I want to write tests for destroying objects on ruby on rails with cucumber and capybara and rack_test.
I think that I should write a scenario like:
When I try to delete X object
Then I should be on the index page for ...
And I should see "Object deleted successfully"
I don't want to test it by clicking on a link on the index page or the show/edit page because I think I should be testing something different than purely the delete action.
Also, clicking on a link means that the interface is ok, but what happens if someone send a request to delete an object he can't and the interface doesn't show the link?
Given this, I need to send a delete request in the first step and then inspect the response to that request.
I've find troubles trying to do that because when I implement the first step as delete url then when I try page.should have_contain(text) I get the error No response yet. Request a page first. (Rack::Test::Error)
What do you think about it?
check out how-do-you-post-to-a-url-in-capybara (answering another question) but there a several examples of how to use visit in a capybara test like
visit "item/:id/delete"
Related
I'm using Capybara to test a Rails app using RSpec. Calling visit "/bars" correctly takes me to the index page for the bars resource. But calling visit "/bars/1" (where a bar has been saved with id 1) mysteriously takes me to the index page as well, even though typing in the URL by hand works just fine.
The index page even has links to specific bars pages (like "/bars/1"), and while those links work perfectly in the browser, Capybara's click_link on those links takes me instead back to—you guessed it—the index page.
In both of these cases, the show method in the bars controller never even gets called when using Capybara, though it is called when I'm clicking around in the browser.
Does anyone have any idea what's going on?
If you have "saved it to db' and can view it in browser, it is development db. Capybara deals with test db which you probably don't have record there.
While you don't have such record and you may set 404 redirect to home, such result in Capybara is expected.
Another note, it's not good to use named id say 1 in test. There is no guarantee your last inserted record must be id 1. Better to use bars_path #bar or bars_path Bar.last
How do I write rspec tests defensively, so that in a scenario at least one expectation must be met yet the failure of others is accepted? (without the input changing). AND is easy enough by listing multiple expectations, but how is OR expressed?
As an example, a user has many posts, and user Bob hacks a form so that when he submits his create post form it sends the id of user Dunc. Currently the application ignores the passed Dunc id, and uses Bob's id as Bob is creating the post. So we could test that the newly created Post has Bob's user_id. However, if in future the code is refactored so that it returns an error message instead of assuming Bob's id, that test would wrongly fail. I want to test the intent, not the implementation.
So i need to test that either no post is created, or that if one is created, its for Bob.
This example is so simple it can be solved by testing
expect { run }.not_to change( Post.where(user_id: #other_user.id), :count )
However I'm looking for the general solution, in more complex cases there can be many conditions. How is "OR" achieved in Rspec? (or is it not possible?)
I don't think it is possible.
I do think you are mistaken when you say that you would be testing implementation, instead of intent in your example.
When you write a test, you test whether what comes out matches your expectation.
Creating a user is something completely different than returning error messages.
In my opinion it would be strange to say: when I do this, I expect this, or that, or that, or that to happen.
In my opinion you should write one test, that tests whether a user is created when you send the correct parameters, and another test that deals with what happens when a user tries to send illegal parameters.
Check edit at bottom of page
My boss has a sitemap up- it's basically just every route as a link, with a button to click that says "valid?" or "ignore" which will mark it valid or ignore it on the page.
He asked me to manually go through and click each link, test that page isn't a 500 or 404, and then mark it valid if it isn't.
This seems silly to me, as it is basically just a user facing test for working routes.
I could, in the same time, write out routing specs in Rspec for all those, but I guess he wants some sort of documentation that this is happening on the front end for himself and users.
I was thinking a fun way to work around this boring clicking would be to do it with some programming WHILE writing the specs. Makes him happy, and also adds actual value and test to the app that can be reused.
Is there a way to, in a spec, write something like:
links = page.all('a.routing-links)
link.each do |link|
link.click
if page status != 404 || 500
Link.find(id).update_attribute("verified", true)
end
end
I tried putting that in my spec, but when link.click hits an incorrect route, it stops the test (which makes sense, as that route is broken and this is a test.
What I'd like is to be able to take that error and use it to update the attribute of my model.
Am I going about this completely wrong? Any better ideas or inspiration?
Thanks
Edit
I agree with the poster who said this is better left to a script or rake task.
I'm a bit lost on how to write a script that will go to a page, find every link, record its status_code, and then find and update a model. Any suggestions or tips? Ideally it would be run within in the application, so that I could have access to my models and controllers.
Thanks
Personally I wouldn't actually put this in a spec since you're not actually expecting anything to fail.
