CallbackCache: simplify calling function/method

[marc-mabe]

Q	A
Documentation	no
Bugfix	no
BC Break	no
New Feature	no
RFC	no
QA	yes

Description

Simplified calling the underlying function/method of CallbackCache (and ObjectCache and ClassCache)

[boesing]

LGTM so far.
Thanks for contributing this!
I'll merge this when I've implemented GHA for CI.
For more details, see laminas/technical-steering-committee#61

[boesing]

Hm, I was wondering if I could backport this to v2 but actually, this is a BC break.
I've implemented GHA already and it turns out that changing the callback to [$className, $method] changes its type from string to array and thus, leads to an error when the callback key is generated.

https://github.com/laminas/laminas-cache/pull/63/files#diff-affcf97709fa42da6b51246d0e1d19dfbd80479d3cfbf2c35eecae7358107e6dR63

laminas-cache/src/Pattern/ClassCache.php

Line 99 in 15e0b44

$callbackKey = md5(strtolower($callback));

I will rebase this PR against 3.0.x as it contains GHA so we can see the failing tests in here.

[boesing]

Wait what? I've tried to rebase against 2.12.x but that led into failing tests.
Why the heck isn't that the case in 3.0.x - I am confused...

---

Okay, somehow the names of the test classes changed due to some of the changes made back in february. I've fixed that with #103 and rebased against 3.0.x now.

[marc-mabe]

@boesing should be fine now

[boesing]

Thanks, Marc.
Looks good. As far as I can see, this wont be a BC break, right?
I'll probably retarget this against 2.12.x if its not considered BC break so we get this out of the door somewhen in the next weeks rather than months.
WDYT?

[marc-mabe]

I don't see any bc break here. It's just an internal optimisation.

[boesing]

GHA have major outage as of https://www.githubstatus.com/incidents/zbpwygxwb3gw

Verified functionality locally on all three PHP versions.
Thanks, @marc-mabe!