pytest+yaml实现接口自动化框架

前言

httprunner 用 yaml 文件实现接口自动化框架很好用，最近在看 pytest 框架，于是参考 httprunner的用例格式，写了一个差不多的 pytest 版的简易框架

项目结构设计

项目结构完全符合 pytest 的项目结构，pytest 是查找 test_.py 文件，我这里是查找 test_.yml 文件，唯一不同的就是这个地方
以前写test_*.py 的测试用例，现在完全不用写了，全部写yaml 文件就行，项目结构参考

只需在 conftest.py 即可实现，代码量超级少
pytest 7.x最新版

def pytest_collect_file(parent, file_path):
    # 获取文件.yml 文件,匹配规则
    if file_path.suffix == ".yml" and file_path.name.startswith("test"):
        return YamlFile.from_parent(parent, path=file_path)

pytest 5.x以上版本

import pytest
import requests


def pytest_collect_file(parent, path):
    # 获取文件.yml 文件,匹配规则
    if path.ext == ".yml" and path.basename.startswith("test"):
        # print(path)
        # print(parent)
        # return YamlFile(path, parent)
        return YamlFile.from_parent(parent, fspath=path)


class YamlFile(pytest.File):
    # 读取文件内容
    def collect(self):
        import yaml
        raw = yaml.safe_load(self.fspath.open(encoding='utf-8'))
        for yaml_case in raw:
            name = yaml_case["test"]["name"]
            values = yaml_case["test"]
            yield YamlTest.from_parent(self, name=name, values=values)

class YamlTest(pytest.Item):
    def __init__(self, name, parent, values):
        super(YamlTest, self).__init__(name, parent)
        self.name = name
        self.values = values
        self.request = self.values.get("request")
        self.validate = self.values.get("validate")
        self.s = requests.session()

    def runtest(self):
        # 运行用例
        request_data = self.values["request"]
        # print(request_data)
        response = self.s.request(**request_data)
        print("\n", response.text)
        # 断言
        self.assert_response(response, self.validate)

    def assert_response(self, response, validate):
        '''设置断言'''
        import jsonpath
        for i in validate:
            if "eq" in i.keys():
                yaml_result = i.get("eq")[0]
                actual_result = jsonpath.jsonpath(response.json(), yaml_result)
                expect_result = i.get("eq")[1]
                print("实际结果：%s" % actual_result)
                print("期望结果：%s" % expect_result)
                assert actual_result[0] == expect_result

pytest 4.x 以下版本

import pytest
import requests
# 作者-上海悠悠 QQ交流群:717225969
# blog地址 https://www.cnblogs.com/yoyoketang/


def pytest_collect_file(parent, path):
    # 获取文件.yml 文件,匹配规则
    if path.ext == ".yml" and path.basename.startswith("test"):
        # print(path)
        # print(parent)
        return YamlFile(path, parent)



class YamlFile(pytest.File):
    # 读取文件内容
    def collect(self):
        import yaml
        raw = yaml.safe_load(self.fspath.open(encoding='utf-8'))
        for yaml_case in raw:
            name = yaml_case["test"]["name"]
            values = yaml_case["test"]
            yield YamlTest(name, self, values)


class YamlTest(pytest.Item):
    def __init__(self, name, parent, values):
        super(YamlTest, self).__init__(name, parent)
        self.name = name
        self.values = values
        self.request = self.values.get("request")
        self.validate = self.values.get("validate")
        self.s = requests.session()

    def runtest(self):
        # 运行用例
        request_data = self.values["request"]
        # print(request_data)
        response = self.s.request(**request_data)
        print("\n", response.text)
        # 断言
        self.assert_response(response, self.validate)

    def assert_response(self, response, validate):
        '''设置断言'''
        import jsonpath
        for i in validate:
            if "eq" in i.keys():
                yaml_result = i.get("eq")[0]
                actual_result = jsonpath.jsonpath(response.json(), yaml_result)
                expect_result = i.get("eq")[1]
                print("实际结果：%s" % actual_result)
                print("期望结果：%s" % expect_result)
                assert actual_result[0] == expect_result

断言这部分，目前只写了判断相等，仅供参考，支持jsonpath来提取json数据

yaml格式的用例

在项目的任意目录，只要是符合test_开头的yml文件，我们就认为是测试用例
test_login.yml的内容如下

- test:
    name: login case1
    request:
        url: http://49.235.x.x:7000/api/v1/login/
        method: POST
        headers:
            Content-Type: application/json
            User-Agent: python-requests/2.18.4
        json:
            username: test
            password: 123456
    validate:
        - eq: [$.msg, login success!]
        - eq: [$.code, 0]


- test:
    name: login case2
    request:
        url: 49.235.x.x:7000/api/v1/login/
        method: POST
        headers:
            Content-Type: application/json
            User-Agent: python-requests/2.18.4
        json:
            username: test
            password: 123456
    validate:
        - eq: [$.msg, login success!]
        - eq: [$.code, 0]

运行用例

运行用例，完全符合pytest的只需用例风格，支持allure报告

pytest -v

D:\soft\api_pytest_1208>pytest -v
====================== test session starts ======================
platform win32 -- Python 3.6.6, pytest-4.5.0, py-1.9.0,
cachedir: .pytest_cache
rootdir: D:\soft\api_pytest_1208
plugins: allure-pytest-2.8.6
collected 4 items

data/test_login.yml::login case1 PASSED                    [ 25%]
data/test_login.yml::login case2 PASSED                    [ 50%]
data/test_login1.yml::login case1 PASSED                   [ 75%]
data/test_login1.yml::login case2 PASSED                   [100%]

=================== 4 passed in 1.34 seconds ====================

allure报告

pytest --alluredir ./report

目前是把 yaml 文件下每个 test 当一个用例执行，后续还可以加上提取参数，参数关联更高级的功能！

这可能是B站最详细的pytest自动化测试框架教程，整整100小时，全程实战！！！