AT_abc433_f [ABC433F] 1122 Subsequence 2
Description
You are given a string $ S $ consisting of digits.
A string $ T $ is called a **1122-string** if it satisfies all of the following conditions. (The definition is the same as in Problem C.)
- $ T $ is a non-empty string consisting of digits.
- $ |T| $ is even, where $ |T| $ denotes the length of string $ T $ .
- All characters from the $ 1 $ -st through the $ \frac{|T|}2 $ -th character of $ T $ are the same digit.
- All characters from the $ (\frac{|T|}2+1) $ -th through the $ |T| $ -th character of $ T $ are the same digit.
- Adding $ 1 $ to the digit of the $ 1 $ -st character of $ T $ gives the digit of the $ |T| $ -th character.
For example, `1122`, `01`, and `444555` are 1122-strings, but `1222` and `90` are not 1122-strings.
Find the number, modulo $ 998244353 $ , of **(not necessarily contiguous) subsequences** of $ S $ that are 1122-strings.
Two subsequences are counted separately if they are extracted from different positions, even if they are identical as strings.
Input Format
The input is given from Standard Input in the following format:
> $ S $
Output Format
Output the number, modulo $ 998244353 $ , of subsequences of $ S $ that are 1122-strings.
Explanation/Hint
### Sample Explanation 1
The following five subsequences satisfy the condition.
- `12` extracted from the $ 1 $ -st and $ 3 $ -rd characters of $ S $
- `12` extracted from the $ 1 $ -st and $ 4 $ -th characters of $ S $
- `12` extracted from the $ 2 $ -nd and $ 3 $ -rd characters of $ S $
- `12` extracted from the $ 2 $ -nd and $ 4 $ -th characters of $ S $
- `1122` extracted from the $ 1 $ -st through $ 4 $ -th characters of $ S $
Thus, output $ 5 $ .
Note that two subsequences are counted separately if they are extracted from different positions, even if they are identical as strings.
### Sample Explanation 2
There may be no subsequence that is a 1122-string.
### Constraints
- $ S $ is a string consisting of digits with length between $ 1 $ and $ 10^6 $ , inclusive.