Instead I'd create a quick script, or even rake task to run through the links as you described.
That being said, this article: http://agileleague.com/2012/12/rails-3-2-custom-error-pages-the-exceptions_app-and-testing-with-capybara/ details how to bypass the normal fail in these circumstances, namely:
In your config/environments/test.rb
config.consider_all_requests_local = false
config.action_dispatch.show_exceptions = true
Though this would affect all tests, which is quite possibly not what you want.
Also, a minor thing that you'd probably figure out in no time when testing this - you'll either need to revisit the list page after clicking the link, or rather relying on link clicks, you could visit the href instead which would be a bit quicker.
links = page.all('a.routing-links')
link.each do |link|
visit link[:href]
if page.status != 404 || 500
Link.find(id).update_attribute("verified", true)
end
end
I haven't tested that, so not sure if it would work like that, but you should be able to get the idea.
I'm trying to create a cucumber scenario that checks to see if elements are loaded for an 'edit posting' page. My trouble, however, is that I don't know how to create a path that will direct it to the page.
The general path is something like: /posting/id/edit
i.e. /posting/11/edit
Here is my posting.feature scenario
# Editing existing post
Scenario: Saving the edits to an existing post
Given I am logged in
Given there is a posting
Given I am on the edit posting page
When I fill in "posting_title" with "blah"
And I fill in "posting_location" with "blegh"
When I press "Update posting"
Then I should see "Posting was successfully updated."
I dabbled around with some Factory Girl stuff, but I don't have the knowledge to use it appropriately (if it offers a solution), and wasn't able to find a relevant example.
I've also seen a lot of suggestions with regards to 'Pickle', but if possible I'd like to avoid that route to keep things simple seeing as I have very limited experience. Thanks!
Is there a link on your website that would take someone to the edit page? Then you could do something like:
Given I am on the homepage
And I follow "Posts"
And I follow "Edit"
This assumes that there is a link on your homepage whose text is Posts, and then another one in the resulting page called Edit. This is the best way to accomplish this, because there should be a direct route to whatever page you are testing. Those steps are also provided in web_steps.rb
You could also make a custom step like you did there with Given I am on the edit posting page and the code would be something like:
Given /^I am on the edit posting page$/ do
visit("/posting/11/edit")
end
Which you of course could also generalize like I am on the edit posting page for posting 11. But in general, cucumber tests are acceptance tests, which means not bypassing things like this. You should have a link to the edit page that can be clicked.
I came up with a solution, but I am not sure of its validity in terms of how clean it is. I ended up using Factory Girl (installed the gem).
I kept my scenario the same.
Under features/step_definitions I created posting_steps.rb
Given /^there is a posting$/ do
Factory(:posting)
end
Under features/support I created a file factories.rb with the following inside:
Factory.define :posting do |f|
f.association :user
f.title 'blah'
f.location 'Some place'
end
In my paths.rb I used
when /the edit posting page/
edit_posting_path(Posting.first)
How it works (or at least how I think it works) is that as
Given there is a posting
is executed, the posting_step.rb is invoked (Factory(:posting) is basically Factory.create(:posting)), which in turn uses the factory definition I created in factories.rb. This leads to an instance of a posting being created.
Then in my paths.rb
when /the edit posting page/
edit_posting_path(Posting.first)
gets passed the id from the instance, to ultimately get a path that could resemble /posting/1/edit , and the test continues on its way!
If there are any corrections to be made, please let me know as I am just learning the ropes.
Hopefully this will help other newbies out there!
The scenario is as follows. My Order model has an after_create that contacts a remote payment gateway to retrieve a payment URL. In my Cucumber tests I don't want to perform this action, but return an arbitrary URL. My current cucumber tests looks like this:
Given there is a product "Product X"
When I enter my credentials
And I click "Order Now"
Then I should be redirected to "arbitrary url"
The problem is where/how do I make sure that my order model sets the url correctly and does not contact the remote payment gateway?
The wiki also has some tips on stubbing.
In features/support/env.rb I monkey-patched my Order model to set the arbitrary URL. This could possible be done with Mocha or something else as well, but there is not point in this case.
In my steps I can check the response for the correct redirect like this:
Then /^I should be redirected to the payment gateway$/ do
response.status.should eql("302 Found")
response.location.should eql(Order.last.payment_url)
end
Hope this helps for others as well. I'd still like to know if there's a better/cleaner way of achieving this goal.
If I understand what you are trying to do correctly, have a look at FakeWeb.