Rspec没有捕捉到有目的的失败

我试图检查我的一个类方法是否响应无效输入有异常,但Rspec没有它.

我的班级文件:

class WhateverClass
  def run(options)
    if options['input'].nil? || options['input'].empty?
      fail ArgumentError, 'No input object provided in configuration.'
    end
  end
end

我的Rspec测试:

RSpec.describe WhateverClass do
  it 'should raise an ArgumentError when provided invalid input' do
    invalid_input = { 'nonsense' => 'here' }
    expect(WhateverClass.new.run(invalid_input)).to raise_error(ArgumentError)
  end
end

运行上述测试的结果如下:

1) WhateverClass should raise an ArgumentError when provided invalid input
   Failure/Error: expect(WhateverClass.new.run(invalid_input)).to raise_error(ArgumentError)
   ArgumentError:
     No input object provided in configuration.
   # ./whatever_class.rb:9:in `run'
   # ./spec/whatever_class_spec.rb:14:in `block (2 levels) in '
Finished in 0.00048 seconds (files took 1.51 seconds to load)
1 example, 1 failure

================================================== =

请注意,如果我这样做:

RSpec.describe WhateverClass do
  it 'should raise an ArgumentError when provided invalid input' do
    expect{ fail ArgumentError }.to raise_error(ArgumentError)
    invalid_input = { 'nonsense' => 'here' }
    expect(WhateverClass.new.run(invalid_input)).to raise_error(ArgumentError)
  end
end

第一个期望通过,但第二个预期失败.

如何使此测试正常运行？

事实证明,改变它:

expect(WhateverClass.new.run(invalid_input)).to raise_error(ArgumentError)

至:

expect { WhateverClass.new.run(invalid_input) }.to raise_error(ArgumentError)

解决了这个问题.希望这有助于其他人.