You are given a string $*S*$ of length $*n*$ with each character being one of the first $*m*$ lowercase English letters.

Calculate how many different strings $*T*$ of length $*n*$ composed from the first $*m*$ lowercase English letters exist such that the length of LCS (longest common subsequence) between $*S*$ and $*T*$ is $*n* - 1$.

Recall that LCS of two strings $*S*$ and $*T*$ is the longest string $*C*$ such that $*C*$ both in $*S*$ and $*T*$ as a subsequence.

Input

The first line contains two numbers $*n*$ and $*m*$ denoting the length of string $*S*$ and number of first English lowercase characters forming the character set for strings ($1 ≤ *n* ≤ 100 000$, $2 ≤ *m* ≤ 26$).

The second line contains string $*S*$.

Output

Print the only line containing the answer.

Note

For the first sample, the $6$ possible strings $*T*$ are: aab, aac, aba, aca, baa, caa.

For the second sample, the $11$ possible strings $*T*$ are: aaa, aac, aba, abb, abc, aca, acb, baa, bab, caa, cab.

For the third sample, the only possible string $*T*$ is b.