Effects with Multiple outputs

[bux225]

I'm doing a standard parse of an API cal for both the status code (for error UX) and for the response. The below works but I have no idea if it's the right/best way to do this. I'm having a difficult time writing tests against it. Specifically, the two outputs at the end

In Reducer:

` case .authenticate:

        return .run { [state = state] send in

            let url = URL(string: "https://account.maxar.com/auth/realms/mds/protocol/openid-connect/token")!
            var request = URLRequest(url: url)
            request.httpMethod = "POST"
            request.setValue("application/x-www-form-urlencoded", forHTTPHeaderField: "Content-Type")
            request.httpBody = "username=\(state.username)&password=\(state.password)&client_id=\(state.clientId)&grant_type=password".data(using: .utf8)
            //print(String(decoding: request.httpBody!, as: UTF8.self))
            let (data, body) = try await URLSession.shared
                .data(for: request)
            var code = 0
            var response = MGPAuthResponse()
            if let http = body as? HTTPURLResponse {
                code = http.statusCode
                if http.statusCode == 200 {
                    response = try JSONDecoder().decode(MGPAuthResponse.self, from: data)
                }
            }
            await send(.setAuthResponse(response))
            await send(.setHttpStatusCode(code))

        }`

[mbrandonw]

Hi @bux225, this code isn't very testable because you are reaching out into the real world to make a network request (the line where you do URLSession.shared). Instead you should design a dependency that puts an interface in front of this request and use that instead. Are you familiar with that technique?

[bux225]

I appreciate the response. No, I don't think I'm familiar with that technique. Would you have a pointer to docs or examples? Much appreciated

[mbrandonw]

We have documentation on dependencies here, and if you follow the Essentials tutorial you will see how to use dependencies from beginning to end.

[mbrandonw]

That code is from the "Side effect" section of the Essentials tutorial. But if you keep reading on, in particular the "Testing your feature" section, you will see how to control the dependency and make it testable.

[bux225]

I see it now in section 4. Thanks!!

[bux225]

OK, I've read and coded the example of numbersapi behind a dependency and the testing. I still think I have a very naive question but the test doesn't seem to do anything at all. I can change the url even and the test passes since it's not even testing the call. Why is this desired? I'd like to ensure that I could do the following tests on my actual API in my app:

• Test a 200 happy path and that I populate a codable model (tests that the API schema has not changed underneath me)
• Test a 403 to ensure I am providing the user a good UX

[jshier]

For the 200 case I suggest you directly test the live version of your networking dependency then, without involving TCA at all. But the 403 case sounds like something you'd simulate and propagate through your stores, so that would be done using typical TCA testing.

You probably want to start a new discussion for this, as it doesn't seem to have anything to do with the topic at hand.

[mbrandonw]

Yes, please start a discussion because it is an important topic of why you want to sometimes test with mocks and why you may sometimes want to test against a live API.

[bux225]

Will do. The OP was about multiple effects. In this case it is the URLResponse status code as well as the data payload itself from an API call. But I'll reframe into a new discussion