Fix false positives for ignored mismatches.

[calavera]

Result#matched? returns true when it was actually an ignored mismatch.

In the publishing example in the README, for instance:

class MyExperiment
  def publish(result)
    # Store the timing for the control value,
    $statsd.timing "science.#{name}.control", result.control.duration
    # for the candidate (only the first, see "Breaking the rules" below,
    $statsd.timing "science.#{name}.candidate", result.candidates.first.duration

    # and counts for match/ignore/mismatch:
    if result.matched?
      $statsd.increment "science.#{name}.matched"
    elsif result.ignored?
      $statsd.increment "science.#{name}.ignored"
    else
      $statsd.increment "science.#{name}.mismatched"
      # Finally, store mismatches in redis so they can be retrieved and examined
      # later on, for debugging and research.
      store_mismatch_data(result)
    end
  end

Given this experiment:

science "test" do |e|
  e.ignore { true }
  e.use { true }
  e.try { false }
end

The publish method will increment the counter for science.test.matched rather than incrementing science.test.ignored.

[calavera]

Btw, I'm aware I can fix this in my code by switching the order in which I check Result#matched? and Result#ignored?, but I still think it's unexpected to return true when the observations didn't really match.

[calavera]

@jesseplusplus, @jbarnette I'd really appreciate some comments about this behavior.

[jesseplusplus]

Definitely a bug.

👍 nice catch and fix

[zerowidth]

I vaguely recall doing this on purpose, but this makes more sense. 👍 and thanks for merging @jesseplusplus