当我的测试失败或发生异常时，我需要始终注销我的应用程序。我该怎么做？

Question

我有一个测试用例如下：

@Test

public void checkSomething()

{

//line1

//line2

//line3

//line4[Exception occurs here]

//line5

//line6

//line7 homepage.Logout();

}

现在，例如，如果第 4 行中发生异常，则我的应用程序将永远不会注销 [line7]。这将导致我的进一步测试用例失败，因为它们将无法登录，因为用户会话将处于活动状态。我如何使注销始终在测试过早失败时发生？

我尝试将注销逻辑放在@AfterMethod中。它工作正常，但这是在配置方法（如@AfterMethod）中编写测试代码的最佳实践吗？

Answer 1

将注销放入会很好，但请确保您以有效的方式执行此操作。@AfterMethod如果仅测试失败，则检查注销避免使用尝试捕获，因为它等待给定的时间（隐式等待）来检查存在的元素，然后进入捕获块而不是使用列表引用以下代码使用@AfterMethod&nbsp;@AfterMethod&nbsp;&nbsp;public void screenShot(ITestResult result){&nbsp; &nbsp; &nbsp; &nbsp;if(ITestResult.FAILURE==result.getStatus()){&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; List<WebElement> username = driver.findElement(By.locator); // element which displays if user is logged in&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; if(!username.isEmpty())&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; // steps to logout will go here&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; }&nbsp; &nbsp; &nbsp; &nbsp;}&nbsp; }另一种选择是您可以使用TestNG监听器。在类中实现并重写方法，如下所示ITestListeneronTestFailure@Overridepublic void onTestFailure(ITestResult result) {&nbsp; &nbsp; &nbsp; if(ITestResult.FAILURE==result.getStatus()){&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; List<WebElement> username = driver.findElement(By.locator); // element which displays if user is logged in&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; if(!username.isEmpty())&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; // steps to logout will go here&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; }&nbsp; &nbsp; &nbsp; &nbsp;}}在测试中添加下面的标签.xml<listeners>&nbsp; &nbsp;<listener class-name="com.pack.listeners.TestListener"/> // your created class name with package which implemented ITestListener</listeners>

Answer 2

我使用 C# 工作，但概念在所有语言中很可能是相同的。在我的例子中，我在我的基类中使用所谓的“TearDown”标签来标记一个应该在测试后始终运行的方法。所有测试都从基类继承此方法，并进行相应的处理。在过去的几年里，这已经很好了，据我所知，任何类似的概念都被认为是最佳实践。在伪代码中：&nbsp; &nbsp; [TearDown]&nbsp; &nbsp; public void Cleanup()&nbsp; &nbsp; {&nbsp; &nbsp; &nbsp; &nbsp; try&nbsp; &nbsp; &nbsp; &nbsp; {&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; Logout();&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; OtherStuffLikeClosingDriver();&nbsp; &nbsp; &nbsp; &nbsp; }&nbsp; &nbsp; &nbsp; &nbsp; catch (Exception ex)&nbsp; &nbsp; &nbsp; &nbsp; {&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; Log(ex);&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; // Obviously, this logging function needs to generate logs that are easily readable, based on the given exception.&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; FinishTest(testInstance, testName); // Handles critical flows that should always be finished (and "should" not be able to error out)&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; throw ex;&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;// In my case, throwing the exception again makes sure that the exception is shown in the test output directly. This often speeds up the first diagnose of a failed test run.&nbsp; &nbsp; &nbsp; &nbsp; }&nbsp; &nbsp; }只要确保处理异常等：@AfterMethod中的逻辑不应该被意外问题打断。