测试平台系列(73) 设计测试计划功能.md

大家好~我是米洛！
我正在从0到1打造一个开源的接口测试平台, 也在编写一套与之对应的完整教程，希望大家多多支持。
欢迎关注我的公众号米洛的测开日记，获取最新文章教程!

回顾

上一节我们简单介绍了下APScheduler，这一节我们来编写测试计划相关内容。

设计测试计划表

测试计划，其实也可以叫测试集合，它是我们一组用例的集合。并且有着对应的特征:

• 定时执行

• 执行完毕后通知方式

• 通过率低于多少发送邮件/钉钉等通知

• 优先级

• 涵盖多少case

• case失败重试间隔 等等

  这里可能比较怪的就是把测试计划+测试集合给耦合到一起了。

  基于上面的思路，我们就可以来设计测试计划表了, 具体字段的含义可以参照注释。

  （数据表的设计不一定完美，后续一般会根据业务需求扩展）

from sqlalchemy import Column, String, TEXT, UniqueConstraint, BOOLEAN, SMALLINT, INT
from sqlalchemy.dialects.mysql import TINYTEXT

from app.models.basic import PityBase


class PityTestPlan(PityBase):
    project_id = Column(INT, nullable=False)
    # 测试计划执行环境, 可以多选
    env = Column(String(64), nullable=False)
    # 测试计划名称
    name = Column(String(32), nullable=False)
    # 测试计划优先级
    priority = Column(String(3), nullable=False)
    # cron表达式
    cron = Column(String(12), nullable=False)
    # 用例列表
    case_list = Column(TEXT, nullable=False)
    # 并行/串行(是否顺序执行)
    ordered = Column(BOOLEAN, default=False)
    # 通过率低于这个数会自动发通知
    pass_rate = Column(SMALLINT, default=80)
    # 通知用户，目前只有邮箱，后续用户表可能要完善手机号字段，为了通知
    receiver = Column(TEXT)
    # 通知方式 0: 邮件 1: 钉钉 2: 企业微信 3: 飞书 支持多选
    msg_type = Column(TINYTEXT)
    # 单次case失败重试间隔，默认2分钟
    retry_minutes = Column(SMALLINT, default=2)
    # 测试计划是否正在执行中
    state = Column(SMALLINT, default=0, comment="0: 未开始 1: 运行中")

    __table_args__ = (
        UniqueConstraint('project_id', 'name', 'deleted_at'),
    )

    __tablename__ = "pity_test_plan"

    def __init__(self, project_id, env, case_list, name, priority, cron, ordered, pass_rate, receiver, msg_type,
                 retry_minutes, user, state=0, id=None):
        super().__init__(user, id)
        self.env = ",".join(map(str, env))
        self.case_list = ",".join(map(str, case_list))
        self.name = name
        self.project_id = project_id
        self.priority = priority
        self.ordered = ordered
        self.cron = cron
        self.pass_rate = pass_rate
        self.receiver = ",".join(map(str, receiver))
        self.msg_type = ",".join(map(str, msg_type))
        self.retry_minutes = retry_minutes
        self.state = state

这里值得注意的是，我们定义的FORM数据，环境列表env、接收人receiver、用例列表case_list都是数组，我们需要进行一波转换。

编写CRUD方法

• 抽离异步分页方法和where方法

在DatabaseHelper类中添加pagination方法，接受page和size，session和sql参数，先读出sql匹配到的总数，如果为0则直接return，否则通过offset和limit获取到对应分页的数据。

where方法是用于改进我们平时的多条件查询，类似这种:

• 编写新增测试计划方法

可以看到改造之后，我们只需要调用where方法，不需要写if name != "":这样的语句了。

因为我们不允许同一个项目里面出现同名的测试计划，所以条件是项目id+name不能重复。

• 编写增改删方法

from sqlalchemy import select

from app.models import async_session, DatabaseHelper
from app.models.schema.test_plan import PityTestPlanForm
from app.models.test_plan import PityTestPlan
from app.utils.logger import Log


class PityTestPlanDao(object):
    log = Log("PityTestPlanDao")

    @staticmethod
    async def list_test_plan(page: int, size: int, project_id: int = None, name: str = ''):
        try:
            async with async_session() as session:
                conditions = [PityTestPlan.deleted_at == 0]
                DatabaseHelper.where(project_id, PityTestPlan.project_id == project_id, conditions) \
                    .where(name, PityTestPlan.name.like(f"%{name}%"), conditions)
                sql = select(PityTestPlan).where(conditions)
                result, total = await DatabaseHelper.pagination(page, size, session, sql)
                return result, total
        except Exception as e:
            PityTestPlanDao.log.error(f"获取测试计划失败: {str(e)}")
            raise Exception(f"获取测试计划失败: {str(e)}")

