I've downloaded sources and saw"compactDistance" is default method used in in Levenshtein comparator.
"distance" method described as original, naive implementation, using the Wagner & Fischer algorithm from 1974.
Comparing two strings "emma" and "ema".
When using default "compactDistance" in compare method it returns 0 (dist=0) and I receive result 1 from comparator. So it supposes those strings are exactly the same but they are not.
When using "distance" method it returns 1 and I receive result 0.6666666666666667 from comparator. I think it's good.
So does the "compactDistance" method work incorrectly?