apply method

@override double apply(String s1, String s2, { double scoreCutoff = 0.0, })

Returns the score of similarity computed from s1 and s2 Optionally takes a scoreCutoff which will return early if the score is below the threshold

Implementation

@override
double apply(String s1, String s2, {double scoreCutoff = 0.0}) {
  // RapidFuzz returns 0 if cutoff > 100
  if (scoreCutoff > 100) return 0.0;

  var len1 = s1.length;
  var len2 = s2.length;

  // RapidFuzz returns 0 when either string is empty
  if (len1 == 0 || len2 == 0) {
    return 0.0;
  }

  var tryPartials = tryPartialsConst;
  var unbaseScale = unbaseScaleConst;
  var partialScale = partialScaleConst;

  var base = ratio(s1, s2, scoreCutoff: scoreCutoff);
  var lenRatio = max(len1, len2) / min(len1, len2);

  tryPartials = lenRatio >= 1.5;
  if (lenRatio > 8) partialScale = 0.6;

  if (tryPartials) {
    // Update score_cutoff when using the results of one algorithm as cutoff for the next
    scoreCutoff = max(scoreCutoff, base) / partialScale;
    var partial =
        partialRatio(s1, s2, scoreCutoff: scoreCutoff) * partialScale;

    scoreCutoff = max(scoreCutoff, partial) / (unbaseScale * partialScale);
    var partialSor =
        tokenSortPartialRatio(s1, s2, scoreCutoff: scoreCutoff) *
        unbaseScale *
        partialScale;

    var partialSet =
        tokenSetPartialRatio(s1, s2, scoreCutoff: scoreCutoff) *
        unbaseScale *
        partialScale;

    return [base, partial, partialSor, partialSet].reduce(max);
  } else {
    scoreCutoff = max(scoreCutoff, base) / unbaseScale;
    var tokenSort =
        tokenSortRatio(s1, s2, scoreCutoff: scoreCutoff) *
        unbaseScale;
    var tokenSet =
        tokenSetRatio(s1, s2, scoreCutoff: scoreCutoff) *
        unbaseScale;

    return [base, tokenSort, tokenSet].reduce(max);
  }
}