Hi Jon, I get what you’re saying, and I agree in general, but this case is a little different. Every account that gets activated gets sent to Stripe, so unless I’m testing the activation process itself, the account will get sent to Stripe. And as the account is built out to facilitate whatever test we are running, #send_to_stripe can be sent more than once. In fact, it’s designed to be able to be called more often than strictly necessary without a bad outcome. So I’m not worried about extra calls, but I need to stop any calls from actually proceeding into the routine and calling the Stripe API.
I am looking into alternatives to using allow_any_instance_of, but then I have to put my tests in the middle of where the user gets instantiated (for example, when a web request comes in it does a #find_by(auth_token: xxx), which means in every controller test for a logged in user I will have to stub that call so that I can return an object that I have access to. In an unrelated question, I’ve been using Webrat for matchers like has_selector?, but when I moved to Rspec 3.99, it stopped recognizing Webrat matchers. Is there a require and/or include that must be added to get access to those? I looked at the webrat docs, and what they recommended (e.g. require “webrat-rails”) does not work. I saw one StackOverflow answer that indicated that we should use Capybara for this. Is that the case? > On Sep 25, 2020, at 9:28 AM, Jon Rowe <[email protected]> wrote: > > Hello! > > We consider `allow_any_instance_of(…)` a feature for working with legacy code > and not to be the first choice, but it’s currently still supported, you can > find the documentation here: > > https://relishapp.com/rspec/rspec-mocks/v/3-9/docs/working-with-legacy-code/any-instance > > To replace it we recommend you be specific about your instances. > > e.g for your example: > > ``` > config.before(:each) do > allow(User).to receive(:find) do > user = instance_double(User) > allow(user).to receive(:send_to_stripe).and_return(true) > user > end > end > ``` > > But really, I wouldn’t do that at all, because why does every test need this? > I’d more likely recommend stubbing the right call at the right time, with the > right calls. Otherwise my code could call send_to_stripe at any point and my > tests would never notice... > > Cheers > Jon > ---------------- > [email protected] > https://jonrowe.co.uk > > On 25 September 2020 at 16:31, Jack Royal-Gordon wrote: >> I’ve just upgraded from Rspec 2.99 to 3.9, and have discovered that >> “allow_any_instance_of” is no longer supported. I’m not sure how best to go >> about replacing it, as it is in use in most of my tests. I have a User >> class, and most tests involve a user. Behind the scenes, many of the user >> methods call the Stripe API to adjust subscriptions, view and preview >> invoices, etc. Currently, I’ve got the following in “spec_helper.rb”: >> >> config.before(:each) do >> allow_any_instance_of(User).to receive(:send_to_stripe).and_return(true) >> end >> >> In most cases the calls to the Stripe API go through that method (I used >> instance stubs as necessary in tests that use other facets of the Stripe >> API). I understand that the preferred approach is to stub an instance rather >> than a whole class. But to do this, everywhere I create a user I would have >> to add “allow(xxx).to receive(:send_to_stripe).and_return(true). Most (if >> not all) of my users are created with calls to FactoryBot.In >> <http://factorybot.in/> the interest of DRYing my tests, is there a way I >> can leverage that to set a stub on the instance being created by FactoryBot? > > -- > 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] > <mailto:[email protected]>. > To view this discussion on the web visit > https://groups.google.com/d/msgid/rspec/dejalu-217-5242bc7e-977b-41a6-b172-a7e4525c8701%40jonrowe.co.uk > > <https://groups.google.com/d/msgid/rspec/dejalu-217-5242bc7e-977b-41a6-b172-a7e4525c8701%40jonrowe.co.uk?utm_medium=email&utm_source=footer>. -- 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 view this discussion on the web visit https://groups.google.com/d/msgid/rspec/C41FB2DB-5FE4-4F70-8B60-C9C25197677D%40pobox.com.
