Most Unknown Elements (Part 3 :)

Compute the Jaccard similarity of character multisets. The source's example shows that repeated characters count and gives 0.4; this site adds non-empty strings and the standard input protocol.

输入格式

Two lines with non-empty strings containing only lowercase English letters, each of length 1..10^5.

输出格式

Output the multiset Jaccard similarity: Σmin(countA[c],countB[c]) / Σmax(countA[c],countB[c]), with an error of at most 1e−6.

样例 1

输入：

baa
abbc

输出：

0.400000000000

样例 2

输入：

abc
xyz

输出：

0.000000000000

样例 3

输入：

aaaa
aa

输出：

0.500000000000