Protected StaticCountHow many cases came from each label source.
StaticNormalizeA label as every golden-set comparison (scoring and labeler agreement) sees it: trimmed, lower-cased; null is ''.
StaticScoreScores a golden set. Outputs are those named in the cases' Expected or ExpectAbstain; a failed record is wrong.
Protected StaticScoredThe cases that score an output (it has Expected labels or expects an abstention), each with its verdict.
Protected StaticScoreStaticWilson95% Wilson score interval for k successes out of n ({0, 0} when n = 0).
Golden-set scoring; pure and deterministic.