docs: Web + App UI 自动化测试最佳实践 v1.0
This commit is contained in:
@@ -0,0 +1,835 @@
|
||||
# Web + App UI 自动化测试最佳实践
|
||||
|
||||
> **版本**: v1.0 | **日期**: 2026-07-15 | **维护者**: QE Fleet Team
|
||||
>
|
||||
> **定位**: 面向测试工程师和 AI Agent 的统一 UI 自动化实践指南。人类工程师可按章查阅、逐条落地;AI Agent(`web-executor`、`mobile-executor`)可将本文档规则作为脚本生成依据。
|
||||
>
|
||||
> **前置阅读**: 建议先阅读 `knowledge_base/01_standards/test_case_template.md`(用例规范)和 `definition_of_done.md`(完成定义)。
|
||||
|
||||
---
|
||||
|
||||
## 目录
|
||||
|
||||
1. [总则与适用范围](#1-总则与适用范围)
|
||||
2. [技术栈与框架选型](#2-技术栈与框架选型)
|
||||
3. [Web UI 自动化最佳实践](#3-web-ui-自动化最佳实践)
|
||||
4. [App UI 自动化最佳实践](#4-app-ui-自动化最佳实践)
|
||||
5. [通用工程实践](#5-通用工程实践)
|
||||
6. [UI 自动化用例范例](#6-ui-自动化用例范例)
|
||||
7. [附录](#7-附录)
|
||||
|
||||
---
|
||||
|
||||
## 1. 总则与适用范围
|
||||
|
||||
### 1.1 适用场景
|
||||
|
||||
| 场景 | Web (Playwright) | App (Appium) | 说明 |
|
||||
| :--- | :---: | :---: | :--- |
|
||||
| PC Web 管理后台 | ✅ | — | 平台端、商家端等中后台系统 |
|
||||
| 移动 H5 页面 | ✅ | — | 微信 H5、手机浏览器页面 |
|
||||
| Android 原生 APP | — | ✅ | UiAutomator2 驱动 |
|
||||
| iOS 原生 APP | — | ✅ | XCUITest 驱动 |
|
||||
| 微信小程序 | ✅ | — | 小程序本质是 WebView,用 Playwright 操作 |
|
||||
| 混合场景(Web + App 联动) | ✅ | ✅ | 如后台发起操作 → APP 端验证结果 |
|
||||
|
||||
### 1.2 与项目文档体系的关系
|
||||
|
||||
```text
|
||||
本文档(最佳实践)
|
||||
├── 引用 agents/execute/web_executor.md(Web 执行规则)
|
||||
├── 引用 agents/execute/mobile_executor.md(App 执行规则)
|
||||
├── 符合 knowledge_base/01_standards/test_case_template.md(用例规范)
|
||||
├── 符合 knowledge_base/01_standards/definition_of_done.md(完成定义)
