On Sun, Jul 6, 2008 at 5:43 PM, s.ross <[EMAIL PROTECTED]> wrote:
> Hello--
>
> I'm not sure if I'm pushing too far out of specing a given model, but here's
> what I want to express:
>
> class A < AR::Base
> has_many :bs
>
> def okay(segment)
> end
> end
>
> class B < AR::Base
> belongs_to :a
> end
>
> it A, "should increase the vote count for a given segment if okayed" do
> @a = A.create(stuff)
> @a.okay(10) # okay segment 10
> # here's where I'm having trouble...
> # a.bs should have one row, and it should have a segment number of 10 and
> various other stuff should happen
> # subsequent @a.okay with different values should transform numbers in a
> predictable way. This is handled in
> # the A model because the transforms occur only in Bs that belong to a
> given A.
> @a.bs.find_by_segment_number(10).thingie.should be(1) # not too expressive
> @a.okay(9)
> @a.bs.find_by_segment_number(9).thingie.should be(1) # still not too
> expressive
> @a.bs. find_by_segment_number(9).transformed_thingie.should
> be(something_else)
> end
>
> This kind of spec smells to me, yet I am having trouble figuring out how to
> explain, using rspec, exactly what the behavior is to be in a multi-step
> sequence.
>
> Anybody have thoughts regarding how this might be done better?
Hi,
You could do
lambda {
@a.okay(9)
}.should change { @a.bs.find_by_segment_number(9) }.
from(nil).to(1)
I do wonder if you might be able to make it a bit more
behavior-oriented, maybe something like
lambda {
@a.okay(9)
}.should change { @a.count_votes_for(9) }.by(1)
I don't know exactly what you're trying to get at though, so I can't
be more specific. At any rate, does using "should change" make it
more readable to you?
Pat
_______________________________________________
rspec-users mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/rspec-users
