Thank you Myron for the answer. It's very clear.
I built the above example just to explain my question/case. Let's not worry
about the particulars of this example.
It is the idea of validating the context and your answer covers it very
well. I am also on the same page like you are.
However, there are some people that they find strange using "expect(...)"
to validate the test context. In order to make it clear, they like the idea
to check the validity
of the test context. What they don't like is the usage of the "expect(...)"
method to do that.
Is there any other RSpec suggested method to do that? For example, when
using Jasmine, I use the "fail(...)" method which I call if
the test context is not valid (in the before block again). If not, I was
thinking about creating a helper method like "validate_context(...)" that
would
raise an error like "RSpec::ContextValidityError", and that I could call
for such cases in which I want to validate the context.
What are your thoughts?
Panos
On Friday, September 2, 2016 at 1:49:42 PM UTC+1, Myron Marston wrote:
>
> In general, yes, I've also found it useful to express pre-conditions as
> expectations in some situations. I usually do this when I've realized that
> the test I'm writing has a non-obvious assumption that, if it was ever
> violated due to a config change or some other change, would cause the test
> to pass without exercising the code I intend it to. You don't want a
> future change to cause a test to start passing only because it's no longer
> exercising the logic it is intended to cover.
>
> That said, in the example you gave, it's not clear to me what the value
> is. Are you concerned that `build` might one day change to no longer
> return a Hash?
>
> HTH,
> Myron
>
> On Friday, September 2, 2016 at 5:25:46 AM UTC-7, Panayotis Matsinopoulos
> wrote:
>>
>> Hi,
>>
>> I would like to have your opinion on the following.
>>
>> I have found very useful to verify the context of an example. What do I
>> mean? See the following example:
>>
>> describe MyClass do
>> describe '#process' do
>> context 'when something is a Hash' do
>> let(:something) { build(:something) }
>>
>> before do
>> expect(something).to be_a(Hash)
>> end
>>
>> it 'returns an open struct' do
>> expect(subject.process(something)).to be_a(OpenStruct)
>> end
>> end
>> end
>> end
>>
>> Do you see that in the before block I am verifying that the context of
>> the "it" is correct?
>>
>> To me, this has been proven very useful. Because sometimes, the context
>> setup code might not be doing what we are expecting it to do. The
>> verification code does the double check.
>>
>> What is your opinion about this practice? Is this a practice that you
>> apply too? Or am I overprotective?
>>
>> Panos
>>
>>
--
You received this message because you are subscribed to the Google Groups
"rspec" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/rspec/11f36542-5540-4538-bf5d-955df5783d28%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
