A foundry runs a line of n metal stamps past a die press. Each stamp already carries a positive whole-number rating painted on it, but the press's imprinting head is worn and simply refuses to strike the digit 0 -- every other digit still gets imprinted normally, in its original left-to-right order, onto the stamp's fresh face. Reading the stamps in line order and writing down each stamp's imprinted digits one after another (each stamp's own digits stay in their original left-to-right order), you get one long number for the whole batch: the batch serial. Because no stamp's rating is ever written with a leading zero, its very first digit always survives the press, so the batch serial is guaranteed to be non-empty and to never start with a zero.

The foreman also wants a quality-control figure. Before any digits were dropped, add up the ORIGINAL ratings of every stamp on the line to get the line weight. The plant's final batch code is the batch serial, read as an ordinary integer, multiplied by the line weight.

Given the stamps' ratings in line order, compute the batch code.

Input format

• Line 1: a single integer n, the number of stamps on the line.
• Line 2: n space-separated positive integers a_1, a_2, ..., a_n, the rating painted on each stamp, in line order.

Output format

• A single line containing one integer: the batch code. It can be far larger than fits in a 64-bit integer, so print it in full and exactly.

Constraints

• 1 <= n <= 100
• 1 <= a_i <= 1000 for every i

3
12 34 5

629595

2
100 205

38125