编写http接口api及接口自动化测试

2023-12-17

片言
此文中代码都是笔者工作中源码，所以不会很完整，主要摘常见场景的api片段用以举例说明

另:此文主要针对自动化测试人员，尤其有python基础阅读更佳。

笔者使用 python3.6 + postgresql10 + flask 0.12 的环境

场景介绍
我们要实现后台管理中的用户管理功能，实现对用户的增、修改、删除、查询操作。

文中场景省去登录和用户权限部分，但我会在接口代码片段中进行注释说明

接口：新增用户
（定义访问的http链接地址，以及请求使用的 method） http://127.0.0.1:7070/guardian/users

 
@app.route('/guardian/users', methods=['POST'])    # 使用 POST 方法
@require_role('guardian_admin')    # 登录用户的角色权限必须是guardian_admin 此处是自定义的更角色管理中的设置管理员角色一致即可。不需要角色权限也可以注释此行即可
@login_required    # 验证用户登录 flask_login插件
def create_user():
    '''
    下面json串是新增时候前台表单提交的内容以及格式，也就是我们接口测试时候需要添加的测试数据
    {
            "user": {
                "domain_id": "default",
                "enabled": True,
                "name": "tsbc1",
                "password": "111111",
                "realname": "Ray",
                "email": "tsbc@vip.qq.com",
                "tel": "+86-029-88880000",
                "phone": "13800138000",
                "org_id": "7d22d25a-b648-4ad3-a834-2398eb989a9e",
                "org_array":[],
                "comments": "备注信息"
            }
        }
    '''
    try:
        domain_id = None
        user_name = None
        user_passwd = None
        project_id = None
        real_name = None
        email = None
        phone = None
        org_id = None
        org_array = None
        comments = None
        try:
            # 接收前端请求数据，并使用json.loads 把数据转换成Python可以读格式
            resq_obj = request.data
            resq_json = None
            if type(resq_obj) is bytes:
                resq_json = resq_obj.decode()
            else:
                resq_json = resq_obj
            user_obj = jsonpickle.loads(resq_json)    # 此处使用了jsonpickle的lib，使用json lib也是一样的
            logging.debug('create user {0} '.format(user_obj))
            if 'default_project_id' in user_obj['user']:
                project_id = user_obj['user']['default_project_id']
 
            domain_id = user_obj['user']['domain_id']
            user_name = user_obj['user']['name']
            user_passwd = user_obj['user']['password']
 
        except Exception as ee:
            abort(400)
            #raise ee
 
        if domain_id is not None and user_name is not None and user_passwd is not None:
            # 笔者使用 keystone 进行用户管理 此处在keystone创建帐号，然后获取id,把id及其它信息会存放在自定义的数据库表中
            resp = usermgt.create_user(domain_id, user_name, user_passwd, project_id)    # resp返回值是 dict 格式
            if 'id' in resp['user']:
                try:
                    # 把前端POST的表单值进行变量赋值 传给后台进行存库
                    user_id = resp['user']['id']
                    if 'realname' in user_obj['user']:
                        real_name = user_obj['user']['realname']
                    if 'email' in user_obj['user']:
                        email = user_obj['user']['email']
                    if 'tel' in user_obj['user']:
                        tel = user_obj['user']['tel']
                    if 'phone' in user_obj['user']:
                        phone = user_obj['user']['phone']
                    if 'org_id' in user_obj['user']:
                        org_id = user_obj['user']['org_id']
                    if 'org_array' in user_obj['user']:
                        org_array = user_obj['user']['org_array']
                    if 'comments' in user_obj['user']:
                        comments = user_obj['user']['comments']
 
                except Exception as e:
                    # raise e
                    abort(400)
                if user_obj is not None:
                    try:
                        # 调用方法进行后台存库
                        GuardianModelDao._save_update_users_info(user_id, user_name, real_name, email, tel, phone, org_id, org_array, comments)
                        return wrap_response(SUCCESS_MSG, 200)    # 执行成功 返回操作成功信息及状态码 200
                    except Exception as e:
                        LOG.debug(e)
                        abort(500)    # 否则 后台500 错误
            else:
                abort(500)

后台函数

用于实现对接口中传递数据进行处理入库、修改、查询等等。

 
 @classmethod
    def _save_update_users_info(cls, user_id, username, real_name, email, tel, phone, org_id, org_array, comments):
        LOG.debug('start create users info...')
        try:
            with TxConn() as conn: # 连接数据库
                with conn.cursor() as cursor:
                    # save user data
                    now_time = time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())
                    # 接收变量赋值 user_data
                    users_data = {
                        "user_id":user_id,
                        "username":username,
                        "realname":real_name,
                        "email":email,
                        "tel":tel,
                        "phone":phone,
                        "org":org_id,
                        "org_array":org_array,
                        "comments":comments,
                        "now_time":now_time
                    }
                    cursor.execute(cls.insert_users_info, users_data)    # 插入数据
                conn.commit()
        except Exception as e:
            LOG.exception('error saving users info', e)
            raise e

