CF196A.Lexicographically Maximum Subsequence

You've got string s, consisting of only lowercase English letters. Find its lexicographically maximum subsequence.

We'll call a non-empty string s[_p_1_p_2... p__k] = _s__p_1_s__p_2... s__p__k(1 ≤ _p_1 < _p_2 < ... < p__k ≤ |s|) a subsequence of string s = _s_1_s_2... s|s|.

String x = _x_1_x_2... x|x| is lexicographically larger than string y = _y_1_y_2... y|y|, if either |x| > |y| and _x_1 = _y_1, _x_2 = _y_2, ... , x|y| = y|y|, or exists such number r (r < |x|, r < |y|), that _x_1 = _y_1, _x_2 = _y_2, ... , x__r = y__r and x__r + 1 > y__r + 1. Characters in lines are compared like their ASCII codes.

你有一个仅由小写英文字母组成的字符串 $s$。请找出其字典序最大的子序列。

我们称一个非空字符串 $s[p_1p_2\ldots p_k] = s_{p_1}s_{p_2}\ldots s_{p_k}$（其中 $1 \leq p_1 < p_2 < \ldots < p_k \leq |s|$）为字符串 $s = s_1s_2\ldots s_{|s|}$ 的一个子序列。

字符串 $x = x_1x_2\ldots x_{|x|}$ 在字典序上大于字符串 $y = y_1y_2\ldots y_{|y|}$，当且仅当以下任一条件成立：

• $|x| > |y|$ 且 $x_1 = y_1,\, x_2 = y_2,\, \ldots,\, x_{|y|} = y_{|y|}$；
• 或存在某个数 $r$（满足 $r < |x|$ 且 $r < |y|$），使得 $x_1 = y_1,\, x_2 = y_2,\, \ldots,\, x_r = y_r$，且 $x_{r+1} > y_{r+1}$。
  字符串中字符的比较依据其 ASCII 码值。

The single line contains a non-empty string s, consisting only of lowercase English letters. The string's length doesn't exceed 105.

单行包含一个非空字符串 $s$，该字符串仅由小写英文字母组成。字符串的长度不超过 $10^5$。

Print the lexicographically maximum subsequence of string s.

输出字符串 $s$ 的字典序最大的子序列。

Let's look at samples and see what the sought subsequences look like (they are marked with uppercase bold letters).

The first sample: aBaBBA

The second sample: abbCbCCaCbbCBaaBA

我们来看几个样例，观察所求子序列的形态（它们用大写粗体字母标出）。

第一个样例：aBaBBA

第二个样例：abbCbCCaCbbCBaaBA