Consider two strings X = x1x2 ... xm and Y = y1y2 ... yn over an alphabet set sigma = {A, G,C, T}. Denote sigma* = sigma + {-}, where "-" (dash) is the symbol that represents a space (or blank) in strings. A string alignment is to align X and Y and form two strings X*, Y* over the alphabet sigma* such that: 1. the two strings X*, Y* have the same lengths, and 2. ignoring dashes, the string X* is the same as the string X, and the string Y* is the same as the string Y. As an example, an alignment of two strings 'GATCCGA' and 'GAAAGCAGA' is as follows: G-A--TCCGA GAAAG-CAGA. There are three gaps in the above alignment; here a gap is defined as a string of consecutive dashes. Now, let us consider the following alignment: GA---TCCGA GAAAG-CAGA. Here are two gaps within this alignment. The rule of measuring the intermittent gap punishment alignment score (abbreviated by GPS) is as follows: If xi is aligned with yj, the score is o If a (consecutive) subsequence of xi's (or yj's) is aligned with a gap of length k, the score is defined as -(4 + k). That is, in the first alignment example given above, its GPS is 2 - (4 + 1) + 2 - (4 + 2) - (4 + 1) + 2 - 1 + 2 + 2 = -7. For the second alignment, its GPS is 2 + 2 - (4 + 3) - (4 + 1) + 2 - 1 + 2 + 2 = -3. Given two strings, the problem we would like to solve is to find an alignment such that its GPS is maximized. Thus, in our example, the best alignment is GA--TCCGA GAAAGCAGA. Its GPS is 2+2-(4+2)-1+2-1+2+2=2. In our problem, m and n are at most 500. Furthermore, it is required that no space in one sequence is aligned with a space in another.


The input format is as follows: 1. The first line contains an integer n of sequence pairs; the number n is at most 50. 2. The 2nd line is the sequence X of the first pair. 3. The 3rd line is the other sequence Y of the first pair. ... 2i. The (2i)-th line is the sequence X of the i-th pair. 2i+1. The (2i + 1)-th line is the other sequence Y of the i-th pair. ... 2n. The (2n)-th line is the sequence X of the n-th pair. 2n+1. The (2n + 1)-th line is the other sequence Y of the n-th pair.


For each pair of sequences, output the maximum GPS in one line.

Sample Input


Sample Output

18 20 2


Asia 2003, Kaohsiung (Taiwan China)