|
||||
└── 补充 knowledge_base/03_best_practices/(本目录)的自动化工程视角
|
||||
```
|
||||
|
||||
### 1.3 基础原则
|
||||
|
||||
1. **自动化不是目标,质量才是。** 只自动化那些需要频繁回归、手工执行成本高的用例。
|
||||
2. **稳定性 > 覆盖率。** 100 个可靠的用例比 500 个频繁误报的用例有价值。
|
||||
3. **可维护性优先。** 定位策略、数据准备、断言逻辑必须清晰可读。
|
||||
4. **失败必须可追溯。** 每次失败必须产出截图 + 日志 + 操作序列,让接手者无需复现即可定位。
|
||||
|
||||
---
|
||||
|
||||
## 2. 技术栈与框架选型
|
||||
|
||||
### 2.1 框架选择决策表
|
||||
|
||||
| 被测端 | 推荐框架 | 当前版本 | 驱动/引擎 |
|
||||
| :--- | :--- | :--- | :--- |
|
||||
| PC Web | **Playwright** (Python) | ≥1.40 | Chromium / Firefox / WebKit |
|
||||
| 移动 H5 | **Playwright** (Python) | ≥1.40 | Chromium (移动视口模拟) |
|
||||
| Android APP | **Appium** (Python) | ≥3.0 | UiAutomator2 |
|
||||
| iOS APP | **Appium** (Python) | ≥3.0 | XCUITest |
|
||||
| 微信小程序 | **Playwright** (Python) | ≥1.40 | Chromium (特定 User-Agent) |
|
||||
|
||||
### 2.2 为什么选 Playwright 而非 Selenium?
|
||||
|
||||
| 维度 | Playwright | Selenium |
|
||||
| :--- | :--- | :--- |
|
||||
| 自动等待 | 内置(actionability checks) | 需手动 `WebDriverWait` |
|
||||
| 网络拦截 | `page.route()` 原生支持 | 需 CDP 或代理 |
|
||||
| 多浏览器 | 单 API,同代码 | 各浏览器 driver 独立 |
|
||||
| 调试工具 | `--debug`、Trace Viewer、Codegen | 较弱的原生日志 |
|
||||
| 执行速度 | 快(单一 WebSocket 连接) | 慢(HTTP 往返) |
|
||||
|
||||
> **团队决策**: 本项目统一使用 Playwright。旧 Selenium 用例应逐步迁移。
|
||||
|
||||
### 2.3 为什么选 Appium 而非专用框架?
|
||||
|
||||
| 维度 | Appium | Espresso / XCUITest 原生 |
|
||||
| :--- | :--- | :--- |
|
||||
| 跨平台复用 | 高(一套 API 双平台) | 低(两套独立代码) |
|
||||
| 学习成本 | 低(WebDriver 协议) | 高(平台特有 API) |
|
||||
| CI 集成 | 成熟(Server-Client 架构) | 需额外工具链 |
|
||||
| 脚本语言 | Python / Java / JS 等 | Java/Kotlin (Android) / Swift (iOS) |
|
||||
|
||||
> **团队决策**: 本项目统一使用 Appium (Python Client)。除非有性能压测或白盒测试需求,不引入原生框架。
|
||||
|
||||
---
|
||||
|
||||
## 3. Web UI 自动化最佳实践
|
||||
|
||||
### 3.1 定位策略(Element Locator)
|
||||
|
||||
**优先级从高到低:**
|
||||
|
||||
| 优先级 | 策略 | 示例 | 稳定性 |
|
||||
| :---: | :--- | :--- | :---: |
|
||||
| 1 | `data-testid` | `page.getByTestId('submit-btn')` | ⭐⭐⭐⭐⭐ |
|
||||
| 2 | `role + name` | `page.getByRole('button', {name: '提交'})` | ⭐⭐⭐⭐ |
|
||||
| 3 | `text` | `page.getByText('确认删除')` | ⭐⭐⭐ |
|
||||
| 4 | `placeholder` | `page.getByPlaceholder('请输入手机号')` | ⭐⭐⭐ |
|
||||
| 5 | `label` | `page.getByLabel('用户名')` | ⭐⭐⭐ |
|
||||
| 6 | CSS selector | `page.locator('.form .submit-btn')` | ⭐⭐ |
|
||||
| 7 | XPath | `page.locator('//div[@class="foo"]//button')` | ⭐ |
|
||||
|
||||
**原则**:
|
||||
|
||||
- ✅ **永远优先使用语义化定位**(`data-testid`、`role`、`text`)。它们不随 UI 重构而断裂。
|
||||
- ❌ **避免使用 CSS class**(尤其是 `_hash_abc123` 这类编译期生成的 class)。
|
||||
- ❌ **避免 XPath 长链**(`/html/body/div[3]/div[2]/span[1]`),改一个嵌套层级就全挂。
|
||||
- ⚠️ **不确定的 selector 用注释标注 `# ⚠️ 需确认 selector`**,不要捏造。
|
||||
|
||||
**反例 vs 正例:**
|
||||
|
||||
```python
|
||||
# ❌ 反例:CSS class 不稳定,UI 改动就失效
|
||||
page.locator('.ant-btn.ant-btn-primary.btn-submit-form-v2').click()
|
||||
|
||||
# ❌ 反例:深层 XPath,嵌套改动就挂
|
||||
page.locator('/html/body/div[3]/div[2]/div[1]/form/div[4]/button[1]').click()
|
||||
|
||||
# ✅ 正例:data-testid,与视觉样式解耦
|
||||
page.getByTestId('order-submit-btn').click()
|
||||
|
||||
# ✅ 正例:语义化 role + name
|
||||
page.getByRole('button', name='提交订单').click()
|
||||
|
||||
# ✅ 正例:文本匹配
|
||||
page.getByText('确认收货').click()
|
||||
```
|
||||
|
||||
### 3.2 等待策略
|
||||
|
||||
Playwright 内置自动等待(actionability checks),大多数场景**不需要**手动等待。只有在以下情况才显式加等待:
|
||||
|
||||
| 场景 | 推荐写法 | 说明 |
|
||||
| :--- | :--- | :--- |
|
||||
| 页面跳转后等待元素 | `page.waitForSelector('[data-testid="result"]')` | 等待关键元素出现 |
|
||||
| 等待网络请求完成 | `page.waitForResponse(lambda r: '/api/order' in r.url)` | 等待 API 返回 |
|
||||
| 等待特定状态文本 | `expect(page.getByText('提交成功')).toBeVisible()` | 等待 UI 状态变化 |
|
||||
| 等待元素消失 | `expect(page.locator('.loading')).toBeHidden()` | 等待 Loading 结束 |
|
||||
|
||||
**反例 vs 正例:**
|
||||
|
||||
```python
|
||||
# ❌ 反例:硬编码 sleep——浪费时间且不可靠
|
||||
import time
|
||||
page.click('button')
|
||||
time.sleep(3) # 网速慢时 3 秒不够,快时浪费
|
||||
assert page.getByText('成功').is_visible()
|
||||
|
||||
