字符串的匹配 KMP、Horspool、BM和Sunday

KMP

出现次数

#include<cstdio>
#include<iostream>
#include<cstring>
using namespace std;
char a[10005],b[1000000];
void getnext(char *t,int *next)
{
    int i=0,j=-1,len=strlen(t);
    next[0]=-1;
    while(i<len)
    {
        if(j==-1 || t[i]==t[j])
        {
            i++,j++;
            next[i]=j;
        }
        else
            j=next[j];
    }
}
int ans;
void kmp(char *s,char *t)
{
    int i=0,j=0,len1=strlen(s),len2=strlen(t),next[10005];
    getnext(t,next);
    while(i<len1 && j<len2)
    {
        if(j==-1 || s[i]==t[j])
            i++,j++;
        else
            j=next[j];

        if(j==len2)//统计子串出现次数
        {
            j=next[j];
            ans++;
        }
    }
}
int main()
{
    int t;
    cin>>t;
    while(t--)
    {
        cin>>a>>b;
        ans=0;
        kmp(b,a);
        cout<<ans<<endl;
    }
    return 0;
}

出现位置

#include<iostream>
#include<cstring>
#include<cstdio>
#include<algorithm>
#include<cstdlib>
#include<string>
using namespace std;
void getnext(string t,int len,int *next)
{
    int i=0,j=-1;
    next[0]=-1;
    while(i<len)
    {
        if(j==-1 || t[i]==t[j])
        {
            i++,j++;
            next[i]=j;
        }
        else
            j=next[j];
    }
}
int kmp(string s,string t)
{
    int i=0,j=0,next[10005];
    int len1=s.size(),len2=t.size();
    getnext(t,len2,next);
    while(i<len1 && j<len2)
    {
        if(j==-1 || s[i]==t[j])
            i++,j++;
        else
            j=next[j];
    }
    if(j==len2)
        return i-len2+1;
    else
        return -1;
}
string a,b;
int main()
{
    while(getline(cin,b))
    {
        getline(cin,a);
        cout<<kmp(a,b)<<endl;
    }
    return 0;
}

Horspool

#include <iostream>
#include <string>
using namespace std;
int Horspool(string s,string t,int pos)
{
    int len1=s.length(),len2=t.length();
    int ti=len2-1; //目标字符串匹配位置
    int si=pos+ti; //源字符串匹配位置
    if(len1-pos<len2)
        return -1;
    while(ti>=0 && si<len1)
    {
        if(s[si]==t[ti])
            ti--,si--;
        else
        {
            while(ti>=0 && s[si]!=t[ti])
                ti--;
            si+=len2-1-ti;
            ti=len2-1;
        }
    }
    if(si<len1)
        return si+1;
    else
        return -1;
}
int main()
{
    string src,tar;
    while(cin>>tar>>src)
    {
        cout<<Horspool(src,tar,0)<<endl;
    }
    return 0;
}

BM

#include <iostream>
#include <string>
#include <cstring>
const int maxn=102005;
using namespace std;
void build_bad(string pattern,int bad[])
{
    int len=pattern.length();
    memset(bad,len,sizeof(bad));
    for(int i=0;i<len;i++)
        bad[pattern[i]-'A']=len-1-i;
}
void build_good(string pattern,int good[])
{
    int len=pattern.length(),k=0;
    memset(good,len,sizeof(good));
    good[len-1]=1;
    char p[maxn];
    strcpy(p,pattern.c_str());//将string转化为char{]
    for(int i=len-1;i>=0;i--)
    {
        for(int j=0;j<i;j++)
        {
            if(memcmp(p+i,p+j,(len-i)*sizeof(char))==0)//比较内存区域buf1和buf2的前count个字节。
            {
                if(j==0)
                    k=len-i;
                else if(p[i-1]!=p[j-1])
                    good[i-1]=j-1;
            }
        }
    }
    for(int i=0;i<len-1;i++)
    {
        if(good[i]!=len)
            good[i]=len-1-good[i];
        else
        {
            good[i]=len-1-i+good[i];
            if(k && k<=len-1-i)
                good[i]-=k;
        }
    }
}
int Boyer_Moore(string pattern,string text)
{
    int bad[maxn],good[maxn],len1=pattern.length(),len2=text.length();
    build_bad(pattern,bad);
    build_good(pattern,good);
    int i=len1-1,j=len1-1;
    while(j<len2)
    {
        while(i && pattern[i]==text[j])
            i--,j--;
        if(i==0 && pattern[i]==text[j])
            return j;
        else
            j+=max(good[i],bad[text[j]-'A']);
        i=len1-1;
    }
    return -1;
}
int main()
{
    string a,b;
    while(cin>>a>>b)
    {
        cout<<Boyer_Moore(a,b)<<endl;
    }
    return 0;
}

Sunday

#include <iostream>
#include <string>
#include <cstring>
using namespace std;
int Sunday(string s,string t)
{
    int len1=s.length(),len2=t.length(),pos=0,next[26],i,j;
    memset(next,len2,sizeof(next));//init
    for(i=0;i<len2;i++)//set
        next[t[i]-'A']=len2-i;//移动的越大
    while(pos<len1-len2+1)
    {
        i=pos;
        for(j=0;j<len2;j++,i++)
        {
            if(s[i]!=t[j])
            {
                pos+=next[s[len2+pos]-'A'];//一旦不匹配，原串就按照next跳转  
                break;
            }
        }
        if(j==len2)
            return pos;
    }
    return -1;
}
int main()
{
    string src,tar;
    while(cin>>tar>>src)
    {
        cout<<Sunday(src,tar)<<endl;
    }
    return 0;
}