Your approach seems fine to me.
Generally speaking, I do not think that there is something like a "standard way" to selectively execute unit tests based on iOS version.
On the one hand, your approach is the same that can be used to implement one feature selectively or in different ways according to iOS version. Possibly you already know about this discussion on CocoaWithLove: Tips and Tricks from conditional.... It's pretty old, but the approaches described here to hold still.
You do not specify the way your unit tests are executed, but the real cool way to handle this, IMO, would be to be able to specify which tests to execute outside of the implementation, so that you state which ones are for iOS7 only and do not pollute your test implementation.
This could done, e.g., through a version number associated to each test; before calling the unit test implementation you check the version number in the function calling, e.g., testSomethingThatOnlyWorksOniOS7
.
EDIT:
Actually, things could be easier that what I thought in the first place. Now I am going a bit hacky...
You could modify the place in OCUnit where the actual test method call is done (no idea about this, sorry, but I do not think it should be hard to find out...) so that it checks on the selector name: you convert the selector name into a string (NSStringFromSelector
) and if it matches some regular expression, you take some specific branch (which would simply be in your case, ignoring that test).
If you are worried by modifying OCUnit, which might not be sensible, what I said above could be done through method swizzling: only if the selector name does not match your expression, you call the original implementation, otherwise do nothing.