# ✅ 正例:使用 Playwright 自动等待 + 显式等待结合
|
||||
page.getByTestId('submit-btn').click()
|
||||
expect(page.getByText('提交成功')).toBeVisible(timeout=10000) # 最多等 10 秒
|
||||
```
|
||||
|
||||
### 3.3 Page Object 模式
|
||||
|
||||
将页面元素定位和操作封装为独立的 Page Object 类,实现**用例逻辑与页面结构的分离**。
|
||||
|
||||
**结构约定**:
|
||||
|
||||
```python
|
||||
# pages/order_page.py
|
||||
from playwright.sync_api import Page, expect
|
||||
|
||||
class OrderPage:
|
||||
"""订单确认页"""
|
||||
|
||||
def __init__(self, page: Page):
|
||||
self.page = page
|
||||
# 所有定位器集中定义在 __init__
|
||||
self.submit_btn = page.getByTestId('order-submit-btn')
|
||||
self.amount_text = page.getByTestId('order-total-amount')
|
||||
self.success_tip = page.getByText('下单成功')
|
||||
|
||||
def get_total_amount(self) -> str:
|
||||
"""获取订单总金额"""
|
||||
return self.amount_text.inner_text()
|
||||
|
||||
def submit(self) -> None:
|
||||
"""提交订单并等待成功提示"""
|
||||
self.submit_btn.click()
|
||||
expect(self.success_tip).toBeVisible(timeout=10000)
|
||||
|
||||
def verify_amount(self, expected: str) -> None:
|
||||
"""验证订单金额"""
|
||||
expect(self.amount_text).toContainText(expected)
|
||||
```
|
||||
|
||||
**原则**:
|
||||
|
||||
- 每个 Page Object 对应**一个页面或一个功能区域**(弹窗、抽屉可独立)。
|
||||
- 定位器集中声明在 `__init__()`,方法只调用已声明的定位器。
|
||||
- 断言尽可能封装在 Page Object 方法内(如 `submit()` 自带成功断言)。
|
||||
- 跨页面复用操作封装为**公共组件**(如 `LoginComponent`、`NavBar`),而不是在每个 Page Object 中重复。
|
||||
|
||||
### 3.4 截图策略
|
||||
|
||||
本项目的截图策略与 `agents/execute/web_executor.md` 对齐:
|
||||
|
||||
| 策略 | 配置参数 | 说明 |
|
||||
| :--- | :--- | :--- |
|
||||
| **失败截图**(默认开启) | `screenshot_on_failure: true` | 断言失败/异常时自动截图 |
|
||||
| **每步截图**(默认关闭) | `screenshot_on_step: false` | 量大,仅在调试阶段开启 |
|
||||
| **关键节点截图** | 始终开启 | P0 用例的关键验证点 |
|
||||
|
||||
```python
|
||||
# ✅ 正例:失败自动截图(conftest.py 全局 fixture)
|
||||
@pytest.fixture(autouse=True)
|
||||
def screenshot_on_failure(request, page):
|
||||
yield
|
||||
if request.node.rep_call.failed:
|
||||
path = f"output/screenshots/{request.node.name}_{datetime.now():%Y%m%d_%H%M%S}.png"
|
||||
page.screenshot(path=path, full_page=True)
|
||||
|
||||
# ✅ 正例:关键节点显式截图
|
||||
page.getByTestId('pay-btn').click()
|
||||
page.screenshot(path=f"output/screenshots/after_payment_{order_id}.png")
|
||||
```
|
||||
|
||||
截图命名规范: `{用例编号}_{操作描述}_{时间戳}.png`,如 `ORDER_001_click_submit_20260715_143020.png`。
|
||||
|
||||
### 3.5 重试与容错
|
||||
|
||||
| 配置项 | 推荐值 | 说明 |
|
||||
| :--- | :--- | :--- |
|
||||
| 失败用例重试 | 1 次 | `retry_on_failure: 1` |
|
||||
| 操作超时 | 30 秒 | 单次操作(点击/输入)的超时 |
|
||||
| 断言超时 | 10 秒 | `expect().toBeVisible()` 的等待上限 |
|
||||
| 页面加载超时 | 60 秒 | `page.goto()` 的导航超时 |
|
||||
|
||||
```python
|
||||
# ✅ 正例:关键操作加重试(仅在网络相关失败时)
|
||||
from playwright.sync_api import TimeoutError as PlaywrightTimeout
|
||||
|
||||
def safe_click(locator, max_retries=2):
|
||||
for attempt in range(max_retries):
|
||||
try:
|
||||
locator.click(timeout=30000)
|
||||
return
|
||||
except PlaywrightTimeout:
|
||||
if attempt == max_retries - 1:
|
||||
raise
|
||||
time.sleep(2)
|
||||
```
|
||||
|
||||
> ⚠️ **注意**: 重试仅适用于网络抖动、临时渲染问题。断言失败(业务逻辑错误)不应重试,直接标记失败。
|
||||
|
||||
### 3.6 多浏览器覆盖
|
||||
|
||||
```python
|
||||
# conftest.py — 按 fleet_config.yml 配置的 browsers 参数化
|
||||
import pytest
|
||||
from playwright.sync_api import sync_playwright
|
||||
|
||||
BROWSERS = ["chromium", "firefox", "webkit"] # 或从 fleet_config.yml 读取
|
||||
|
||||
@pytest.fixture(params=BROWSERS)
|
||||
def browser_context(request):
|
||||
with sync_playwright() as p:
|
||||
browser = getattr(p, request.param).launch(headless=True)
|
||||
context = browser.new_context(viewport={"width": 1920, "height": 1080})
|
||||
yield context
|
||||
context.close()
|
||||
browser.close()
|
||||
```
|
||||
|
||||
> **建议**: P0 用例覆盖 Chromium + 至少一个额外浏览器;P1 用例默认仅 Chromium。
|
||||
|
||||
---
|
||||
|
||||
## 4. App UI 自动化最佳实践
|
||||
|
||||
### 4.1 定位策略(Element Locator)
|
||||
|
||||
**优先级从高到低:**
|
||||
|
||||
| 优先级 | 策略 | Android 示例 | iOS 示例 | 稳定性 |
|
||||
