Stub_chain together with should_receive

Testing RSpec
1

Hello.

I am trying to test if in a method calling chain one of the methods
get a specific parameter. In the below code for example MyModel must
receive the parameter 0 for the method “offset”. Unfortunately the
code below does not work. It seems it is not possible to mix
should_receive and stub_chain. How could I solve this?

MyModel.should_receive(:offset).with(0).stub_chain(:tag_counts, :offset,
:limit, :order).and_return([])

does not work!

Regards,
Kai

2

On Nov 23, 2010, at 7:12 PM, medihack wrote:

Hello.

I am trying to test if in a method calling chain one of the methods
get a specific parameter. In the below code for example MyModel must
receive the parameter 0 for the method “offset”. Unfortunately the
code below does not work. It seems it is not possible to mix
should_receive and stub_chain. How could I solve this?

MyModel.should_receive(:offset).with(0).stub_chain(:tag_counts, :offset, :limit,
:order).and_return([])

does not work!

As you’re finding out, chains like this are hard to test. That, and the
fact that writing code before tests significantly increases the
likelihood that the code will be hard to test.

There is no easy way to do what you’re trying to do with this code.
You’d have to create a series of test doubles, like this:

offset_return_value = double
offset_return_value.stub_chain(“limit.order”) { [] }
tag_counts_return_value = double
tag_counts_return_value.should_receive(:offset).with(0).and_return(offset_return_value)
MyModel.stub(:tag_counts) { tag_counts_return_value }

The problem with this sort of structure, besides it being hard to read,
is that you can’t change anything about the implementation of the method
that is invoking this chain without changing this example and likely
many others.

I’d recommend wrapping that chain in a method and specifying how that
method behaves when a 0 makes its way to the offset method vs how it
behaves otherwise.

HTH,
David

Regards,
Kai

rspec-users mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/rspec-users

Cheers,
David