回归算法第一天:字符串哈希
意思就是把字符串反射成一个整型来记录;
贴模板:
unsigned int BKDRHash(char *str) {
unsigned int seed = 31, key = 0;
while(*str) {
key = key * seed + (*str++);
cout << key << endl;
}
return key & 0x7fffffff;
}
这里的返回值是unsigned int
,key
的属性也是unsigned int
,它会自动取模,最后对0x7fffffff
,返回一个unsighed int
属性的只,代表这个字符串的hash值。
但是平常我们不仅需要这个字符串的hash值,还需要字串的hash值,这里就可以用下面的方法写:
const int seed = 31;
const int maxn = 4e4 + 50;
typedef unsigned long long ull;
ull _hash[maxn], base[maxn];
char ch[maxn];
ull string_hash(int l, int r) {
return _hash[r] - _hash[l - 1] * base[r - l + 1];
}
base[0] = 1;
for (int i = 1; i < maxn; i++) base[i] = base[i - 1] * seed;
_hash[0] = 0;
for (int i = 1; i <= len; i++) _hash[i] = _hash[i - 1] * seed + ch[i - 1] - 'a';
这里主要是要明白string_hash
的原理。
base
数组的意思是seed
的几次方的值。比如base[1] = 31, base[2] = 31 * 31
;
举个例子:31342,我们想求34的值,就可以用3134 - 31 * 100 = 34。
同理,我们有一个字符串"fvcmojiew"
,我们想求hash("cm")
值,就可以用hash("fvcm") - hash("fv") * 31 * 31
。所以当我们想求第l到第r的hash值时,就只要用_hash[r] - _hash[l - 1] * base[r - l + 1]
,这里的r - l + 1
就是要多乘几个seed。
这里再提供一个前缀和后缀的写法,时leetcode的1616题:
typedef unsigned long long ull;
const int seed = 31;
const int maxn = 1e5 + 50;
ull base[maxn];
ull hash_a_pre[maxn], hash_a_suf[maxn], hash_b_pre[maxn], hash_b_suf[maxn];
class Solution {
public:
bool check1(int index, int n) {
// 1 index + 1 index + 2 n 25489
ull prefix = hash_a_pre[index + 1] * base[n - index - 1] + (hash_b_pre[n] - hash_b_pre[index + 1] * base[n - index - 1]);
//0 index index + 1 n - 1
ull suffix = hash_b_suf[index + 1] * base[index + 1] + (hash_a_suf[0] - hash_a_suf[index + 1] * base[index + 1]);
return prefix == suffix;
}
bool check2( int index, int n) {
ull prefix = hash_b_pre[index + 1] * base[n - index - 1] + (hash_a_pre[n] - hash_a_pre[index + 1] * base[n - index - 1]);
ull suffix = hash_a_suf[index + 1] * base[index + 1] + (hash_b_suf[0] - hash_b_suf[index + 1] * base[index + 1]);
return prefix == suffix;
}
bool checkPalindromeFormation(string a, string b) {
int n = a.length();
base[0] = 1;
for (int i = 1; i <= n; i++) base[i] = base[i - 1] * seed;
hash_a_pre[0] = hash_b_pre[0] = hash_b_suf[n] = hash_a_suf[n] = 0;
for (int i = 1; i <= n; i++) hash_a_pre[i] = hash_a_pre[i - 1] * seed + a[i - 1] - 'a';
for (int i = 1; i <= n; i++) hash_b_pre[i] = hash_b_pre[i - 1] * seed + b[i - 1] - 'a';
for (int i = n - 1; i >= 0; i--) hash_a_suf[i] = hash_a_suf[i + 1] * seed + a[i] - 'a';
for (int i = n - 1; i >= 0; i--) hash_b_suf[i] = hash_b_suf[i + 1] * seed + b[i] - 'a';
if (hash_a_pre[n] == hash_a_suf[0]) return true;
if (hash_b_pre[n - 1] == hash_b_suf[0]) return true;
for (int i = 0; i < n; i++) {
if (check1(i,n)) return true;
if (check2(i,n)) return true;
}
return false;
}
};