Alphabet

Johnny is a little boy who is learning the alphabet. For his birthday his father gave him a large collection of tokens, each token marked with a letter from A to Z, and Johnny invented a game to help himself learn.

First, Johnny takes several tokens and arranges them in a circle. He then chooses one token to start from and picks a number $k$. On each turn he finds the winning token by starting at the current starting token and moving $k$ tokens forward along the circle (the starting token itself is not counted). He then inserts a new token immediately after the winning token. The new token carries the letter that follows the winning token's letter in the alphabet: B follows A, C follows B, and so on; if the winning token is Z, then A is inserted. There are always enough spare tokens to supply the required letter.

After inserting the token, Johnny begins the next turn from the newly inserted token. The number $k$ is chosen only once and never changes between turns. The figure below shows the first four turns of a sample game whose initial tokens are J, O, H, N, N, Y (J is the starting token) with $k = 3$.

Georgie is Johnny's elder brother. Being a schoolboy already, he plays the same game in his head and amazes Johnny by predicting the letter inserted on any turn. As Johnny makes more and more turns it becomes harder to keep up, so Georgie decides to write a program that finds the answer quickly.

The first line contains three integers $n$, $k$, and $m$: $n$ is the number of tokens initially placed in the circle (1 ≤ n ≤ 10000), $k$ is the number of tokens counted on each turn (1 ≤ k ≤ 10000), and $m$ is the number of turns (1 ≤ m ≤ 10⁹).

The second line contains a string of $n$ uppercase letters (A to Z) — the tokens initially placed in the circle, listed starting from the initial token.

Output a single uppercase letter: the letter written on the token inserted on the $m$-th turn.