如何遍历 @pytest.mark.parametrize 数据或将其打印在测试中

Question

我有这些数据基本上是我想传递给函数的参数：

data2 = { "user": "bender", "resource": "machine/2", "action": "delete"}
data3 = { "user": "bender", "resource": "machine/2", "action": "delete"}
data4 = { "user": "bender", "resource": "machines/*", "action": "*"}
data5 = { "user": "leela", "resource": "machine/1", "action": "delete"}
data6 = { "user": "leela", "resource": "machine/*", "action": "*"}

我使用@pytest.mark.parametrize进行了此测试：

@pytest.mark.parametrize("test_input",  [
    data2,
    data3,
    data4,
    data5,
    data6,
])
@patch("dev_maintenance.roles.db")
def test_check_if_user_has_permissions(db,test_input):
    tmp = {}
    tmp["user"] = "bender"
    tmp["resource"] = "machine/1"
    tmp["action"] = "delete"

    db.session.query.return_value.filter_by.return_value.join.return_value.all.return_value = [(tmp["user"], tmp["resource"], tmp["action"])]

    assert check_if_user_has_permissions(user = "bender",resource= "machine/1", access_token= "delete") == True

我想要实现的是在check_if_user_has_permissions(user, resource, action)上传递每个数据集check_if_user_has_permissions(user, resource, action)以便我可以断言是值，如果断言失败，因为断言是True == False我希望该测试用例通过，因为在我的情况如果发生这种情况，则意味着用户没有权限，并且测试应该通过。

我试图在像print(test_input)这样的测试上print(test_input) ，看看它是否会打印所有数据集，但只打印一个。

Answer 1

你好像在问两个不同的问题。 如果你想断言否定，只需使用assert not 。 如果您使用-v标志运行 pytest，它将输出它正在运行的所有测试，对于参数化测试，它还将显示它运行测试的每个参数值。

您可以像这样编写测试：

$ cat test.py
import pytest

data = [
    { "user": "bender", "resource": "machine/2", "action": "delete"},
    { "user": "bender", "resource": "machine/2", "action": "delete"},
    { "user": "bender", "resource": "machines/*", "action": "*"},
    { "user": "leela", "resource": "machine/1", "action": "delete"},
    { "user": "leela", "resource": "machine/*", "action": "*"}
]


def check_if_user_has_permissions(*args, **kwargs):
    # Dummy implementation for demonstration
    import random
    print(kwargs)
    return random.choice([True, False])


@pytest.mark.parametrize('test_input', data)
def test_check_if_user_has_permissions(test_input):
    # ... additional test setup ...
    assert not check_if_user_has_permissions(**test_input)

然后运行：

$ pytest -v test.py
============================= test session starts ==============================
platform cygwin -- Python 3.6.8, pytest-5.3.1, py-1.8.0, pluggy-0.13.1 -- /usr/bin/python3
cachedir: .pytest_cache
rootdir: .
collected 5 items

test.py::test_check_if_user_has_permissions[test_input0] PASSED          [ 20%]
test.py::test_check_if_user_has_permissions[test_input1] PASSED          [ 40%]
test.py::test_check_if_user_has_permissions[test_input2] PASSED          [ 60%]
test.py::test_check_if_user_has_permissions[test_input3] FAILED          [ 80%]
test.py::test_check_if_user_has_permissions[test_input4] FAILED          [100%]

=================================== FAILURES ===================================
_______________ test_check_if_user_has_permissions[test_input3] ________________

test_input = {'action': 'delete', 'resource': 'machine/1', 'user': 'leela'}

    @pytest.mark.parametrize('test_input', data)
    def test_check_if_user_has_permissions(test_input):
>       assert not check_if_user_has_permissions(**test_input)
E       AssertionError: assert not True
E        +  where True = check_if_user_has_permissions(**{'action': 'delete', 'resource': 'machine/1', 'user': 'leela'})

test.py:21: AssertionError
----------------------------- Captured stdout call -----------------------------
{'user': 'leela', 'resource': 'machine/1', 'action': 'delete'}
_______________ test_check_if_user_has_permissions[test_input4] ________________

test_input = {'action': '*', 'resource': 'machine/*', 'user': 'leela'}

    @pytest.mark.parametrize('test_input', data)
    def test_check_if_user_has_permissions(test_input):
>       assert not check_if_user_has_permissions(**test_input)
E       AssertionError: assert not True
E        +  where True = check_if_user_has_permissions(**{'action': '*', 'resource': 'machine/*', 'user': 'leela'})

test.py:21: AssertionError
----------------------------- Captured stdout call -----------------------------
{'user': 'leela', 'resource': 'machine/*', 'action': '*'}
========================= 2 failed, 3 passed in 0.19s ==========================

Answer 2

一段时间后，我找到了一个更简单的解决方案。

我做了一个mark.parametrize这样的：

data = 
[
["bender", "machine_4", "action1", False],
["leela", "machine_2", "action1", False],
["leela", "file/2", "action4", True],
["leela", "file/2", "action2", False],
["leela", "application/1", "action1", True],
["leela", "application/2", "action1", False],
["fry", "application/1", "action1", True],
["zoidberg", "application/1", "action1", False],
]


@pytest.mark.parametrize("user,resource,action, expected", data)
@patch("dev_maintenance.roles.db")
def test_check_if_user_has_permissions(db,user,resource,action,expected):

    db.session.query.return_value.filter_by.return_value.join.return_value.all.return_value = \
        [
            ("bender", "machine_1", "action1"),
            ("bender", "machine_2", "action1"),
            ("bender", "machine_3", "action1"),
            ("bender", "machine_3", "action2"),
            ("leela", "file/*", "action4"),
            ("leela", "application/1", "*"),
            ("fry", "*", "*"),
        ]

    assert check_if_user_has_permissions(user, resource, action) == expected

我在这里所做的是将user 、 resource和action参数传递给函数，并断言具有这些参数的函数的响应等于expected参数。 所以它会像 True == True 一样工作，具体取决于参数。