I have a bit of a question on how people are organizing their controller
specs, once you take user roles into account. I'm not entirely sure that
I've found a way to do it that feels "natural" to me.
So, say I've got a controller that I want to ensure is locked down to a
particular set of users. I can't decide how the layer the
describes/contexts:
describe PostsController do
context "as a normal user" do
before { logged_in }
describe "POST create" do
it "is forbidden" do
post :create, :post => {}
response.should be_forbidden
end
end
... Other specs ....
end
context "as an editor" do
before { logged_in.with_role :editor }
describe "POST create" do
...
end
end
This is the direction that the flow of the language seems right to me, when
it's dumped in the specdocs -- "PostsController, as a normal user POST
create is forbidden", but from another standpoint, it breaks up the
specification of a single method into a couple of different locations in the
file, and may require duplicating quite a bit of setup.
How does everyone else deal with this?
--
// anything worth taking seriously is worth making fun of
// http://blog.devcaffeine.com/
_______________________________________________
rspec-users mailing list
[email protected]
http://rubyforge.org/mailman/listinfo/rspec-users
