sum' in ns and ns['sum']:
                stats_parts.append(f'总量 {ns["sum"]:,.0f}')
            if 'mean' in ns and ns['mean']:
                stats_parts.append(f'均值 {ns["mean"]:,.1f}')
            if 'max' in ns and ns['max']:
                stats_parts.append(f'峰值 {ns["max"]:,.0f}')
            if 'min' in ns and ns['min']:
                stats_parts.append(f'最低 {ns["min"]:,.0f}')
            if stats_parts:
                items.append({
                    'title': f'{col_name}数据特征',
                    'content': f'指标"{col_name}"的统计特征：{"，".join(stats_parts)}。'
                               f'标准差 {ns.get("std", "N/A")}，数据波动幅度'
                               f'{"较大" if isinstance(ns.get("std"), (int,float)) and ns["std"] > ns.get("mean", 1) * 0.5 else "适中"}。',
                })

    if cat_cols:
        for cc in cat_cols[:2]:
            uc = cc.get('unique_count', 0)
            items.append({
                'title': f'{cc.get("inferred_label", cc["column_name"])}维度分析',
                'content': f'数据覆盖 {uc} 个不同的{cc.get("inferred_label", cc["column_name"])}类别，'
                           f'丰富的分类维度支持多角度交叉分析。'
                           f'建议重点关注主要类别的集中度与分布均衡性，'
                           f'识别高价值类别与低效类别之间的差异特征。',
            })

    if time_cols:
        tc = time_cols[0]
        items.append({
            'title': '时间维度覆盖',
            'content': f'数据包含时间列"{tc.get("inferred_label", tc["column_name"])}"，'
                       f'支持按时间维度进行趋势分析。通过对时间序列数据的分解，'
                       f'可识别周期性波动、趋势变化及异常时间节点，为预测与规划提供依据。',
        })

    items.append({
        'title': '数据质量评估',
        'content': f'数据质量评分 {score}/100，'
                   f'{"数据完整可靠，" if score >= 90 else "数据质量良好，建议关注缺失值" if score >= 80 else "数据需重点关注质量控制"}'
                   f'缺失率 {q.get("null_rate", 0)*100:.1f}%。'
                   f'本报告中的分析与图表均基于现有数据进行自动化生成，确保数据准确性。',
    })

    high_null = q.get('high_null_columns', [])
    if high_null:
        items.append({
            'title': '数据完整性说明',
            'content': f'以下列缺失值比例较高：{", ".join(high_null[:5])}。'
                       f'在分析涉及这些列时已进行空值排除处理，'
                       f'建议后续数据录入环节关注这些字段的完整填写，以提升分析精度。',
        })

    total_rows = data_profile.get('total_rows', 0)
    if total_rows:
        items.append({
            'title': '数据规模概述',
            'content': f'本期报告基于 {total_rows} 条数据记录进行分析，'
                       f'样本量{"充足，统计结果具有较好的代表性" if total_rows >= 100 else "适中，统计结果可作为参考" if total_rows >= 30 else "有限，分析结果仅供参考"}。',
        })

    return items


if __name__ == '__main__':
    import sys
    if len(sys.argv) > 1:
        from data_loader import load_workbook_metadata, load_daily
        fp = sys.argv[1]
        meta = load_workbook_metadata(fp)
        d0 = meta['date_range'][0]
        df = load_daily(fp, d0)
        m = calc_daily_metrics(df)
        print(f"Daily metrics for {d0.date()}:")
        print(f"  Tracking orders: {m['tracking_orders']}")
        print(f"  Total qty: {m['total_qty']}")
        print(f"  Avg order size: {m['avg_order_size']}")
        print(f"  Status dist: {m['status_dist']}")