You are given a row of n tokens in a row colored red and blue.

In one move, you can choose to perform a capture. A capture chooses a token, and makes a jump over exactly one other token, and removes a token of the opposite color.

More specifically, given three adjacent tokens we can convert it into the following state with two adjacent tokens:

• \(RxB \to xR\)
• \(BxR \to xB\)
• \(RxB \to Bx\)
• \(BxR \to Rx\)

where x is a token of an arbitrary color.

Given the initial row of tokens, return the minimum possible length of the resulting row after a series of captures.

Input Format:

The first line will contain the number of test cases T.
Each test case can be described with one line.
This line will contain a string consisting of only characters 'R' and 'B' denoting the colors in the row.

Output Format:

Output T numbers, the answers to each problem.

Constraints

There is no partial credit for this problem.

There is only one file for this problem. The file has the following constraints.
\(T = 1000\)
\(1 ≤ n ≤ 50\)