dongsheng8664 2010-08-12 17:37
浏览 109
已采纳

没有用phpUnit测试异常?

I'm writing some unit test with phpUnit to test a Zend Framework application and I've got some issues with testing an exception in the changePassword function. The test doesn't fail, but in the coverage tool which generates html the "throw new Exception($tr->translate('userOldPasswordIncorrect'));" line isn't tested.

public function changePassword(array $data, $id)
{
    $user = $this->_em->find('Entities\User', (int) $id);

    $oldPassword = sha1(self::$_salt . $data['oldPassword']);
    if ($user->getPassword() !== $oldPassword) {
        $tr = PC_Translate_MySQL::getInstance();
        throw new Exception($tr->translate('userOldPasswordIncorrect'));
    }

    $user->setPassword(sha1(self::$_salt . $data['password']));

    $this->_em->persist($user);
    $this->_em->flush();
}

The unit test which should test the exception:

/**
 * @depends testFindByAuth
 * @expectedException Exception
 */
public function testChangePasswordWrongOldPassword()
{
    $this->_dummyUser = $this->_user->findByAuth($this->_dummyEmail, $this->_dummyPassword, $this->_reseller);

    // Try to change the password with a wrong oldPassword
    $data['oldPassword'] = 'wrongOldPassword';
    $data['password'] = $this->_dummyNewPassword;

    $this->_user->changePassword($data, $this->_dummyUser->getId());
}

I'll hope somebody can tell me what I'm doing wrong.

Update

The problem was inside the PC_Translate_MySQL::getInstance() method. There was thrown an exception. And as I was testing on getting a general exception this ofcourse passed. Solution don't use a general Exception in the changePassword method.

  • 写回答

1条回答 默认 最新

  • dqusbxh44823 2010-08-12 18:00
    关注

    My guess? There's an exception being thrown from PC_Translate_MySQL::getInstance()...

    That's the trouble with using a single Exception. It makes it tougher to check what exception was thrown. I'd suggest changing the changePassword method to throw a different exception. Perhaps a InvalidArgumentException, or a RuntimeException. And then test for that one.

    Personally, I use custom exceptions all the time for this reason.

    try {
    } catch (DatabaseQueryException $e) {
        // Handle database error
    } catch (DatabaseConnectionException $e) {
        // We never even connected...
    } catch (InvalidArgumentException $e) {
        //...
    }
    

    I, as a rule, never use catch (Exception $e) for that very reason. You never know what exception you caught. (I do have a custom exception handler defined, so I don't fatal if the application doesn't catch. Instead it shows a 500 error and logs the exception)...

    本回答被题主选为最佳回答 , 对您是否有帮助呢?
    评论

报告相同问题?

悬赏问题

  • ¥15 像这种代码要怎么跑起来?
  • ¥15 怎么改成循环输入删除(语言-c语言)
  • ¥15 安卓C读取/dev/fastpipe屏幕像素数据
  • ¥15 pyqt5tools安装失败
  • ¥15 mmdetection
  • ¥15 nginx代理报502的错误
  • ¥100 当AWR1843发送完设置的固定帧后,如何使其再发送第一次的帧
  • ¥15 图示五个参数的模型校正是用什么方法做出来的。如何建立其他模型
  • ¥100 描述一下元器件的基本功能,pcba板的基本原理
  • ¥15 STM32无法向设备写入固件