DX: check deprecations exactly #7610
Conversation
| $this->expectDeprecation($deprecation); | ||
| $this->expectDeprecation($deprecation); | ||
| $this->expectDeprecation($deprecation); |
There was a problem hiding this comment.
Does it make sense to expect exactly the same deprecation message multiple times? I would rather want to ensure that every expected deprecation message happens at least once. What is the gain from checking the exact amount of messages?
There was a problem hiding this comment.
👍
we care to detect that there is deprecation. There is no value in having same deprecation msg once or 10 times.
There was a problem hiding this comment.
Does it make sense to expect exactly the same deprecation message multiple times?
Of course not, I just wanted to show that if we would want to check the exact count (as suggested in #7578 (comment)) there would be multiple expectations of the same message.
There was a problem hiding this comment.
that was totally not my point @kubawerlos .
my point was what if yet another deprecation would happen. let say, the "unexpected" one - we should fail. now we silently move on.
| foreach ($this->expectedDeprecations as $expectedDeprecation) { | ||
| self::assertContains($expectedDeprecation, $this->actualDeprecations); | ||
| } | ||
| self::assertSame($this->expectedDeprecations, $this->actualDeprecations); |
There was a problem hiding this comment.
this way you depends on order of triggered deprecation. if someone done DX refactoring of method, method works exactly same way, but deprecations will get triggered in other order - this assertion will fail.
There was a problem hiding this comment.
Is that bad? SymfonyTestsListener needs expectDeprecation called in order they are triggered.
There was a problem hiding this comment.
Symfony's implementation aside - why order of deprecations should be interesting? In my opinion you either expect some deprecation or not, it does not matter if the executed code triggers it at the beginning or in the end.
There was a problem hiding this comment.
I have no problem with not caring about the order (which is how it is after the latest commit), and at the same time, I can imagine someone wanting to check the order in what deprecations were triggered (like maybe from least detailed to most detailed).
|
recovered #7742 |
Ref. #7578 (comment)