如何使用pytest对多个Fixture进行参数化？

核心解决办法是利用pytest.mark.parametrize的indirect参数，指定需要解析为Fixture的参数名称，让pytest自动把参数中的Fixture名字符串转换成对应Fixture的实例。

针对你的场景修改后的代码 #

@pytest.mark.parametrize(
    "input_payload, expected_output",
    [
        ("input_payload_1", {"value"}),
        ("input_payload_2", {"value2"}),
    ],
    indirect=["input_payload"]  # 关键：告诉pytest这个参数要解析为Fixture
)
def test_something(input_payload, expected_output):
    result = tested_functionality(input_payload)
    assert result == expected_output

原理说明 #

默认情况下，parametrize会直接把参数列表里的字符串传给测试函数，但加上indirect=["input_payload"]后，pytest会把每个参数组里的"input_payload_1"、"input_payload_2"当作Fixture名称，自动执行对应的input_payload_1()和input_payload_2()Fixture函数，把返回的Pydantic实例传给测试函数的input_payload参数，而不是传字符串本身。

扩展：同时参数化多个Fixture #

如果需要同时对多个Fixture进行这类参数化，只需要把多个参数名加入indirect的列表即可。比如有两个Fixture参数的场景：

# 假设conftest.py里还有这两个Fixture
@pytest.fixture
def config_a():
    pass

@pytest.fixture
def config_b():
    pass

# 测试函数参数化多个Fixture
@pytest.mark.parametrize(
    "payload, config, expected",
    [
        ("input_payload_1", "config_a", {"val1"}),
        ("input_payload_2", "config_b", {"val2"}),
    ],
    indirect=["payload", "config"]  # 多个参数都解析为Fixture
)
def test_multi_fixture(payload, config, expected):
    result = tested_func(payload, config)
    assert result == expected

内容的提问来源于stack exchange，提问作者J. Donič