跳到主要内容

测试

问题​

Python 项目如何做测试？pytest 的核心功能有哪些？

答案​

pytest 基础​

tests/test_user.py
import pytest
from myapp.user import create_user, UserExistsError

# 基础测试
def test_create_user():
    user = create_user("Alice", "alice@example.com")
    assert user.name == "Alice"
    assert user.email == "alice@example.com"

# 异常测试
def test_duplicate_user():
    create_user("Bob", "bob@example.com")
    with pytest.raises(UserExistsError, match="already exists"):
        create_user("Bob", "bob@example.com")

# 参数化
@pytest.mark.parametrize("email,valid", [
    ("user@example.com", True),
    ("invalid", False),
    ("@example.com", False),
    ("user@.com", False),
])
def test_validate_email(email: str, valid: bool):
    assert validate_email(email) == valid

Fixture​

import pytest
from sqlalchemy.ext.asyncio import AsyncSession

@pytest.fixture
async def db_session() -> AsyncGenerator[AsyncSession, None]:
    """每个测试使用独立事务，测试后回滚"""
    async with async_session() as session:
        async with session.begin():
            yield session
            await session.rollback()

@pytest.fixture
def user(db_session: AsyncSession) -> User:
    """依赖 db_session fixture"""
    user = User(name="Test", email="test@example.com")
    db_session.add(user)
    return user

async def test_get_user(db_session: AsyncSession, user: User):
    result = await get_user(db_session, user.id)
    assert result.name == "Test"

Mock​

from unittest.mock import AsyncMock, patch, MagicMock

# Mock 外部 API
@patch("myapp.service.httpx.AsyncClient.get")
async def test_fetch_weather(mock_get: AsyncMock):
    mock_get.return_value = MagicMock(
        status_code=200,
        json=lambda: {"temp": 25, "city": "Beijing"},
    )

    result = await fetch_weather("Beijing")
    assert result["temp"] == 25
    mock_get.assert_called_once()

# Mock 数据库
@patch("myapp.repo.get_user_by_email")
async def test_login(mock_get_user: AsyncMock):
    mock_get_user.return_value = User(name="Alice", password_hash="...")
    result = await login("alice@example.com", "password")
    assert result.token is not None

覆盖率​

# 运行测试 + 覆盖率
pytest --cov=myapp --cov-report=html tests/

# pyproject.toml 配置
[tool.pytest.ini_options]
testpaths = ["tests"]
asyncio_mode = "auto"

[tool.coverage.run]
source = ["myapp"]
omit = ["*/tests/*", "*/migrations/*"]

常见面试问题​

Q1: pytest 和 unittest 的区别？​

答案：

特性pytestunittest
断言原生 assertself.assertEqual()
Fixture函数式，可组合setUp/tearDown
参数化@pytest.mark.parametrizesubTest
插件极丰富（1000+）少
异步pytest-asyncioIsolatedAsyncioTestCase

Q2: 测试金字塔在 Python 项目中怎么落地？​

答案：

         /  E2E  \        ←  少量：Playwright / Selenium
        / 集成测试 \       ←  中量：测试 API + DB
       /  单元测试  \      ←  大量：纯逻辑测试，Mock 依赖
  • 单元测试（70%）：测试纯函数、业务逻辑
  • 集成测试（20%）：测试 API 端到端（使用测试数据库）
  • E2E 测试（10%）：核心流程的端到端

Q3: 如何 Mock 异步函数？​

答案：

from unittest.mock import AsyncMock

mock_fn = AsyncMock(return_value={"data": "test"})
result = await mock_fn()
assert result == {"data": "test"}

相关链接​