    @staticmethod
    async def insert_test_plan(plan: PityTestPlanForm, user: int):
        try:
            async with async_session() as session:
                async with session.begin():
                    query = await session.execute(select(PityTestPlan).where(PityTestPlan.project_id == plan.project_id,
                                                                             PityTestPlan.name == plan.name,
                                                                             PityTestPlan.deleted_at == 0))
                    if query.scalars().first() is not None:
                        raise Exception("测试计划已存在")
                    plan = PityTestPlan(**plan.dict(), user=user)
                    await session.add(plan)
        except Exception as e:
            PityTestPlanDao.log.error(f"新增测试计划失败: {str(e)}")
            raise Exception(f"添加失败: {str(e)}")

    @staticmethod
    async def update_test_plan(plan: PityTestPlanForm, user: int):
        try:
            async with async_session() as session:
                async with session.begin():
                    query = await session.execute(
                        select(PityTestPlan).where(PityTestPlan.id == plan.id, PityTestPlan.deleted_at == 0))
                    data = query.scalars().first()
                    if data is None:
                        raise Exception("测试计划不存在")
                    DatabaseHelper.update_model(data, plan, user)
        except Exception as e:
            PityTestPlanDao.log.error(f"编辑测试计划失败: {str(e)}")
            raise Exception(f"编辑失败: {str(e)}")

    @staticmethod
    async def delete_test_plan(id: int, user: int):
        try:
            async with async_session() as session:
                async with session.begin():
                    query = await session.execute(
                        select(PityTestPlan).where(PityTestPlan.id == plan.id, PityTestPlan.deleted_at == 0))
                    data = query.scalars().first()
                    if data is None:
                        raise Exception("测试计划不存在")
                    DatabaseHelper.delete_model(data, user)
        except Exception as e:
            PityTestPlanDao.log.error(f"删除测试计划失败: {str(e)}")
            raise Exception(f"删除失败: {str(e)}")
            
    @staticmethod
    async def query_test_plan(id: int) -> PityTestPlan:
        try:
            async with async_session() as session:
                sql = select(PityTestPlan).where(PityTestPlan.deleted_at == 0, PityTestPlan.id == id)
                data = await session.execute(sql)
                return data.scalars().first()
        except Exception as e:
            PityTestPlanDao.log.error(f"获取测试计划失败: {str(e)}")
            raise Exception(f"获取测试计划失败: {str(e)}")

基本思路差不多，老CRUD了！这里就不多说了，对sqlalchemy的async内容不了解的可以去官网看看demo。

这个query方法，是给定时任务查询测试计划数据使用。由于做了软删除，会导致经常忘记带上deleted_at==0的条件。

编写相关接口（app/routers/testcase/testplan.py）

from fastapi import Depends

from app.dao.test_case.TestPlan import PityTestPlanDao
from app.handler.fatcory import PityResponse
from app.models.schema.test_plan import PityTestPlanForm
from app.routers import Permission
from app.routers.testcase.testcase import router
from config import Config


@router.get("/plan/list")
async def list_test_plan(page: int, size: int, project_id: int = None, name: str = "", user_info=Depends(Permission())):
    try:
        data, total = await PityTestPlanDao.list_test_plan(page, size, project_id, name)
        return PityResponse.success_with_size(PityResponse.model_to_list(data), total=total)
    except Exception as e:
        return PityResponse.failed(str(e))


@router.get("/plan/insert")
async def insert_test_plan(form: PityTestPlanForm, user_info=Depends(Permission(Config.MANAGER))):
    try:
        await PityTestPlanDao.insert_test_plan(form, user_info['id'])
        return PityResponse.success()
    except Exception as e:
        return PityResponse.failed(str(e))


@router.get("/plan/update")
async def update_test_plan(form: PityTestPlanForm, user_info=Depends(Permission(Config.MANAGER))):
    try:
        await PityTestPlanDao.update_test_plan(form, user_info['id'])
        return PityResponse.success()
    except Exception as e:
        return PityResponse.failed(str(e))


@router.get("/plan/delete")
async def delete_test_plan(id: int, user_info=Depends(Permission(Config.MANAGER))):
    try:
        await PityTestPlanDao.delete_test_plan(id, user_info['id'])
        return PityResponse.success()
    except Exception as e:
        return PityResponse.failed(str(e))

这边我们把测试计划的增删改权限赋予MANAGER，但其实最好是能给对应的项目经理，不过那样会稍微复杂点。我们暂时先偷个懒，或许再完善。

今天的内容就分享到这里，下节会介绍测试计划如何和APScheduler结合起来，之后便是编写测试计划的前端页面部分，完成定时任务功能。