| :---: | :--- | :--- | :--- | :---: |
|
||||
| 1 | `accessibility_id` | `driver.find_element(AppiumBy.ACCESSIBILITY_ID, "submit")` | 同 | ⭐⭐⭐⭐⭐ |
|
||||
| 2 | `id` | `driver.find_element(AppiumBy.ID, "com.app:id/btn")` | `driver.find_element(AppiumBy.ID, "submitBtn")` | ⭐⭐⭐⭐ |
|
||||
| 3 | `class_name + index` | `driver.find_elements(AppiumBy.CLASS_NAME, "android.widget.Button")[0]` | 同 | ⭐⭐ |
|
||||
| 4 | XPath | `driver.find_element(AppiumBy.XPATH, "//*[@text='确认']")` | 同 | ⭐ |
|
||||
|
||||
**原则**:
|
||||
|
||||
- ✅ 优先让开发在关键元素上添加 `content-desc`(Android)或 `accessibilityIdentifier`(iOS),对应 `accessibility_id` 定位。
|
||||
- ❌ 避免 XPath,移动端 UI 层级深且频繁变动。
|
||||
- ❌ 避免坐标点击(`tap([x, y])`),不同分辨率/设备上位置完全不同。
|
||||
- ⚠️ 不确定的定位方式用注释标注 `# ⚠️ 需确认定位方式`。
|
||||
|
||||
**反例 vs 正例:**
|
||||
|
||||
```python
|
||||
# ❌ 反例:坐标点击——换一台设备就挂
|
||||
driver.tap([(300, 500)])
|
||||
|
||||
# ❌ 反例:深层 XPath
|
||||
driver.find_element(AppiumBy.XPATH, "//android.view.View[3]/android.widget.Button[2]")
|
||||
|
||||
# ✅ 正例:accessibility_id 最稳定
|
||||
driver.find_element(AppiumBy.ACCESSIBILITY_ID, "submit_order_btn").click()
|
||||
|
||||
# ✅ 正例:id 定位(Android 资源 ID)
|
||||
driver.find_element(AppiumBy.ID, "com.example.app:id/btn_pay").click()
|
||||
```
|
||||
|
||||
### 4.2 Android vs iOS 双平台差异
|
||||
|
||||
```python
|
||||
# ✅ 正例:平台感知的 Driver Factory
|
||||
import platform
|
||||
from appium import webdriver
|
||||
from appium.options.android import UiAutomator2Options
|
||||
from appium.options.ios import XCUITestOptions
|
||||
|
||||
def create_driver():
|
||||
if platform.system() == "Android":
|
||||
options = UiAutomator2Options()
|
||||
options.device_name = "Android Emulator"
|
||||
options.app_package = "com.example.app" # ⚠️ 必改
|
||||
options.app_activity = ".MainActivity" # ⚠️ 必改
|
||||
options.no_reset = True
|
||||
options.automation_name = "UiAutomator2"
|
||||
return webdriver.Remote("http://localhost:4723", options=options)
|
||||
else:
|
||||
options = XCUITestOptions()
|
||||
options.device_name = "iPhone 15"
|
||||
options.bundle_id = "com.example.app" # ⚠️ 必改
|
||||
options.no_reset = True
|
||||
options.automation_name = "XCUITest"
|
||||
return webdriver.Remote("http://localhost:4723", options=options)
|
||||
```
|
||||
|
||||
| 差异点 | Android | iOS |
|
||||
| :--- | :--- | :--- |
|
||||
| 驱动 | UiAutomator2 | XCUITest |
|
||||
| 应用标识 | `appPackage` + `appActivity` | `bundleId` |
|
||||
| 等待策略 | `WebDriverWait` + `expected_conditions` | 同,但部分 condition 支持有限 |
|
||||
| 手势 API | `driver.swipe()` | `driver.swipe()`(API 一致) |
|
||||
| 截图 | `driver.save_screenshot(path)` | 同 |
|
||||
|
||||
### 4.3 手势操作
|
||||
|
||||
```python
|
||||
# 滑动(从下往上)
|
||||
def swipe_up(driver, duration_ms=500):
|
||||
size = driver.get_window_size()
|
||||
driver.swipe(
|
||||
start_x=size['width'] // 2,
|
||||
start_y=int(size['height'] * 0.8),
|
||||
end_x=size['width'] // 2,
|
||||
end_y=int(size['height'] * 0.2),
|
||||
duration=duration_ms
|
||||
)
|
||||
|
||||
# 长按
|
||||
from appium.webdriver.common.appiumby import AppiumBy
|
||||
from appium.webdriver.common.touch_action import TouchAction
|
||||
|
||||
def long_press(driver, element, duration_ms=1000):
|
||||
TouchAction(driver).long_press(element, duration=duration_ms).release().perform()
|
||||
|
||||
# 双指缩放(地图/图片)
|
||||
def pinch(driver, scale=0.5):
|
||||
# ⚠️ 推荐使用 MultiAction API,因设备差异较大,需实际调试
|
||||
pass
|
||||
```
|
||||
|
||||
### 4.4 等待策略
|
||||
|
||||
移动端网络不稳定,等待策略比 Web 更关键:
|
||||
|
||||
```python
|
||||
from selenium.webdriver.support.ui import WebDriverWait
|
||||
from selenium.webdriver.support import expected_conditions as EC
|
||||
from appium.webdriver.common.appiumby import AppiumBy
|
||||
|
||||
# ✅ 正例:显式等待元素可见后再操作
|
||||
wait = WebDriverWait(driver, timeout=30)
|
||||
|
||||
element = wait.until(
|
||||
EC.visibility_of_element_located((AppiumBy.ACCESSIBILITY_ID, "order_status"))
|
||||
)
|
||||
assert element.text == "已完成"
|
||||
|
||||
# ❌ 反例:硬编码 sleep
|
||||
import time
|
||||
time.sleep(5) # 网络差时 5 秒不够
