Given a string s
of zeros and ones, return the maximum score after splitting the string into two non-empty substrings (i.e. left substring and right substring).
The score after splitting a string is the number of zeros in the left substring plus the number of ones in the right substring.
Example 1:
Input: s = "011101" Output: 5 Explanation: All possible ways of splitting s into two non-empty substrings are: left = "0" and right = "11101", score = 1 + 4 = 5 left = "01" and right = "1101", score = 1 + 3 = 4 left = "011" and right = "101", score = 1 + 2 = 3 left = "0111" and right = "01", score = 1 + 1 = 2 left = "01110" and right = "1", score = 2 + 1 = 3
Example 2:
Input: s = "00111" Output: 5 Explanation: When left = "00" and right = "111", we get the maximum score = 2 + 3 = 5
Example 3:
Input: s = "1111" Output: 3
Constraints:
2 <= s.length <= 500
- The string
s
consists of characters '0' and '1' only.
先不说暴力算法,提示里有说,先计算一遍以当前位置 i 为索引的左边的1的个数,那扫描到最后,所有1的个数就知道了,将每个位置的左边1的个数保存在prefix数组中;
然后再次扫描,累计0的个数,根据先前的prefix,用1的总数减去prefix[i]就得到以当前位置为分割,右边1的个数,然后与当前位置0的个数相加就是当前分割的score.
class Solution {
public:
int maxScore(string s) {
vector<int> prefix(s.size(), 0);
int pre = 0;
for(int i = 0; i < s.size(); i++)
{
if(s[i] == '1')
{
pre += 1;
}
prefix[i] = pre; //注意要更新0的位置对应的pre值
}
int zeroC = 0;
int res = INT_MIN;
for(int i = 0; i < s.size() - 1; i++)//这里一定不能扫描到最后一个,因为就没法分成左右两边了
{
if(s[i] == '0')
{
zeroC++;
}
res = max(res, zeroC + (pre - prefix[i]));
}
return res;
}
};
注意了上面2点,就OK了,小细节啊,真得注意。
Runtime: 0 ms, faster than 100.00% of C++ online submissions for Maximum Score After Splitting a String.
Memory Usage: 6.7 MB, less than 100.00% of C++ online submissions for Maximum Score After Splitting a String.