在测试期间模拟环境变量

Question

我有一个非常简单的 fastapi 应用程序，我想测试一下，dummy_api.py 的代码如下：

import os
from fastapi import FastAPI


app = FastAPI()


@app.get(os.getenv("ENDPOINT", "/get"))
def func():
    return {
        "message": "Endpoint working !!!"
    }

当我想测试这个时，我正在使用以下文件：

from fastapi.testclient import TestClient
import dummy_api


def test_dummy_api():
    client = TestClient(dummy_api.app)
    response = client.get("/get")
    assert response.status_code == 200


def test_dummy_api_with_envar(monkeypatch):
    monkeypatch.setenv("ENDPOINT", "dummy")
    client = TestClient(dummy_api.app)
    response = client.get("/dummy")
    assert response.status_code == 200

但是我无法模拟环境变量部分，因为其中一项测试失败并显示 404。

pytest -s -v
================================================================= test session starts ==================================================================
platform linux -- Python 3.8.5, pytest-6.2.2, py-1.9.0, pluggy-0.13.1 -- /home/subhayan/anaconda3/envs/fastapi/bin/python
cachedir: .pytest_cache
rootdir: /home/subhayan/Codes/ai4bd/roughdir
collected 2 items                                                                                                                                      

test_dummy_api.py::test_dummy_api PASSED
test_dummy_api.py::test_dummy_api_with_envar FAILED

======================================================================= FAILURES =======================================================================
______________________________________________________________ test_dummy_api_with_envar _______________________________________________________________

monkeypatch = <_pytest.monkeypatch.MonkeyPatch object at 0x7ff8c4cf1430>

    def test_dummy_api_with_envar(monkeypatch):
        monkeypatch.setenv("ENDPOINT", "dummy")
        client = TestClient(dummy_api.app)
        response = client.get("/dummy")
>       assert response.status_code == 200
E       assert 404 == 200
E         +404
E         -200

test_dummy_api.py:15: AssertionError
=============================================================== short test summary info ================================================================
FAILED test_dummy_api.py::test_dummy_api_with_envar - assert 404 == 200
============================================================= 1 failed, 1 passed in 0.19s ==============================================================

谁能指出我哪里错了！！

Answer 1

在我看来是这样的：

当测试文件被 pytest 加载时，所有可以在 dummy_api 中执行的东西都会被执行，因为它是导入的。
这意味着装饰函数（即@app.get(...)）的装饰器将被执行。此时猴子修补还没有开始。
测试功能启动后，环境变量设置太晚了。

Answer 2

您可以使用 parametrized fixtures and the importlib.reload function 来测试是否确实使用了环境变量。

我的测试目录如下所示：

.
└── tests
    ├── conftest.py
    ├── dummy_api.py
    └── test_api.py

这是我的 conftest.py:

import pytest
from fastapi.testclient import TestClient
from importlib import reload
import dummy_api


@pytest.fixture(params=["/get", "/dummy", "/other"])
def endpoint(request, monkeypatch):
    monkeypatch.setenv("ENDPOINT", request.param)
    return request.param


@pytest.fixture()
def client(endpoint):
    app = reload(dummy_api).app
    yield TestClient(app=app)

这里是 test_api.py 文件：

import os


def test_dummy_api(client):
    endpoint = os.environ["ENDPOINT"]
    response = client.get(endpoint)
    assert response.status_code == 200
    assert response.json() == {"message": f"Endpoint {endpoint} working !"}

运行pytest后的测试输出：

collected 3 items                                                                                                                                    

tests/test_api.py::test_dummy_api[/get] PASSED                                                                                                   [ 33%]
tests/test_api.py::test_dummy_api[/dummy] PASSED                                                                                                  [ 66%]
tests/test_api.py::test_dummy_api[/other] PASSED                                                                                                  [100%]

在测试期间模拟环境变量

mocking environment variables during testing

python

mocking

pytest

fastapi

pytest-mock