Reputation: 13816
In this mess and different schools of thought around xDD, I am looking for the right nomenclature of a specific type of test.
First of all, the architecture looks like this: the bulk of the application domain (the logic, which brings in business values, as it's often preached in DDD circles) is implemented as a library, decoupled from things like the user interface (be it GUI, web, or CLI) or database.
This library gets its input via commands, which are basically DTOs which only do some basic validation - for instance a number should be a number.
I want to write tests which just create a specific command, pass it over to the library, and wait for the result(s).
What I am looking for is the correct name for this type of test.
It is a kind of integration test, and it is a black-box testing (but I do inject the in memory database driver and other dependencies, prior to executing the test).
It could be an end-to-end test or a functional test, if I were to also include the actual user interface in my test - however I am not, the test looks as it has been written for the programmer, not the customer, since I am instantiating the DTO and I send it over.
It could also be an acceptance test, but, beside the previous paragraph (which still holds), this test is not executed by the customer.
So my question is: how do people call this type of test, in the different schools of thought? Some references would be welcome.
Upvotes: 1
Views: 90
Reputation: 14064
Maybe a subcutaneous test : http://martinfowler.com/bliki/SubcutaneousTest.html ?
Upvotes: 1