测试脚本（新增用户）

python unittest做单元测试

 
import unittest, requests
 
class GuardianTestCase(unittest.TestCase):
    guardian_base = 'http://127.0.0.1:7070'
    user_token = keystone.user_token('default', 'tsbc', '123456') # 调用keystone生成登录用户的token
    
def test_create_user_api(self):
        # 添加对应headers 及 tonken 用于数据传参和登录认证使用
        myheaders = {'Content-Type': 'application/json'}
        myheaders['X-Subject-Token'] = self.user_token
        # 接口测试数据
        user = {
            "user": {
                "domain_id": "default",
                "enabled": True,
                "name": "tsbc1",
                "password": "123456",
                "realname": "Ray",
                "email": "tsbc@123.com",
                "tel":"+86-029-88880000",
                "phone": "13800138000",
                "org_id": "7d22d25a-b648-4ad3-a834-2398eb989a9e",
                "org_array":None,
                "comments": "备注信息"
            }
        }
        # 使用requests的post方法进行请求路由
        _resp = requests.post(self.guardian_base + '/guardian/users', json=user, headers=myheaders, verify=True)
        resp_json = _resp.content
        # 对返回结果及状态进行校验
        if _resp.ok:
            self.assertIsNotNone(resp_json)
            LOG.info('create users, response: {0}'.format(resp_json))
        else:
            LOG.error('error create users, response: {0}, status: {1}'.format(resp_json, _resp.status_code))

接口：修改用户

修改用户和新增用户传递的参数差不多，使用的是http的patch请求，route要添加user_id的参数，其他实现和新增差不多代码不重复贴了

 
@app.route('/guardian/users/<string:user_id>', methods=['PATCH'])
@require_role('guardian_admin')
@login_required
def update_user(user_id):
    """
    此处省略接收请求数据代码
    """
    try:
            resp = usermgt.update_user(domain_id, user_id, username, user_passwd, project_id, enabled) # 更新keystone 用户密码、状态
            if resp is not None:
                # 更新用户基本信息与新增是同一个接口，执行SQL时使用PostGresql特性，相同user_id数据只更新不会再插入
                GuardianModelDao._save_update_users_info(user_id, username, real_name, email, tel, phone, org_id, org_array, comments)
                return wrap_response(SUCCESS_MSG, 200)    # 更新成功返回 200
        except Exception as e:
            LOG.exception('error modify user', e)
            abort(500)    # 更新失败500

测试脚本(修改用户)

 
def test_update_user_api(self):
 
        myheaders = {'Content-Type': 'application/json'}
        myheaders['X-Subject-Token'] = self.user_token
 
        user = {
            "user": {
                "domain_id": "default",
                "enabled": True,
                "name": "tsbc1",
                "password": "111111",
                "realname": "xioohua",
                "email": "haoydun@123.com",
                "tel": "+86-029-88880000",
                "phone": "13800138000",
                "org_id": "7d22d25a-b648-4ad3-a834-2398eb989a9e",
                "comments": "备注信息"
            }
        }
 
        _resp = requests.patch(self.guardian_base + '/guardian/users/0d26176579e94d558c4c2d864b003241',   # 路由中添加要修改的用户ID requests使用 patch方法
                               json=user, headers=myheaders, verify=True)
        resp_json = _resp.content
        # 返回值进行校验
        if _resp.ok:
            self.assertIsNotNone(resp_json)
            LOG.info('update users, response: {0}'.format(resp_json))
        else:
            LOG.error('error update users, response: {0}, status: {1}'.format(resp_json, _resp.status_code))

接口：查询用户

根据用户名查询用户，使用http的get请求

 
@app.route('/guardian/users/<string:user_name>', methods=['GET'])    # 查询使用 GET 方法
@require_role('guardian_admin')
@login_required
def query_user_by_username(user_name):
 
    try:
        user_info = GuardianModelDao._query_users_info_by_name(user_name)    # 调用后台查询函数
        if len(user_info):
            return wrap_response(jsonpickle.dumps(user_info), 200)    # 成功返回200 并把返回内容转换成 json格式
        else:
            return wrap_response("[]", 200)   # 否则返回空
    except Exception as e:
        LOG.exception('error query user by name', e)
        abort(500)

后台函数

 
@classmethod
    def _query_users_info_by_id(cls, user_id):
        LOG.debug('start list users info...')
        users = []
        with TxConn() as conn:
            try:
                with conn.cursor() as cursor:
                    data = (user_id,)
                    cursor.execute(cls.query_users_info_by_id, data)
                    # 查询结构如果是行数据，放进list中进行返回
                    for (user_id, real_name, email, tel, phone, org_id, org_array, comments, create_time, update_time) in cursor:
                        user = Users(user_id, real_name, email, tel, phone, org_id, org_array, comments, create_time, update_time)
                        users.append(user)
            except Exception as e:
                LOG.exception('error query users !')
                raise e
        return users

