"""共享测试夹具 — 所有测试文件共用同一文件级 SQLite 数据库。

使用同步引擎在模块级创建表，避免跨事件循环导致表缺失。
"""

import json
import os
from unittest.mock import AsyncMock, patch

import pytest
from fastapi.testclient import TestClient
from sqlalchemy import create_engine
from sqlalchemy.ext.asyncio import AsyncSession, async_sessionmaker, create_async_engine
from sqlalchemy.pool import StaticPool

from app.core.database import Base, get_db
from app.main import app

TEST_DB_PATH = "./test.db"
TEST_DATABASE_URL = f"sqlite+aiosqlite:///{TEST_DB_PATH}"
SYNC_DATABASE_URL = f"sqlite:///{TEST_DB_PATH}"

# --- 同步引擎：模块级创建表 ---
_sync_engine = create_engine(SYNC_DATABASE_URL, echo=False)


def _create_tables() -> None:
    """用同步引擎创建全部表。"""
    Base.metadata.create_all(_sync_engine)


def _drop_tables() -> None:
    """用同步引擎销毁全部表。"""
    Base.metadata.drop_all(_sync_engine)


# 模块导入时立即创建表
if os.path.exists(TEST_DB_PATH):
    os.remove(TEST_DB_PATH)
_create_tables()

# --- 异步引擎：测试运行时使用 ---
# StaticPool 保持单连接复用，connect_args 允许跨线程访问
test_engine = create_async_engine(
    TEST_DATABASE_URL,
    echo=False,
    poolclass=StaticPool,
    connect_args={"check_same_thread": False},
)
test_session_factory = async_sessionmaker(test_engine, class_=AsyncSession, expire_on_commit=False)


async def override_get_db():
    """覆盖数据库依赖，使用测试 SQLite。"""
    async with test_session_factory() as session:
        try:
            yield session
            await session.commit()
        except Exception:
            await session.rollback()
            raise


app.dependency_overrides[get_db] = override_get_db


@pytest.fixture(scope="session", autouse=True)
async def setup_db():
    """确保表已创建，测试结束后清理。"""
    # 表已在模块级同步创建，此处仅做清理
    yield
    await test_engine.dispose()
    _drop_tables()
    _sync_engine.dispose()
    if os.path.exists(TEST_DB_PATH):
        os.remove(TEST_DB_PATH)


@pytest.fixture(autouse=True)
async def mock_llm():
    """Mock LLM 客户端 — 所有 AI 相关测试不依赖真实 API。

    chat() 尝试解析 JSON 返回 Python 对象（dict/list），
    使 service 层 isinstance(result, dict/list) 检查通过。
    无法解析时返回原始字符串。
    """
    mock_dict = {"summary": "AI 分析结果（测试 mock）", "score": 75, "confidence": 0.8}
    mock_list = [{"type": "tech", "title": "测试协同", "match_reason": "mock"}]

    async def mock_chat_stream(self, messages, temperature=0.3, max_tokens=2000):
        yield json.dumps(mock_dict)

    async def mock_chat(self, messages, temperature=0.3, max_tokens=2000):
        # service 层传的可能是 str 或 list[dict]
        # 根据 prompt 内容判断期望返回类型
        if isinstance(messages, str):
            prompt_text = messages
        elif isinstance(messages, list):
            prompt_text = " ".join(str(m.get("content", "")) for m in messages)
        else:
            prompt_text = str(messages)

        # 如果 prompt 中包含 "JSON 数组" 或 "数组格式"，返回 list
        if "数组" in prompt_text or "JSON 数组" in prompt_text:
            return mock_list
        # 默认返回 dict
        return mock_dict

    with patch("app.services.llm_client.LLMClient.chat", mock_chat), \
         patch("app.services.llm_client.LLMClient.chat_stream", mock_chat_stream):
        yield


@pytest.fixture
def client():
    """创建测试客户端。"""
    return TestClient(app)


@pytest.fixture
def auth_headers(client: TestClient):
    """注册并登录投资人用户，返回认证头。"""
    client.post(
        "/api/v1/auth/register",
        json={
            "email": "smoke_test@example.com",
            "password": "password123",
            "name": "冒烟测试用户",
            "tenant_name": "冒烟测试机构",
            "role": "investor",
        },
    )
    resp = client.post(
        "/api/v1/auth/login",
        json={"email": "smoke_test@example.com", "password": "password123"},
    )
    token = resp.json()["data"]["access_token"]
    return {"Authorization": f"Bearer {token}"}


@pytest.fixture
def company_id(client: TestClient, auth_headers: dict):
    """创建测试企业，返回企业 ID。"""
    resp = client.post(
        "/api/v1/companies",
        json={"name": "冒烟测试公司", "industry": "AI"},
        headers=auth_headers,
    )
    return resp.json()["data"]["id"]