poj 字符串相关之1816 Wild Words

poj 字符串相关之1816 Wild Words

一看到字符串匹配的问题，就想到了之前做的动态规划的问题，转移状态公式为：

if (pattern[j][x] == '?' || pattern[j][x] == words[i][y])
                    dp[x][y] = dp[x - 1][y - 1];
                if (pattern[j][x] == '*')
                {
                    for (k = y; k >= 0; k--)
                    if (dp[x - 1][k])
                    {
                        dp[x][y] = 1;
                        break;
                    }
                }

于是得到了以下代码，每一个pattern和word之间匹配用dp

#include<cstdio>
#include<string>
#include<iostream>
#include<algorithm>
#include<memory.h>
#include<map>
#define PNUM 100005
#define WNUM 105
using namespace std;
char pattern[PNUM][7], words[WNUM][21];
bool dp[7][21];
int N, M;
int main()
{
    int i, j, x, y, k, t, count;
    scanf("%d%d", &N, &M);
    for (i = 0; i < N; i++)
        scanf("%s", pattern[i] + 1);
    for (i = 0; i < M; i++)
        scanf("%s", words[i] + 1);
    for (i = 0; i < M; i++)
    {
        count = 0;
        for (j = 0; j < N; j++)
        {
            memset(dp, 0, sizeof(dp));
            dp[0][0] = 1;
            for (x = 1; x <= strlen(pattern[j] + 1); x++)
            for (y = 0; y <= strlen(words[i] + 1); y++)
            {
                if (pattern[j][x] == '?' || pattern[j][x] == words[i][y])
                    dp[x][y] = dp[x - 1][y - 1];
                if (pattern[j][x] == '*')
                {
                    for (k = y; k >= 0; k--)
                    if (dp[x - 1][k])
                    {
                        dp[x][y] = 1;
                        break;
                    }
                }
            }
            if (dp[strlen(pattern[j] + 1)][strlen(words[i] + 1)])
            {
                printf("%d ", j);
                count++;
            }
        }
        if (count > 0)
            printf("\n");
        else
            printf("Not match\n");
    }
}

不幸的是超时了，因为数据量很大，于是想到了通配符还可以做的是Trie树。代码如下

#include<cstdio>
#include<string>
#include<iostream>
#include<algorithm>
#include<memory.h>
#include<vector>
#include<map>
#define PNUM 100005
using namespace std;
char words[25], pattern[10];
bool result[PNUM];
int N, M;
class node
{
public:
    vector<int> index;
    node* next[28];
    node()
    {
        for (int i = 0; i < 28; i++)
            next[i] = NULL;
    }
};
class Trie
{
public:
    node* root;
    Trie()
    {
        root = new node();
    }
    void insert(int id, char * s)
    {
        node* curr = root;
        int i, j, len;
        len = strlen(s);
        for (i = 0; i < len; i++)
        {
            if (s[i] == '?')//26
            {
                if (curr->next[26] == NULL)
                    curr->next[26] = new node();
                j = 26;
            }
            else if (s[i] == '*')
            {
                if (curr->next[27] == NULL)
                    curr->next[27] = new node();
                j = 27;
            }
            else
            {
                j = s[i] - 'a';
                if (curr->next[j] == NULL)
                    curr->next[j] = new node();
            }
            curr = curr->next[j];
        }
        curr->index.push_back(id);
    }
    void find(char * s, node* root, int i)
    {
        if (i == strlen(s))
        {
            node* curr = root;
            while (curr != NULL)
            {
                if (curr->index.size())
                for (int t = 0; t < curr->index.size(); t++)
                    result[curr->index[t]] = 1;
                curr = curr->next[27];
            }
            return;
        }
        if (root->next[s[i] - 'a'] != NULL)
            find(s, root->next[s[i] - 'a'], i + 1);
        if (root->next[26] != NULL)
            find(s, root->next[26], i + 1);
        int t;
        for (t = i; t <= strlen(s); t++)
        if (root->next[27] != NULL)
            find(s, root->next[27], t);
    }
};
int main()
{
    //freopen("1.txt", "r", stdin);
    Trie T;
    int i, j, x, y, k, t, count;
    scanf("%d%d", &N, &M);
    getchar();
    for (i = 0; i < N; i++)
    {
        gets(pattern);
        T.insert(i, pattern);
    }
    for (i = 0; i < M; i++)
    {
        count = 0;
        memset(result, 0, sizeof(result));
        gets(words);
        T.find(words, T.root, 0);
        for (j = 0; j < N; j++)
        if (result[j])
        {
            if (count)
                printf(" ");
            printf("%d", j);
            count++;
        }
        if (count == 0)
            printf("Not match");
        puts("");
    }
}

注意一下就是pattern可能有重复的，所以记录pattern信息的时候用一下vector
2756K 375M

总结了一下!!!!字符串相关，数据量比较大的时候首先应该考虑字典树。