测试脚本(查询用户)

 
 def test_query_user_api(self):
 
        myheaders = {'Content-Type': 'application/json'}
        myheaders['X-Subject-Token'] = self.user_token
 
        _resp = requests.get(self.guardian_base + '/guardian/users/tsbc', headers=myheaders, verify=True) # 链接中加入用户名
        resp_json = _resp.text
        # 验证判断
        if _resp.ok:
            self.assertIsNotNone(resp_json)
            LOG.info('query users, response: {0}'.format(resp_json))
        else:
            LOG.error('error query users, response: {0}, status: {1}'.format(resp_json, _resp.status_code))

接口：删除用户

根据用户ID进行删除，使用http的delete请求

 
@app.route('/guardian/users/<string:user_id>', methods=['DELETE']) # 传入用户id 使用 delete请求
@require_role('guardian_admin')
@login_required
def delete_user(user_id):
    if user_id is None:
        abort(400)
    try:
        s_code = usermgt.delete_user(user_id)
        if s_code == 204:
            return wrap_response(SUCCESS_MSG, status=s_code) # 返回状态码
    except Exception as e:
        LOG.exception('error delete users info', e)
        abort(500)

后台函数

 
def delete_user(user_id):
    myheaders = {'Content-Type': 'application/json'}
    myheaders['X-Auth-Token'] = settings.KEYSTONE_ADMIN_TOKEN
    _resp = requests.delete(settings.KEYSTONE_BASE + '/v3/users/'+user_id,
                            headers=myheaders, verify=False)
 
    if _resp.status_code == 204:
        LOG.debug('delete users -- response {0}'.format(_resp))
        return _resp.status_code
    else:
        LOG.error('error list users, response: {0}'.format(_resp))
        raise KeystoneException('error delete user {0} with response status {1}'
                                .format(user_id, _resp.status_code), _resp.status_code)

测试脚本(删除用户)

 
def test_delete_users_api(self):
 
        myheaders = {'Content-Type': 'application/json'}
        myheaders['X-Subject-Token'] = self.user_token
        _resp = requests.delete(self.guardian_base + '/guardian/users/tsbc1', headers=myheaders, # 使用 delete 请求
                                verify=True)
        resp_json = _resp.content
        # 根据返回状态进行校验
        if _resp.ok:
            self.assertIsNotNone(resp_json)
            LOG.info('delete users api, response: {0}'.format(resp_json))
        else:
            LOG.error('error delete users api, response: {0}, status: {1}'.format(resp_json, _resp.status_code))

下面提供给前端API文档

## CREATE USER

1. URL:' /guardian/users'

2. METHOD: POST

3. body:

 
{
            "user": {
                "domain_id": "default",
                "enabled": True,
                "name": "tsbc1",
                "password": "111111",
                "realname": "Ray",
                "email": "haoydun@123.com",
                "tel":"+86-029-88880000",
                "phone": "13800138000",
                "org_id": "7d22d25a-b648-4ad3-a834-2398eb989a9e",
                "org_array": null,
                "comments": "备注信息"
            }
}

## UPDATE USER

1. URL:' /guardian/users/<string:user_id>'

2. METHOD: PATCH2. body:

 
{
            "user": {
                "domain_id": "default",
                "enabled": True,
                "name": "tsbc1",
                "password": "111111",
                "realname": "Ray",
                "email": "tsbc@vip.qq.com",
                "tel":"+86-029-88880000",
                "phone": "13800138000",
                "org_id": "7d22d25a-b648-4ad3-a834-2398eb989a9e",
                "org_array": null,
                "comments": "备注信息"
            }
  }

该接口可以用于修改密码，禁用用户

## QUERY USER BY NAME
1. URL:' /guardian/users/<string:user_name>'

2. METHOD: GET

## DELETE USER
1. URL: '/guardian/users/<string:user_id>'

2. METHOD: DELETE

总结：

感谢每一个认真阅读我文章的人！！！

作为一位过来人也是希望大家少走一些弯路，如果你不想再体验一次学习时找不到资料，没人解答问题，坚持几天便放弃的感受的话，在这里我给大家分享一些自动化测试的学习资源，希望能给你前进的路上带来帮助。

软件测试面试文档

我们学习必然是为了找到高薪的工作，下面这些面试题是来自阿里、腾讯、字节等一线互联网大厂最新的面试资料，并且有字节大佬给出了权威的解答，刷完这一套面试资料相信大家都能找到满意的工作。

视频文档获取方式：
这份文档和视频资料，对于想从事【软件测试】的朋友来说应该是最全面最完整的备战仓库，这个仓库也陪伴我走过了最艰难的路程，希望也能帮助到你！以上均可以分享，点下方小卡片即可自行领取。

本文内容由网友自发贡献，版权归原作者所有，本站不承担相应法律责任。如您发现有涉嫌抄袭侵权的内容，请联系:hwhale#tublm.com(使用前将#替换为@)

软件测试

技术分享

自动化测试

http

网络协议