动态规划--最长公共子串

公共子串问题

在计算机科学中，最长公共子串问题是寻找两个或多个已知字符串最长的子串。此问题与最长公共子序列问题的区别在于子序列不必是连续的，而子串却必须是。链接: 百度百科

费曼算法

The Feynman Algorithm:

1. Write down the problem.
2. Think real hard.
3. Write down the solution.

动态规划算法思路

1. 给定两个字符串T1和T2, 假设dp(i, j)表示T1，T2在位置i, j的公共子串的长度【向位置0的方向，连续相同的字符的数量】
   比如“abcdefg”, “absecde”
   dp(0,0) = 1 ‘a’ == ‘a’
   dp(1,1) = 2 ‘ab’ == ‘ab’
   dp(1,2) = 0 ‘b’ != ‘s’

2. 总结规律，最长公共子串：
   if T1[i] == T2[j] :
   dp(i,j) = dp(i-1,j-1) + 1
   else:
   dp(i,j) = 0

---

最长公共子序列：
if T1[i] == T2[j]:
dp[i,j] = dp(i-1,j-1) +1
else:
dp(i,j) = max(dp(i-1,j), dp(i,j-1))

代码实现

def getLcs(t1:str, t2:str)->str:
    '''
    get the longest common substring of two given string
    '''
    if not t1:
        return t1
    if not t2:
        return t2

    # dp = [[0 for i in range(len(t2)+1)] for j in range(len(t1)+1)]

    max_len = 0
    dp = [[0 for i in range(len(t2)+1)] for j in range(2)]
    for i in range(1, len(t1)+1):
        ind = [0,1][i%2==0]
        for j in range(1, len(t2)+1):
            if t1[i-1]==t2[j-1]:
                dp[ind][j] = dp[1-ind][j-1] + 1
            else:
                dp[ind][j] = 0
            
            if dp[ind][j] >max_len:
                max_len = dp[ind][j]
                p = i

    print("max_len:%d , pos: %d"%(max_len,p))
    return t1[p-max_len:p]

def getLcss(t1:str, t2:str)->str:
    '''
    get the longest common subsequence
    '''
    if not t1:
        return t1
    if not t2:
        return t2
    
    dp = [[0 for i in range(len(t2)+1)] for j in range(2)]
    max_len = 0
    res = ""
    for i in range(1,len(t1)+1):
        ind = [0,1][i%2]
        for j in range(1, len(t2)+1):
            if t1[i-1] == t2[j-1]:
                dp[ind][j] = dp[1-ind][j-1] + 1
                
            else:
                dp[ind][j] = max(dp[ind][j-1], dp[1-ind][j])

            if dp[ind][j] > max_len:
                max_len = dp[ind][j]
                res += t1[i-1]
    # max_len = max(dp[0][-1], dp[1][-1])
    print("max len: %d"%max_len)
    return res