Release 3.0.0

[bastelfreak]

Waiting a bit for @ekohl in case he want's other breaking changes.
CC: @jordanbreen28

[codecov]

Codecov Report

Attention: Patch coverage is 0% with 1 lines in your changes are missing coverage. Please review.

Project coverage is 95.34%. Comparing base (f20625d) to head (c61c713).

Files	Patch %	Lines
lib/rspec-puppet-facts/version.rb	0.00%	1 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #166   +/-   ##
=======================================
  Coverage   95.34%   95.34%           
=======================================
  Files           2        2           
  Lines         172      172           
=======================================
  Hits          164      164           
  Misses          8        8           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[ekohl]

I'd like to flip the default for facterdb_string_keys, but perhaps it's too soon for 3.0.0. Should we issue a deprecation warning in 3.0.0?

An alternative: finish #132 first and provide a proper API to override facts. So you can write:

require 'spec_helper'

describe_on_supported_os 'myclass' do
  it { is_expected.to compile.with_all_deps }

  context 'with selinux' do
    override_facts(selinux: true)
    # or
    override_facts_merged(os: { selinux: { enabled: true }})

    it { is_expected.to compile.with_all_deps }
  end
end

Then it just does the right thing for you with less boilerplate. Only then issue the deprecation.

Though you would still need to deal with the selection on certain OS facts even with the improved DSL.

[bastelfreak]

I also agree that it's too early to flip to strings. Releasing the current changes fixes some bugs, but if the rollout requires the change to strings it's just a too big change.

A deprecation warning is a good idea but I've no idea how/where to properly implement this.

[bastelfreak]

I added a deprecation notice to the CHANGELOG.md