import json
import random
from datetime import date, timedelta
from pathlib import Path
from typing import Dict, List, Optional


class CreditPayloadService:
    """根据征信字段 schema 生成稳定随机的 mock payload。"""

    def __init__(self, schema_path: str):
        self.schema_path = schema_path
        self.schema = self._load_schema()

    def generate_payload(
        self,
        model: str,
        h_type: str,
        filename: str,
        subject_identity: Optional[Dict[str, str]] = None,
    ) -> dict:
        identity = subject_identity or {}
        rng = random.Random(self._build_seed(model, h_type, filename, identity))
        payload = {
            "lx_header": {},
            "lx_debt": {},
            "lx_publictype": {},
        }

        for item in self.schema:
            domain = item["domain"]
            field = item["field"]
            field_type = item["type"]
            payload[domain][field] = self._generate_value(field, field_type, item, rng, identity)

        return payload

    def _load_schema(self) -> List[dict]:
        schema_file = Path(self.schema_path)
        if not schema_file.is_absolute():
            schema_file = Path(__file__).resolve().parent.parent / schema_file
        return json.loads(schema_file.read_text(encoding="utf-8"))

    @staticmethod
    def _build_seed(
        model: str,
        h_type: str,
        filename: str,
        subject_identity: Dict[str, str],
    ) -> str:
        return (
            f"{model}|{h_type}|{filename}|"
            f"{subject_identity.get('staff_name', '')}|"
            f"{subject_identity.get('staff_id_card', '')}"
        )

    def _generate_value(
        self,
        field: str,
        field_type: str,
        item: dict,
        rng: random.Random,
        subject_identity: Dict[str, str],
    ) -> str:
        if field_type == "string":
            return self._generate_string(field, rng, subject_identity)
        if field_type == "amount":
            return f"{rng.uniform(0, 500000):.2f}"
        if field_type == "count":
            return str(rng.randint(0, 20))
        if field_type == "status":
            return rng.choice(item["options"])
        raise ValueError(f"Unsupported field type: {field_type}")

    def _generate_string(
        self,
        field: str,
        rng: random.Random,
        subject_identity: Dict[str, str],
    ) -> str:
        if field == "query_cert_no" and subject_identity.get("staff_id_card"):
            return subject_identity["staff_id_card"]
        if field == "query_cust_name" and subject_identity.get("staff_name"):
            return subject_identity["staff_name"]
        if field == "query_cert_no":
            return self._generate_cert_no(rng)
        if field == "query_cust_name":
            return self._generate_name(rng)
        if field == "report_time":
            return self._generate_report_date(rng)
        return f"mock_{rng.randint(1000, 9999)}"

    @staticmethod
    def _generate_cert_no(rng: random.Random) -> str:
        area_code = "330781"
        start_date = date(1980, 1, 1)
        birthday = start_date + timedelta(days=rng.randint(0, 14000))
        sequence = f"{rng.randint(100, 999)}"
        check_code = rng.choice("0123456789X")
        return f"{area_code}{birthday.strftime('%Y%m%d')}{sequence}{check_code}"

    @staticmethod
    def _generate_name(rng: random.Random) -> str:
        surnames = ["张", "王", "李", "赵", "陈", "刘", "周", "吴"]
        given_names = ["伟", "芳", "娜", "敏", "静", "磊", "洋", "婷", "超", "洁"]
        return f"{rng.choice(surnames)}{rng.choice(given_names)}{rng.choice(given_names)}"

    @staticmethod
    def _generate_report_date(rng: random.Random) -> str:
        base_date = date(2024, 1, 1)
        report_date = base_date + timedelta(days=rng.randint(0, 365))
        return report_date.strftime("%Y-%m-%d")