|
||||
driver.find_element(AppiumBy.ACCESSIBILITY_ID, "order_status")
|
||||
```
|
||||
|
||||
### 4.5 截图策略
|
||||
|
||||
| 策略 | 说明 |
|
||||
| :--- | :--- |
|
||||
| **失败截图**(默认开启) | 断言失败/异常时立即截图 |
|
||||
| **关键页面截图** | 登录后首页、核心操作页、结果页 |
|
||||
| **状态变化截图** | 状态流转前后各截一张 |
|
||||
|
||||
```python
|
||||
def take_screenshot(driver, name: str):
|
||||
timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
|
||||
path = f"output/screenshots/{name}_{timestamp}.png"
|
||||
driver.save_screenshot(path)
|
||||
return path
|
||||
|
||||
# 失败自动截图(conftest.py)
|
||||
@pytest.hookimpl(hookwrapper=True)
|
||||
def pytest_runtest_makereport(item, call):
|
||||
outcome = yield
|
||||
if outcome.get_result().failed and "driver" in item.funcargs:
|
||||
driver = item.funcargs["driver"]
|
||||
take_screenshot(driver, f"FAIL_{item.name}")
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 5. 通用工程实践
|
||||
|
||||
以下实践同时适用于 Web 和 App 自动化。
|
||||
|
||||
### 5.1 用例独立性
|
||||
|
||||
每个测试用例必须是**独立且可任意顺序执行**的。
|
||||
|
||||
```python
|
||||
# ✅ 正例:每个用例独立准备数据 + 独立清理
|
||||
class TestOrderSubmit:
|
||||
|
||||
@pytest.fixture(autouse=True)
|
||||
def setup_and_teardown(self):
|
||||
self.test_data = create_test_order() # 每个用例创建独立数据
|
||||
yield
|
||||
cleanup_test_order(self.test_data['id']) # 用例结束后清理
|
||||
|
||||
def test_normal_submit(self):
|
||||
result = submit_order(self.test_data['id'])
|
||||
assert result.status == "SUCCESS"
|
||||
|
||||
def test_duplicate_submit(self):
|
||||
submit_order(self.test_data['id']) # 第一次提交
|
||||
result = submit_order(self.test_data['id']) # 重复提交
|
||||
assert result.status == "ALREADY_SUBMITTED"
|
||||
|
||||
# ❌ 反例:用例之间存在数据依赖
|
||||
def test_a_create_order():
|
||||
global order_id # 全局变量,换执行顺序就挂
|
||||
order_id = create_order()
|
||||
|
||||
def test_b_pay_order():
|
||||
pay_order(order_id) # 依赖 test_a 先执行,且 order_id 不为空
|
||||
```
|
||||
|
||||
### 5.2 数据驱动
|
||||
|
||||
将测试数据与用例逻辑分离,同一套用例覆盖多组数据:
|
||||
|
||||
```python
|
||||
# ✅ 正例:参数化测试数据
|
||||
import pytest
|
||||
|
||||
TEST_DATA = [
|
||||
# (支付方式, 金额, 预期结果)
|
||||
("支付宝", 100, "支付成功"),
|
||||
("微信支付", 50, "支付成功"),
|
||||
("银行卡", 1000, "支付成功"),
|
||||
("支付宝", 0, "金额不能为0"),
|
||||
("支付宝", -10, "金额不能为负"),
|
||||
]
|
||||
|
||||
class TestPayment:
|
||||
@pytest.mark.parametrize("pay_method,amount,expected", TEST_DATA)
|
||||
def test_payment(self, pay_method, amount, expected):
|
||||
result = pay(method=pay_method, amount=amount)
|
||||
assert result.message == expected
|
||||
```
|
||||
|
||||
### 5.3 环境配置管理
|
||||
|
||||
所有环境相关信息统一管理,不硬编码在测试脚本中:
|
||||
|
||||
```python
|
||||
# config/test_config.yaml
|
||||
base_url: "https://test.example.com"
|
||||
timeout_seconds: 120
|
||||
accounts:
|
||||
admin: { phone: "13800138000", password: "Test123456" }
|
||||
user: { phone: "13800138001", password: "Test123456" }
|
||||
browsers: ["chromium", "firefox"]
|
||||
```
|
||||
|
||||
```python
|
||||
# ✅ 正例:从配置读取
|
||||
import yaml
|
||||
from pathlib import Path
|
||||
|
||||
def load_config():
|
||||
with open(Path(__file__).parent / "config/test_config.yaml", encoding="utf-8") as f:
|
||||
return yaml.safe_load(f)
|
||||
|
||||
config = load_config()
|
||||
page.goto(config["base_url"])
|
||||
page.getByPlaceholder("手机号").fill(config["accounts"]["admin"]["phone"])
|
||||
|
||||
# ❌ 反例:硬编码
|
||||
BASE_URL = "https://test.example.com" # 换环境要改代码
|
||||
ADMIN_PHONE = "13800138000" # 换账号要改代码
|
||||
```
|
||||
|
||||
### 5.4 报告与日志
|
||||
|
||||
```python
|
||||
import logging
|
||||
from datetime import datetime
|
||||
|
||||
# 日志配置
|
||||
logging.basicConfig(
|
||||
level=logging.INFO,
|
||||
format="%(asctime)s [%(levelname)s] %(message)s",
|
||||
handlers=[
|
||||
logging.FileHandler(f"output/logs/test_{datetime.now():%Y%m%d_%H%M%S}.log"),
|
||||
logging.StreamHandler()
|
||||
]
|
||||
)
|
||||
logger = logging.getLogger(__name__)
|
||||
|
||||
class TestOrder:
|
||||
def test_submit(self):
|
||||
logger.info(f"开始执行: 验证正常下单")
|
||||
try:
