Is it possible to ‘feed’ expected responses from underlying methods to classes under test in JUnit? Of course, you can splatter the code with if-statements doing something else in case the code is “under test” or introduce an extra bag of objects either forwarding a request to an object (normal obperation) or doing that “something else”, but it leaves a feeling of smelly code and extra bugs. Is there a better approach to keep test code really out of the program code?
I imagine something like if ClassUnderTest.doSomething() calls ClassB.getValue(key) don’t call that function at all, instead, hand value foo as an imaginary return value of ClassB.getValue() back to ClassUnderTest.doSomething()? Can that be done some way? Can I access “key” later to see if my method posed the right question? (See: I don’t want to test the functionality of ClassB here, just the `doSomething()̀ method itself.)
I think what you are looking for are mocking frameworks.
The most popular these days seems to be Mockito.
Related
I am writing junit test cases for a spring boot application. I am having lot of doubts and I have listed them below.
Is it enough to write unit test cases only for service layer?
How to re-use the stubs / mocks created for the models. Each model is created with lot of dependencies. If we don't reuse them, we will be creating the same objects again and again. If we reuse how to accommodate the test values for all test cases?
Is there any best practices when creating the stubs?
Do we need to write the unit test cases for utility methods?
Rest controllers needs unit test cases?
I'll try to give the best general answers, though keep in mind that the specific case might require different approach.
No, generally speaking you should test everything that contains logic which might be subject to maintenance, and therefore unwanted changes.
I'm not quite sure on what is the problem here. The way I understood it, you want to write the stubs/mocks once and use them for multiple tests; maybe, you might use one test class and generate the stubs in a #Before, #BeforeClass annotated methods.
Well, the stub is intended to be used in place of a specific method when provided a given input. So, first, you should identify what inputs your stubbed method is going to receive and be sure you are passing them along (Note: if you provide the wrong inputs the stub won't work). Second, you need to stub the return object or the answer. Anyway, you might need to use sequential stubbing for cases when the method is called multiple times and different returns are required.
Yes, a maintenance change might cause the change in the behavior of such methods heavily affecting the product. You should always use JUnit to constraint the logic. Anyway, utility classes should be trivial and I don't expect it to be difficult for you to test them.
Like I already said, if it contains logic, yes, it should. Anyway I kinda remember there are different frameworks to mock rest calls.
Daniele
We've got a simple webservice for handling queries about our data. I'd like to make a set of asserts/case extentions that would provide high level methods for testing various aspects of the response. For instance I could write assertResultCountMinimum(int). The method would take care of building the query, executing it, and unpacking the response to validate the data. I'd also like the to
I want to make sure I have the right idea in my head about how to go about this.
First create a test case class of my own, with the right setup and teardown methods. For our purposes, MyTestCase. Then provide a series of classes that extend Assert with the new assert methods. The end user of these classes would extend MyTestCase and would use the asserts that I've created. This is the pattern I think I see in jWebUnit.
I feel like I'm mixing and matching junit 3 and 4 concepts. I'd love to have just junit 4 concepts. But I can't seem to line up in my head the proper way to build this. Also, the assert methods that belong to Junit's Assert class are all static. Some of my asserts would require requerying the webservice. This makes me think I should really just provide the asserts as a series of helper functions inside of MytestCase. The later gets the job done, but doesn't feel right.
Any insight, musings, requests for clarification, much appreciated.
Follow up edit:
As Jeanne suggests below, I'm creating a super class with all of my asserts & setup/teardown methods. In reality my asserts are actually helper functions which wrap around the basic junit 4 asserts, which I import into my super class. Any test of mine will just extend this super class. One caveat that I'm considering is making the super class abstract, since there shouldn't be any instance of the super class.
Marc,
I use two patterns in JUnit 4. For "utility type" assertions, I made a static class. For example ReflectionAssertions. Then I use a static import to use those assertions in my JUnit 4 test.
For local type assertions that are only used in one class, I make them regular methods in the JUnit 4 test class itself. For example assertCallingMyBusinessMethodWithNullBlowsUp(). These don't have much reuse value.
I don't consider this mixing concepts because the later group aren't reusable outside my test. If I had reusable assertions that made webservice calls (and therefore needed state), I would create a superclass that did not extend TestCase and use that. My superclass would have the state and #Before methods for setup. As such, it is part of the test.
