Avoid emptying set file for unmanaged sets

[mikesmitty]

(#39) This change applies the inverse of the ignore_content
variable to the replace parameter on *.set file resources in the
ipset::set class, preventing unmanaged ipsets from having their set
files emptied each puppet run

Pull Request (PR) description

Per the ipset::unmanaged section in the readme, unmanaged sets won't be kept if the set options are changed, but the set file is still emptied with each puppet run. It does leave the in-memory set in place, but this is less than ideal as in the event of an unexpected reboot the set could remain empty until the set file is next updated.

This change leaves the set file intact in addition to the in-memory ipset for unmanaged sets. This is more inline with my expectation of how an unmanaged set should behave, but I'm curious what you think.

This Pull Request (PR) fixes the following issues

#39

[truthsolo]

CI failure due to CentOS 6. PR is good!

[bastelfreak]

@mikesmitty thanks for the PR! We fixed the CI issues. Can you please rebase against our latest master branch?

[mikesmitty]

Sure, I just rebased it against master. I do intend to write tests at some point, but I'm a bit lacking in free time at the moment.

[mikesmitty]

I'm pretty new to puppet-rspec so feel free to offer any suggestions. I went ahead and added a test for verifying ignore_contents: true results in replace: false being set on the set file as well as a test to ensure ipset::unmanaged creates an ipset:set with ignore_contents: true

[bastelfreak]

thanks for the tests, I think they are sufficient. Can you take a look at the rubocop complaints?

[mikesmitty]

Sure, sorry about that. Resolved the rubocop error and made that function call more consistent style-wise.

[bastelfreak]

tanks for the fix!

[mikesmitty]

@bastelfreak thanks for merging it. Do you have an idea of when v1.2.4 will end up getting tagged?

[bastelfreak]

we just released v2.0.0: https://forge.puppet.com/modules/puppet/ipset/changelog#v200-2021-06-03