|
||||
# ...
|
||||
logger.info(f"✅ 通过: 验证正常下单")
|
||||
except Exception as e:
|
||||
logger.error(f"❌ 失败: 验证正常下单 — {e}")
|
||||
raise
|
||||
```
|
||||
|
||||
**报告输出要求**:
|
||||
|
||||
- 每个用例执行后产出: 通过/失败/跳过 + 耗时 + 失败原因(如有)+ 截图路径
|
||||
- 批量运行结束后产出汇总: `通过 X / 失败 Y / 跳过 Z`,列出所有失败用例的截图链接
|
||||
|
||||
### 5.5 CI 集成
|
||||
|
||||
```yaml
|
||||
# CI 配置示例 (GitHub Actions / Gitea Actions)
|
||||
jobs:
|
||||
web-tests:
|
||||
runs-on: ubuntu-latest
|
||||
steps:
|
||||
- uses: actions/checkout@v4
|
||||
- uses: actions/setup-python@v5
|
||||
with:
|
||||
python-version: '3.11'
|
||||
- run: pip install -r requirements.txt
|
||||
- run: playwright install chromium
|
||||
- run: pytest tests/web/ --screenshot on_failure --junitxml=report.xml
|
||||
|
||||
app-tests:
|
||||
runs-on: macos-latest # iOS 测试必须 macOS
|
||||
steps:
|
||||
- uses: actions/checkout@v4
|
||||
- run: |
|
||||
# 启动 Appium Server
|
||||
appium --log-level info --log appium.log &
|
||||
sleep 5
|
||||
- run: pytest tests/app/ --platform android --junitxml=report.xml
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 6. UI 自动化用例范例
|
||||
|
||||
> 以下范例直接从 `knowledge_base/03_best_practices/` 下的业务用例中提取自动化实现,展示"好的手工用例"如何映射为"可运行的自动化代码"。
|
||||
|
||||
### 6.1 Web — 支付流程 (Playwright)
|
||||
|
||||
**来源用例**: `payment_flow_cases.md` — `PAY_01_001`(验证正常支付流程成功)
|
||||
|
||||
```python
|
||||
# tests/web/test_payment.py
|
||||
import pytest
|
||||
from playwright.sync_api import Page, expect
|
||||
|
||||
class TestPaymentFlow:
|
||||
"""支付流程自动化测试"""
|
||||
|
||||
@pytest.fixture(autouse=True)
|
||||
def setup(self, page: Page):
|
||||
"""每个用例前登录 + 准备可售商品"""
|
||||
self.page = page
|
||||
login(page, user="user_A") # 封装好的登录函数
|
||||
self.item_id = add_item_to_cart(page, sku="SKU-001") # 添加商品到购物车
|
||||
yield
|
||||
cancel_remaining_orders(page) # 清理未支付订单
|
||||
|
||||
def test_normal_payment_alipay(self, page: Page):
|
||||
"""PAY_01_001: 验证正常支付流程成功"""
|
||||
# Step 1-2: 进入收银台,选择支付宝
|
||||
page.getByTestId("checkout-btn").click()
|
||||
expect(page.getByTestId("cashier-page")).toBeVisible()
|
||||
page.getByTestId("pay-method-alipay").click()
|
||||
|
||||
# Step 3: 完成支付
|
||||
page.getByTestId("confirm-pay-btn").click()
|
||||
|
||||
# 预期 1: 页面跳转支付成功页
|
||||
expect(page.getByText("支付成功")).toBeVisible(timeout=10000)
|
||||
|
||||
# 预期 2: 订单状态变更为"待发货"
|
||||
page.getByTestId("nav-orders").click()
|
||||
order_status = page.getByTestId(f"order-{self.item_id}-status")
|
||||
expect(order_status).toHaveText("待发货")
|
||||
|
||||
# 预期 3: 支付流水金额与订单实付一致
|
||||
page.getByTestId(f"order-{self.item_id}-detail").click()
|
||||
pay_amount = page.getByTestId("pay-record-amount").inner_text()
|
||||
order_amount = page.getByTestId("order-actual-amount").inner_text()
|
||||
assert pay_amount == order_amount, f"支付金额 {pay_amount} ≠ 订单金额 {order_amount}"
|
||||
```
|
||||
|
||||
### 6.2 Web — 异常流程 (Playwright)
|
||||
|
||||
```python
|
||||
def test_payment_balance_insufficient_switch_method(self, page: Page):
|
||||
"""PAY_01_003: 验证支付渠道返回余额不足时可切换支付方式"""
|
||||
page.getByTestId("checkout-btn").click()
|
||||
|
||||
# 选择银行卡支付
|
||||
page.getByTestId("pay-method-bankcard").click()
|
||||
|
||||
# 模拟支付渠道返回余额不足
|
||||
page.getByTestId("confirm-pay-btn").click()
|
||||
expect(page.getByText("余额不足")).toBeVisible(timeout=10000)
|
||||
|
||||
# 订单仍保持待支付状态
|
||||
page.getByTestId("nav-orders").click()
|
||||
order_status = page.getByTestId(f"order-{self.item_id}-status")
|
||||
expect(order_status).toHaveText("待支付")
|
||||
|
||||
# 可切换支付方式继续支付
|
||||
page.getByTestId(f"order-{self.item_id}-pay").click()
