B"Ela loves reading a lot, just like her new co-workers in DTL! On her first day after becoming an engineer in DTL, she is challenged by a co-worker to sort a heap of books into different compartments on the shelf. n books must be split into k compartments on the bookshelf ( n is divisible by k ). Each book is represented by a lowercase Latin letter from 'a' to 'y' inclusively, which is the beginning letter in the title of the book. Ela must stack exactly frac{n}{k} books in each compartment. After the books are stacked, for each compartment indexed from 1 to k , she takes the minimum excluded (MEX) letter of the multiset of letters formed by letters representing all books in that compartment, then combines the resulting letters into a string. The first letter of the resulting string is the MEX letter of the multiset of letters formed by the first compartment, the second letter of the resulting string is the MEX letter of the multiset of letters formed by the second compartment, ... and so on. Please note, under the constraint of this problem, MEX letter can always be determined for any multiset found in this problem because 'z' is not used. What is the lexicographically greatest resulting string possible that Ela can create? A string a is lexicographically greater than a string b if and only if one of the following holds: The minimum excluded (MEX) letter of a multiset of letters is the letter that appears earliest in the alphabet and is not contained in the multiset. For example, if a multiset of letters contains 7 letters 'b', 'a', 'b', 'c', 'e', 'c', 'f' respectively, then the MEX letter of this compartment is 'd', because 'd' is not included in the multiset, and all letters comes before 'd' in the alphabet, namely 'a', 'b' and 'c', are included in the multiset. Each test contains multiple test cases. The first line contains the number of test cases t ( 1 <= t <= 100 ). Description of "...