Remove unnecessary early return in RSpec/ContextWordingcop

[ydah]

Checks if allowed_patterns is empty, but looking at the implementation of matches_allowed_pattern?, it refers to it as allowed_patterns.any?, so return false if allowed_patterns .empty? is a completely unnecessary early return.

https://github.com/rubocop/rubocop/blob/e7ccb0200d013d1c4b75b3face0dea0538620769/lib/rubocop/cop/mixin/allowed_pattern.rb#L29-L31

def matches_allowed_pattern?(line)
  allowed_patterns.any? { |pattern| Regexp.new(pattern).match?(line) }
end

also resolve: #1972

---

Before submitting the PR make sure the following are checked:

• Feature branch is up-to-date with master (if not - rebase it).
• Squashed related commits together.
• [-] Added tests.
• [-] Updated documentation.
• [-] Added an entry to the CHANGELOG.md if the new code introduces user-observable changes.
• The build (bundle exec rake) passes (be sure to run this locally, since it may produce updated documentation that you will need to commit).

[pirj]

I’m quite concerned if we had a method, and it was not fully covered with specs, and now we’re refactoring it to a node pattern where coverage is not calculated, can we say that it does the same thing with certainty?
I live node patterns, they are often more expressive due to their declarative nature.
But does this one read better than the method?

Node patterns support inline comments, but we rarely use that, and it’s not always useful.
Node pattern has this @!method YARD directive, but we often miss the opportunity to add examples of what it does match and what not.

With all this, I think in this case a node pattern is inferior to a method.
Especially since this change conceals a gap in our spec coverage.

[ydah]

@pirj It is certainly not a good choice to replace it with a node pattern for which coverage is not calculated.
So I have updated this PR concern as removing unnecessary early returns. What do you think?

[pirj]

Nice refactoring!
Great to see it’s getting us to 100% coverage ☺️

[pirj]

Another pair of eyes in case I’m missing something this early morning, @rubocop/rubocop-rspec ?

[bquorning]

Isn’t the reason we didn’t have complete spec coverage before that we never hit a case where allowed_patterns.empty?? Should we add a context 'when both Prefixes and AllowedPatterns are empty' to the specs before refactoring?

[corsonknowles]

@bquorning That spec is available here:

• Complete branch coverage for ContextWording #1972

[ydah]

Close this PR because it was found to operate differently from the current specification.