|
||||
page.getByTestId("pay-method-alipay").click()
|
||||
page.getByTestId("confirm-pay-btn").click()
|
||||
expect(page.getByText("支付成功")).toBeVisible(timeout=10000)
|
||||
```
|
||||
|
||||
### 6.3 Web — 表单验证 (Playwright)
|
||||
|
||||
```python
|
||||
# tests/web/test_login.py
|
||||
class TestLogin:
|
||||
"""登录功能自动化测试"""
|
||||
|
||||
@pytest.mark.parametrize("phone,password,expected_error", [
|
||||
("", "Test123456", "请输入手机号"),
|
||||
("13800138000", "", "请输入密码"),
|
||||
("13800138000", "wrong", "密码错误"),
|
||||
("13800000000", "Test123456", "账号不存在"),
|
||||
])
|
||||
def test_login_validation(self, page: Page, phone, password, expected_error):
|
||||
"""验证登录表单校验"""
|
||||
page.goto("/login")
|
||||
|
||||
if phone:
|
||||
page.getByPlaceholder("请输入手机号").fill(phone)
|
||||
if password:
|
||||
page.getByPlaceholder("请输入密码").fill(password)
|
||||
page.getByRole("button", name="登录").click()
|
||||
|
||||
expect(page.getByText(expected_error)).toBeVisible()
|
||||
```
|
||||
|
||||
### 6.4 App — 登录流程 (Appium)
|
||||
|
||||
```python
|
||||
# tests/app/test_login.py
|
||||
import pytest
|
||||
from appium.webdriver.common.appiumby import AppiumBy
|
||||
from selenium.webdriver.support.ui import WebDriverWait
|
||||
from selenium.webdriver.support import expected_conditions as EC
|
||||
|
||||
class TestAppLogin:
|
||||
"""APP 登录自动化测试"""
|
||||
|
||||
def test_login_success(self, driver):
|
||||
"""验证正常登录成功"""
|
||||
wait = WebDriverWait(driver, timeout=30)
|
||||
|
||||
# 输入手机号
|
||||
phone_input = wait.until(
|
||||
EC.visibility_of_element_located((AppiumBy.ACCESSIBILITY_ID, "login_phone_input"))
|
||||
)
|
||||
phone_input.send_keys("13800138000")
|
||||
|
||||
# 输入密码
|
||||
driver.find_element(AppiumBy.ACCESSIBILITY_ID, "login_password_input").send_keys("Test123456")
|
||||
|
||||
# 点击登录
|
||||
driver.find_element(AppiumBy.ACCESSIBILITY_ID, "login_submit_btn").click()
|
||||
|
||||
# 验证跳转主页
|
||||
home_tab = wait.until(
|
||||
EC.visibility_of_element_located((AppiumBy.ACCESSIBILITY_ID, "main_home_tab"))
|
||||
)
|
||||
assert home_tab.is_displayed()
|
||||
```
|
||||
|
||||
### 6.5 App — 订单列表滑动与加载 (Appium)
|
||||
|
||||
```python
|
||||
# tests/app/test_order_list.py
|
||||
class TestOrderList:
|
||||
"""APP 订单列表测试"""
|
||||
|
||||
def test_pull_to_refresh(self, driver):
|
||||
"""验证下拉刷新订单列表"""
|
||||
wait = WebDriverWait(driver, timeout=30)
|
||||
|
||||
# 等待列表加载
|
||||
wait.until(
|
||||
EC.visibility_of_element_located((AppiumBy.ACCESSIBILITY_ID, "order_list"))
|
||||
)
|
||||
|
||||
# 下拉刷新
|
||||
size = driver.get_window_size()
|
||||
driver.swipe(
|
||||
start_x=size['width'] // 2,
|
||||
start_y=int(size['height'] * 0.3),
|
||||
end_x=size['width'] // 2,
|
||||
end_y=int(size['height'] * 0.7),
|
||||
duration=500
|
||||
)
|
||||
|
||||
# 验证刷新成功提示
|
||||
toast = wait.until(
|
||||
EC.visibility_of_element_located((AppiumBy.XPATH, "//*[contains(@text,'刷新成功')]"))
|
||||
)
|
||||
assert toast is not None
|
||||
|
||||
def test_scroll_load_more(self, driver):
|
||||
"""验证上滑加载更多"""
|
||||
wait = WebDriverWait(driver, timeout=30)
|
||||
|
||||
# 获取当前列表项数量
|
||||
initial_items = len(driver.find_elements(AppiumBy.ACCESSIBILITY_ID, "order_card"))
|
||||
|
||||
# 滑动到底部
|
||||
size = driver.get_window_size()
|
||||
driver.swipe(
|
||||
start_x=size['width'] // 2,
|
||||
start_y=int(size['height'] * 0.8),
|
||||
end_x=size['width'] // 2,
|
||||
end_y=int(size['height'] * 0.2),
|
||||
duration=500
|
||||
)
|
||||
|
||||
# 验证加载更多
|
||||
time.sleep(1) # 等待渲染完成
|
||||
after_items = len(driver.find_elements(AppiumBy.ACCESSIBILITY_ID, "order_card"))
