HashSet lookup

object PigLatin {
    private val vowels = hashSetOf('a', 'e', 'i', 'o', 'u')
    private val specials = hashSetOf("xr", "yt")
    private val vowels_y = hashSetOf('a', 'e', 'i', 'o', 'u', 'y')

    fun translate(phrase: String): String {
        return phrase.split(" ").joinToString(separator = " ") { piggyfy(it) }
    }

    private fun piggyfy(word: String): String {
        if (vowels.contains(word[0]) || specials.contains(word.substring(0, 2))) return word + "ay"
        for (pos in 1..word.length) {
            val letter = word[pos]
            if (vowels_y.contains(letter)) {
                val posCalc = if (letter == 'u' && word[pos - 1] == 'q') pos + 1 else pos
                return word.substring(posCalc) + word.substring(0, posCalc) + "ay"
            }
        }
        return word
    }
}

An object declaration is used to define PigLatin as essentially a singleton object instantiation of the class. This is sufficient, since there is no object state that needs to change with each call of the translate method.

Three private vals are defined for the HashSets, using the hashSetOf method.

The translate function is implemented by splitting the input String by a space, and then using the joinToString method with a space as the separator. The lambda of joinToString uses the it keyword to refer to the single String parameter for the lambda, and passes that to the piggyfy function. The translate function returns the rejoined input String as transformed by each word being processed by the piggyfy function.

The piggyfy function takes the input String and checks to see if its first character is in the HashSet of vowels (not including y) or if the first two characters are in the HashSet of "specials" (e.g. "xr" or "yt"). If the word starts with a vowel or a special two-letter combination, then the function returns the word concatenated with ay at the end.

Otherwise, a range is used to iterate an index for the characters of the word in a for loop, starting with the second character and up to but not including the length of the word.

A variable is set from the character in the word at the current index of the range.

If the character is in the HashSet of vowels (including y), then the index is adjusted if the current character is a u and the previous character is a q. A substring is then taken from the character at the index until the end of the word. Concatenated to the end of that is a substring from the beginning of the word up to but not including the index. Finally, concatenated to the end of that is ay. The concatenated String is returned from the function.

In the event that the word has no vowels and the for loop finishes without returning, the function returns the word as is.