I have been reading over design-by-contract posts and examples, and there is something that I cannot seem to wrap my head around. In all of the examples I have seen, DbC is used on a trivial class testing its own state in the post-conditions (e.g. lots of Bank Accounts).
It seems to me that most of the time when you call a method of a class, it does much more work delegating method calls to its external dependencies. I understand how to check for this in a Unit-Test with specific scenarios using dependency inversion and mock objects that focus on the external behavior of the method, but how does this work with DbC and post-conditions?
My second question has to deal with understanding complex post-conditions. It seems to me that to write out a post-condition for many functions, that you basically have to re-write the body of the function for your post-condition to know what the new state is going to be. What is the point of that?
I really do like the notion of DbC and I think that it has great promise, particularly if I can figure out how to reproduce some failure state once I find a validated contract. Over the past couple of hours I have been reading some neat stuff wrt. automatic test generation in Eiffel. I am currently trying to improve my processes in C++ development, but I am open to learning something new if I can figure out how to not lose all of the ground I have made in my current projects. Thanks.
but how does this work with DbC and
post-conditions?
Every function is basically one of these:
A sequence of statements
A conditional statement
A loop
The idea is that you should check any postconditions about the results of the function that go beyond the union of the postconditions of all the functions called.
that you basically have to re-write
the body of the function for your
post-condition to know what the new
state is going to be
Think about it the other way round. What made you write the function in the first place? What were you pursuing? Can that be expressed in a postcondition which is more simple than the function body itself? A postcondition will typically use queries (what in C++ are const functions), while the body usually combines commands and queries (methods that modify the object and methods which only get information from it).
In some cases, yes, you will find out that you can really add little value with postconditions. In these cases, writing a bunch of tests will typically be enough.
See also:
Bertrand Meyer, Contract Driven
Development
Related questions 1, 2
Delegation at the contract level
most of the time when you call a
method of a class, it does much more
work delegating method calls to its
external dependencies
As for this first question: the implementation of a function/method may call many other function/methods, but if the designer of the code had a clear mind, this does not imply that the specification of the caller is the concatenation of the specifications of the callees. For a method that calls many others, the size of the specification can remain contained if the method accomplishes a precise and well-defined task. Which it should if the whole system was well designed.
You are clearly asking your question from the point of view of run-time assertion checking. In this context, the above would perhaps be expressed as "you don't need to re-check in the post-condition of the caller that all the callees have respected their respective contracts. These checks will already be made on each call. In the post-condition of the caller, only check the functionally visible result of the caller."
Understanding complex post-conditions
You may find this "ACSL by example" document interesting (although probably different from what you're used to). It contains many examples of formal contracts for C functions. The language of the contracts is intended for static verification instead of run-time checking, with all the advantages and the drawbacks that it entails. They are a little more sophisticated than the "Bank Accounts" that you mention — these functions implement real algorithms, although simple ones. The document keeps the contracts short and readable by introducing well-thought-out auxiliary predicates (which would be called queries in Eiffel, as Daniel points out in his answer).
I want to test validation logic in a legacy class. The class uses a method to load effective dates from a config file.
I have written a subclass of the class in question and overridden the config method so I can run my unit test against the subclass with any combination of effective dates.
Is this an appropriate strategy? It strikes me as a clean technique for testing code that you don't want to mess with.
I like it, its the most simple and straight forward way to get this done. And since it is a legacy class, it will not change anymore, so you don't run danger of bumping into the fragile base class problem neither.
It seems to be an appropriate strategy to me. Ofcourse with this override you won't
be able to test the code (in the original class) that loads the config data, but if you have other tests to cover this sceario then I think the approach you outlined is fine.
I don't know if I should test my #Entity-annotated Pojos. After all, there are mainly just generated getters/setters. Should I test them?
When it comes to testing DAOs I'm using all those entities - so they are already propely tested, I guess?
Thanks for your thoughts.
Matt
Can your code contain any bugs? If not, what's the point in testing it? In fact, trying to test it would just introduce new bugs (because your tests could be wrong).
So the conclusion is: You should not test getters and setters without code (i.e. those which just assign or read a field without any additional code).
The exception is: When you manually write those getters/setters because you could have made a typo. But even then, some code will use these and there should be a test for that code which in turn tests whether the getters/setters behave correctly.
The only reason I could think of the write tests would be to test the #Entity annotation itself. Testing the storage and retrieval of values seems like one is doubting a fundamental ability of our programming environment :)