|
||||
assert after_items >= initial_items, f"加载更多后应不少于加载前: {after_items} < {initial_items}"
|
||||
```
|
||||
|
||||
---
|
||||
|
||||
## 7. 附录
|
||||
|
||||
### 7.1 Selector 速查表
|
||||
|
||||
#### Playwright (Web)
|
||||
|
||||
| 场景 | 写法 |
|
||||
| :--- | :--- |
|
||||
| 按 data-testid 点击 | `page.getByTestId('my-btn').click()` |
|
||||
| 按文本点击 | `page.getByText('确 定').click()` |
|
||||
| 按角色+名称点击 | `page.getByRole('button', name='提交').click()` |
|
||||
| 按 placeholder 填写 | `page.getByPlaceholder('请输入').fill('hello')` |
|
||||
| 按 label 填写 | `page.getByLabel('手机号').fill('13800138000')` |
|
||||
| 等待可见 | `expect(page.getByText('成功')).toBeVisible()` |
|
||||
| 等待隐藏 | `expect(page.locator('.loading')).toBeHidden()` |
|
||||
| 下拉框选择 | `page.selectOption('#type', 'value1')` |
|
||||
| 文件上传 | `page.setInputFiles('input[type=file]', 'path/to/file')` |
|
||||
| 确认弹窗 | `page.on('dialog', lambda d: d.accept())` |
|
||||
| iframe 内操作 | `page.frameLocator('#iframe').getByTestId('btn')` |
|
||||
|
||||
#### Appium (App)
|
||||
|
||||
| 场景 | Android 写法 | iOS 写法 |
|
||||
| :--- | :--- | :--- |
|
||||
| 按 accessibility_id | `driver.find_element(AppiumBy.ACCESSIBILITY_ID, "xxx")` | 同 |
|
||||
| 按 id | `driver.find_element(AppiumBy.ID, "com.app:id/xxx")` | `driver.find_element(AppiumBy.ID, "xxx")` |
|
||||
| 按文本 | `driver.find_element(AppiumBy.XPATH, "//*[@text='xxx']")` | `driver.find_element(AppiumBy.XPATH, "//*[@label='xxx']")` |
|
||||
| 滑动 | `driver.swipe(x1,y1,x2,y2,duration)` | 同 |
|
||||
| 截图 | `driver.save_screenshot(path)` | 同 |
|
||||
| 上下文切换 | `driver.switch_to.context('WEBVIEW')` | 同 |
|
||||
| 后台运行 | `driver.background_app(5)` | 同 |
|
||||
|
||||
### 7.2 常见坑与解决方案
|
||||
|
||||
| 坑 | 现象 | 解决方案 |
|
||||
| :--- | :--- | :--- |
|
||||
| CSS Module hash | `locator('.btn_abc123')` 找不到 | 用 `getByTestId` 或 `getByText` |
|
||||
| iframe 跨域 | `page.locator()` 在 iframe 内失效 | 用 `page.frameLocator('#iframe')` |
|
||||
| 非固定等待时间 | `time.sleep(3)` 不够或浪费 | 用 `expect(...).toBeVisible(timeout=N)` |
|
||||
| 元素被遮挡 | `click()` 报 Element not clickable | `locator.click(force=True)` 或先滚动到可见区域 |
|
||||
| 新标签页 | 点击后在新 tab 打开,操作还在旧 tab | `page.context().pages[1]` 获取新 tab |
|
||||
| Appium 连接超时 | `Connection refused` | 检查 Appium Server 是否启动: `appium --log-level info` |
|
||||
| iOS 签名过期 | XCUITest 启动失败 | 检查开发者证书有效期;模拟器无需签名 |
|
||||
| 设备屏幕尺寸差异 | 坐标点击错位 | 用相对坐标: `size['width'] // 2` |
|
||||
| Toast 消息难定位 | 一闪而过,无法截图 | 用 `//*[contains(@text,'关键字')]` 或加短延时后截图 |
|
||||
|
||||
### 7.3 与 Agent 执行规则的对应关系
|
||||
|
||||
| 本文档章节 | 对应 Agent 文件 | 关系 |
|
||||
| :--- | :--- | :--- |
|
||||
| §3 Web UI 自动化 | `agents/execute/web_executor.md` | 本文档是其执行规则的详细实践扩展 |
|
||||
| §4 App UI 自动化 | `agents/execute/mobile_executor.md` | 同上 |
|
||||
| §5.1 用例独立性 | `definition_of_done.md` §3 | 本文档给出代码层面的实现方式 |
|
||||
| §3.4 / §4.5 截图策略 | `web_executor.md` / `mobile_executor.md` 截图节 | 本文档与 Agent 配置参数对齐 |
|
||||
| §6 范例 | `knowledge_base/03_best_practices/` | 本文档补充了自动化代码视角 |
|
||||
|
||||
### 7.4 参考资源
|
||||
|
||||
- [Playwright Python API](https://playwright.dev/python/docs/intro)
|
||||
- [Appium Python Client](https://github.com/appium/python-client)
|
||||
- [Appium UiAutomator2 Driver](https://github.com/appium/appium-uiautomator2-driver)
|
||||
- [Appium XCUITest Driver](https://github.com/appium/appium-xcuitest-driver)
|
||||
- 本项目: `docs/EXECUTE_SETUP_GUIDE.md` — 执行环境搭建指南
|
||||
- 本项目: `knowledge_base/01_standards/test_case_template.md` — 用例编写规范
|
||||
- 本项目: `knowledge_base/01_standards/definition_of_done.md` — 完成定义
|
||||
Reference in New Issue
Block a user