【DP】【LCS】最长公共子序列及其应用

  LCS 是很经典的 DP 问题，很多公司面试都会考，也有很多问题可以用 LCS 解决：

1. LeetCode - 1143. Longest Common Subsequence 最长公共子序列（就是 LCS）
2. LeetCode - 583. Delete Operation for Two Strings 将两个字符串删除为相同字符串最少删除字符数
3. LeetCode - 516. Longest Palindromic Subsequence 最长回文子序列（LPS）
4. LeetCode - 1092. Shortest Common Supersequence 最短公共超序列

一、Longest Common Subsequence（LCS）

Given two strings text1 and text2, return the length of their longest common subsequence.
A subsequence of a string is a new string generated from the original string with some characters(can be none) deleted without changing the relative order of the remaining characters. (eg, “ace” is a subsequence of “abcde” while “aec” is not).
A common subsequence of two strings is a subsequence that is common to both strings.
If there is no common subsequence, return 0.

Example 1:
Input: text1 = “abcde”, text2 = “ace”
Output: 3    Explanation: The longest common subsequence is “ace” and its length is 3.

Example 2:
Input: text1 = “abc”, text2 = “abc”
Output: 3    Explanation: The longest common subsequence is “abc” and its length is 3.

Example 3:
Input: text1 = “abc”, text2 = “def”
Output: 0    Explanation: There is no such common subsequence, so the result is 0.

  LCS 重点就是想明白递推公式：

• if text1[i] == text2[j]，那么 lcs[i + 1][j + 1] = lcs[i][j] + 1
• else，lcs[i + 1][j + 1] = max(lcs[i][j + 1], lcs[i + 1][j])

  这种二维矩阵 DP 问题，一般 DP 矩阵都会比数据多一个大小的位置，这样方便初始化，如下图，将所有初始化为0：

  AC 代码如下：

int longestCommonSubsequence(string text1, string text2) {
    const int len1 = text1.length(), len2 = text2.length();
    vector<vector<int>> lcs(len1 + 1, vector<int>(len2 + 1, 0));
    for(int i = 0; i < len1; ++i)
        for(int j = 0; j < len2; ++j) {
            if(text1[i] == text2[j])
                lcs[i + 1][j + 1] = lcs[i][j] + 1;
            else
                lcs[i + 1][j + 1] = max(lcs[i][j + 1], lcs[i + 1][j]);
        }
    return lcs[len1][len2];
}

二、Delete Operation for Two Strings

  像 LeetCode - 583. Delete Operation for Two Strings 这种题，用 LCS 很容易，算出两个字符串的最长公共子序列之后，用两个字符串的长度分别减这个公共子序列的长度，就是了：

int minDistance(string word1, string word2) {
	return word1.length() + word2.length() - 2 * longestCommonSubsequence(word1, word2);
}

三、Longest Palindromic Subsequence（LPS）

  像 LeetCode - 516. Longest Palindromic Subsequence，最长回文子序列的题也可以用 LCS 解决，就是将输入的字符串 s，翻转成 s’然后对 s 和 s’进行 LCS 的计算，当然 LPS 还有自己的 DP 方法，具体可以参考 https://blog.csdn.net/Bob__yuan/article/details/99693725。

int longestPalindromeSubseq(string s) {
    string s_rev = s;
    reverse(s_rev.begin(), s_rev.end());
    return longestCommonSubsequence(s, s_rev);
}

四、Shortest Common Supersequence

  最短公共超序列，就是就是上边第二题的延伸，找出一个字符串 s，通过删除任意位置字符，可以得到 s1 和 s2 两个字符串，找出最短的 s。其实就是找出 lcs 的路径，不止是长度。所以 dp 的二维矩阵中存储的就不是当前 lcs 的长度了，而是当前的 lcs 字符串。然后用最终的 lcs 字符串，用过一个字符一个字符的比较，将需要添加的字符添加到的结果中即可。

string LCS(const string& s1, const string& s2) {
    const int len1 = s1.length(), len2 = s2.length();
    vector<vector<string>> lcs(len1 + 1, vector<string>(len2 + 1, ""));
    for(int i = 0; i < len1; ++i)
        for(int j = 0; j < len2; ++j) {
            if(s1[i] == s2[j])
                lcs[i + 1][j + 1] = lcs[i][j] + s1[i];
            else
                lcs[i+ 1][j + 1] = lcs[i + 1][j].length() > lcs[i][j + 1].length() ? lcs[i + 1][j] : lcs[i][j + 1];
        }
    return lcs[len1][len2];
}

string shortestCommonSupersequence(string str1, string str2) {
    string lcs = LCS(str1, str2), res;
    int i = 0, j = 0;
    for(const char& c : lcs) {
        while(str1[i] != c)
            res += str1[i++];
        while(str2[j] != c)
            res += str2[j++];
        res += c;
        ++i, ++j;
    }
    return res + str1.substr(i) + str2.substr(j);
}