That makes the test more dependent on implementation details: you need
to calculate exactly how many times you expect the method to be called
in the setup stage and then add one (or whatever) to it for the times
you expect it to be called at the stage you're actually testing. It
also makes the intention of the test less clear: somebody reading the
test wouldn't immediately know why you're expecting the method to be
called n times, rather than once or never.
On Sep 27, 5:08 pm, Daniel Cazzulino <
k...@clariusconsulting.net>
wrote: