Let's say I have two classes which implement the same basic API, and I want to test that they are "stochastically equivalent"1, over at least over a subset of their methods.
E.g., I write my own "list" class foo:list
and rather than painstakingly writing a bunch of unit tests for it, I want to compare it to std::list
as a reference. That is, any sequence of operations on foo::list
should produce the same results as the same sequence of std::list
.
I'm OK listing the names of the operations, but hopefully not much more boilerplate than that. A generic solution that can be applied to other pairs of "behaviorally equivalent" classes is ideal.
1 By "stochastically equivalent" I mean that no differences are observed over many series of operations, which obviously falls short of a complete proof of equivalence.
Aucun commentaire:
Enregistrer un commentaire