group_fqcd_jr/app/service/risk_natural_language.py

"""风控预警自然语言筛选条件解析。"""

from __future__ import annotations

import re
from datetime import UTC, date, datetime, time, timedelta
from zoneinfo import ZoneInfo

from app.core.risk_contracts import RiskAlertQuery

SHANGHAI = ZoneInfo("Asia/Shanghai")


def parse_risk_alert_filters(
    message: str,
    *,
    now: datetime | None = None,
) -> dict[str, str]:
    """把常见中文筛选表达转换为风控查询工具参数。"""
    current = now or datetime.now(SHANGHAI)
    if current.tzinfo is None:
        current = current.replace(tzinfo=SHANGHAI)

    filters: dict[str, str] = {}
    level = re.search(r"(低|中|高)(?:风险)?", message)
    if level:
        filters["risk_level"] = level.group(1)

    rule = re.search(r"RW-\d{3}", message, re.IGNORECASE)
    if rule:
        filters["rule_code"] = rule.group(0).upper()

    customer = re.search(
        r"(?:客户编号|客户号|客户账户|客户)\s*[:：#]?\s*"
        r"([A-Za-z0-9_-]{1,64})",
        message,
    )
    if customer:
        filters["customer_no"] = customer.group(1)
    elif matched := re.search(r"\b(CUST-\d+)\b", message, re.IGNORECASE):
        filters["customer_no"] = matched.group(1).upper()

    product_code = re.search(
        r"(?:产品编号|产品代码|产品编码)\s*[:：#]?\s*"
        r"([A-Za-z0-9_-]{1,64})",
        message,
    )
    if product_code:
        filters["product_code"] = product_code.group(1)
    elif matched := re.search(r"\b(P-[A-Za-z0-9_-]{1,60})\b", message):
        filters["product_code"] = matched.group(1)

    product_name = re.search(
        r"产品名称\s*[:：#]?\s*([^\s，,。；;？！]{1,128})",
        message,
    )
    if product_name is None:
        product_name = re.search(
            r"产品\s*[:：]\s*([^\s，,。；;？！]{1,128})",
            message,
        )
    if product_name:
        candidate = product_name.group(1).strip()
        if candidate not in {"编号", "代码", "编码", "名称"}:
            filters["product_name"] = candidate

    start_time, end_time = _parse_time_range(message, current)
    if start_time is not None:
        filters["start_time"] = _to_utc_iso(start_time)
    if end_time is not None:
        filters["end_time"] = _to_utc_iso(end_time)

    validated = RiskAlertQuery.model_validate(filters)
    return validated.model_dump(mode="json", exclude_none=True)


def _parse_time_range(
    message: str,
    current: datetime,
) -> tuple[datetime | None, datetime | None]:
    explicit = re.search(
        r"(\d{4}-\d{2}-\d{2})\s*(?:至|到|~|—)\s*(\d{4}-\d{2}-\d{2})",
        message,
    )
    if explicit:
        start_date = _parse_date(explicit.group(1))
        end_date = _parse_date(explicit.group(2))
        if start_date is not None and end_date is not None:
            return (
                datetime.combine(start_date, time.min, tzinfo=SHANGHAI),
                datetime.combine(end_date, time.max, tzinfo=SHANGHAI),
            )

    if "今天" in message or "今日" in message:
        start = current.replace(hour=0, minute=0, second=0, microsecond=0)
        return start, current

    if "本月" in message:
        start = current.replace(
            day=1,
            hour=0,
            minute=0,
            second=0,
            microsecond=0,
        )
        return start, current

    recent = re.search(r"(?:近|最近)\s*(\d+)\s*(天|日)", message)
    if recent:
        days = int(recent.group(1))
        return current - timedelta(days=days), current

    since = re.search(r"(\d{4}-\d{2}-\d{2})\s*(?:以来|之后|起)", message)
    if since:
        start_date = _parse_date(since.group(1))
        if start_date is not None:
            return datetime.combine(start_date, time.min, tzinfo=SHANGHAI), current

    until = re.search(r"(?:截至|截止)\s*(\d{4}-\d{2}-\d{2})", message)
    if until:
        end_date = _parse_date(until.group(1))
        if end_date is not None:
            return None, datetime.combine(end_date, time.max, tzinfo=SHANGHAI)

    return None, None


def _to_utc_iso(value: datetime) -> str:
    return value.astimezone(UTC).replace(tzinfo=None).isoformat()


def _parse_date(value: str) -> date | None:
    try:
        return date.fromisoformat(value)
    except ValueError:
        return None