Question

时，您将得到一个小的json响应。

我将响应保存在一个变量中（实际）。我也将响应放入另一个变量（预期）中。两个响应都相同。我正在更改值以测试失败的案例。最终目标是比较2并确保它们匹配。

我有2个函数，其中1个比较两个字典的键和值，另一个函数对字典进行排序。下面的代码：

import json
import requests


response = requests.get('https://reqres.in/api/users/2')
#actual_response saves the json as we get it from url above
actual_response= json.loads(response.text)

#expected response is saved after using pretty json that will be used to testing/comparing actual vs expected
expected_response={
    "data": {
        "id": 2,
        "first_name": "Janet",
        "last_name": "Weaver",
        "avatar": "https://s3.amazonaws.com/uifaces/faces/twitter/josephstein/128.jpg"
    }
}

# sort the key values before comparing
def dict_sort(dictA,dictB):
    dictA, dictB = json.dumps(dictA, sort_keys=True), json.dumps(dictB, sort_keys=True)
    dictA == dictB

#if there are any failure due to mismatch in key value the function below will show that
def key_diff(dictA,dictB):
    for key,value in dictA.items():
        for keyB,valueB in dictB.items():
            for k,v in value.items():
                for k2,v2 in valueB.items():
                    if(key!= keyB):
                        print('Expected',key,' but got',keyB)
                    if(k!=k2):
                        print('Expected', k, ' but got', k2)
                    if(v!=v2):
                        print('Expected', v, ' but got', v2)
                    else:
                        print()

dict_sort(actual_response,expected_response)

if(actual_response==expected_response):
    print('Passed')
else:
    print('Failed')
    key_diff(actual_response,expected_response)

问题：没有差异时测试就会通过，但是如果有差异，顺序就会变得疯狂。这是一个示例，其中我将数据更改为预期响应内的数据：
预期数据，但数据达标

期望的ID，但姓氏

预计2但得到了Weaver

排序函数应该比使用sort_keys = True更具体吗？通过考虑** args的方式，但我认为在这种情况下这不是一个好选择。

谢谢您的专家评论和时间。

Answer 1

在3.7以下的Python版本中不能保证键顺序；当您需要创建一个记住按键顺序的对象时，应该使用collections.OrderedDict。

在Python 3.7中，插入顺序被保留，因此您的键将始终匹配。

Answer 2

我建议使用unittest并避免使用太多嵌套的for循环

from unittest import TestCase

import pandas as pd
import requests


def mocked_server_response():
    expected = {"data": {"id": 2, "first_name": "Janet", "last_name": "Weaver",
                         "avatar": "https://s3.amazonaws.com/uifaces/faces/twitter/josephstein/128.jpg"}}
    data = expected['data']
    df = pd.DataFrame(my_dict['data'], index=[0])
    return [expected, data, df]

在这一点上，mocked_server_response()将为您提供：

Out[27]:
   id first_name last_name                                             avatar
0   2      Janet    Weaver  https://s3.amazonaws.com/uifaces/faces/twitter...

现在，您可以轻松地在课程中进行测试。


class TestServerResponse(TestCase):
    real_response = requests.get('https://reqres.in/api/users/2')

    def setUp(self):
        self.actual_response = real_response

    def response(self):
        self.assertEqual(self.actual_response, mocked_server_response()[0])

    def test_data_in_response(self):
        self.assertEqual(self.actual_response['data'], mocked_server_response()[1])

    def test_dataframe(self):
        self.assertEqual(pd.DataFrame(self.actual_response['data'], index=[0]), mocked_server_response()[2])

Answer 3

import requests
import json

# Here I am converting the expected payload in dictionary
expected_payload = json.loads("""[
    {
        "key": "data-center",
        "value": "All",
        "values": [
            "1",
            "2"
        ]
    },
    {
        "key": "router",
        "value": "All",
        "values": [
            "cisco",
            "juniper"
        ]
    },
    {
        "key": "virtual-machine",
        "value": "All",
        "values": [
            "dell",
            "hp",
            "None"
        ]
    },

]""")


def test_get_all_system_attributes():
    url = "http://" + str(ipaddr) + ":" + str(port) + "/Service/" + "system_attribute/"
    payload = {}
    headers = {
        'Content-Type': 'application/json'
    }
    actual_response = requests.request("GET", url, headers=headers, data=payload)
    assert json.loads(actual_response.text) == expected_payload

# json.loads(actual_response.text) will convert the response in dictionary
# using assert I am comparing the actual_response with exepcted_response

if __name__ == '__main__':
    test_get_all_system_attributes()

比较Python中的get请求的预期与实际json响应

3 个答案: