Spark levenshtein left join on string, match one data then break

Question

val resultDF = todoDF.join(sourceDF,
      functions.levenshtein(sourceDF.col("string_right"),todoDF.col("string_left")) <= 3
      &&