"""风险自动检测引擎。

基于月报结构化数据，检测指标越界并自动生成风险事件。
"""

import logging
from typing import Any

logger = logging.getLogger(__name__)

# 风险检测规则
RULES = [
    {
        "name": "现金跑道不足",
        "type": "financial",
        "severity": "critical",
        "condition": lambda d: _get_runway(d) < 3 and _get_runway(d) > 0,
        "title": "现金跑道不足 3 个月",
        "description": "当前现金跑道仅 {runway} 个月，需紧急融资",
        "suggested_action": "立即启动融资对话，评估 bridge loan 可能性",
    },
    {
        "name": "现金跑道预警",
        "type": "financial",
        "severity": "high",
        "condition": lambda d: 3 <= _get_runway(d) < 6,
        "title": "现金跑道低于 6 个月",
        "description": "当前现金跑道 {runway} 个月，需加快融资进度",
        "suggested_action": "与创始人沟通融资时间表，准备备选方案",
    },
    {
        "name": "烧钱率上升",
        "type": "financial",
        "severity": "medium",
        "condition": lambda d: _get_burn_trend(d) == "up",
        "title": "烧钱率持续上升",
        "description": "月度烧钱率呈上升趋势，需关注成本控制",
        "suggested_action": "审查主要支出项，制定成本优化计划",
    },
    {
        "name": "营收下滑",
        "type": "financial",
        "severity": "high",
        "condition": lambda d: _get_yoy(revenue=d) < 0,
        "title": "营收同比下滑",
        "description": "营收同比下降 {yoy}%，需关注业务增长",
        "suggested_action": "分析营收下滑原因，调整商业策略",
    },
    {
        "name": "高人员流失",
        "type": "org",
        "severity": "medium",
        "condition": lambda d: _get_departures(d) > 5,
        "title": "人员流失率较高",
        "description": "本月离职 {departures} 人，需关注团队稳定性",
        "suggested_action": "了解离职原因，评估核心岗位风险",
    },
    {
        "name": "团队净缩减",
        "type": "org",
        "severity": "high",
        "condition": lambda d: _get_net_headcount(d) < -3,
        "title": "团队规模显著缩减",
        "description": "本月团队净减少 {net} 人，需关注组织健康",
        "suggested_action": "与创始人沟通团队规划，评估关键岗位覆盖",
    },
]


def _get_runway(data: dict[str, Any]) -> float:
    """获取现金跑道月数。"""
    cash = data.get("cash_balance", {})
    try:
        return float(cash.get("runway_months", 0) or 0)
    except (ValueError, TypeError):
        return 0.0


def _get_burn_trend(data: dict[str, Any]) -> str:
    """获取烧钱率趋势。"""
    burn = data.get("burn_rate", {})
    return burn.get("trend", "")


def _get_yoy(revenue: dict[str, Any], d: dict[str, Any] = None) -> float:
    """获取营收同比增长率。"""
    if d is None:
        d = revenue
        revenue = d.get("revenue", {})
    yoy = revenue.get("yoy_change", "")
    try:
        return float(str(yoy).replace("%", "").replace("+", "") or 0)
    except (ValueError, TypeError):
        return 0.0


def _get_departures(data: dict[str, Any]) -> float:
    """获取离职人数。"""
    hc = data.get("headcount", {})
    try:
        return float(hc.get("departures", 0) or 0)
    except (ValueError, TypeError):
        return 0.0


def _get_net_headcount(data: dict[str, Any]) -> float:
    """获取团队净变化。"""
    hc = data.get("headcount", {})
    try:
        new = float(hc.get("new_hires", 0) or 0)
        dep = float(hc.get("departures", 0) or 0)
        return new - dep
    except (ValueError, TypeError):
        return 0.0


def detect_risks(
    structured_data: dict[str, Any],
    company_id: str,
) -> list[dict[str, Any]]:
    """从月报结构化数据中检测风险事件。

    Args:
        structured_data: 月报 AI 解析后的结构化数据
        company_id: 企业 ID

    Returns:
        风险事件列表，每项包含 type/severity/title/description/suggested_action
    """
    if not structured_data:
        return []

    risks: list[dict[str, Any]] = []
    runway = _get_runway(structured_data)
    yoy = _get_yoy(structured_data)
    departures = _get_departures(structured_data)
    net_hc = _get_net_headcount(structured_data)

    for rule in RULES:
        try:
            if rule["condition"](structured_data):
                risk = {
                    "company_id": company_id,
                    "type": rule["type"],
                    "severity": rule["severity"],
                    "title": rule["title"],
                    "description": rule["description"].format(
                        runway=runway,
                        yoy=abs(yoy),
                        departures=departures,
                        net=abs(net_hc),
                    ),
                    "suggested_action": rule["suggested_action"],
                }
                risks.append(risk)
                logger.info("检测到风险: %s — %s", rule["name"], risk["title"])
        except Exception as e:
            logger.warning("风险检测规则 '%s' 执行异常: %s", rule["name"], e)

    return risks