Restructure MPM disabling

[ekohl]

This takes the approach of listing the disabling steps once rather than repeating it for every mpm.

I wasn't quite sure on the prefork disabling. Does prefork need to stay enabled for itk?

[david22swan]

Not sure about this change, while it cuts down on repetition it may also make the code harder for new users to read through. In addition, look's like it's causing a mass failure in the spec tests.

[ekohl]

Not sure about this change, while it cuts down on repetition it may also make the code harder for new users to read through

Funny, I had great problems with the duplication and think this is much easier to follow. My logic is:

If MPM is used
  configure it
else
  remove it
end

If there is no configuring it ends up being

if not MPM is used
  remove it
end

And the mass failures can easily be explained by my syntax error. Due to https://tickets.puppetlabs.com/browse/MODULES-11161 I can't run PDK module tests locally so I have to rely on CI. Sadly it makes contributing to Puppetlabs modules really hard for me. Normally I ignore any PDK module but puppetlabs modules are hard to ignore.

Edit: Because of 400c75d it meant it ran the whole test suite instead of verifying it.

[david22swan]

Looking back at this, and after confusing myself and then realising mistake. This looks like a good change.
Thank's for putting in the work :)