Sereja loves number sequences very much. That's why he decided to make himself a new one following a certain algorithm.
Sereja takes a blank piece of paper. Then he starts writing out the sequence in m stages. Each time he either adds a new number to the end of the sequence or takes l first elements of the current sequence and adds them c times to the end. More formally, if we represent the current sequence as a1, a2, ..., an, then after we apply the described operation, the sequence transforms intoa1, a2, ..., an[, a1, a2, ..., al] (the block in the square brackets must be repeated c times).
A day has passed and Sereja has completed the sequence. He wonders what are the values of some of its elements. Help Sereja.
The first line contains integer m (1 ≤ m ≤ 105) — the number of stages to build a sequence.
Next m lines contain the description of the stages in the order they follow. The first number in the line is a type of stage (1 or 2). Type 1 means adding one number to the end of the sequence, in this case the line contains integer xi (1 ≤ xi ≤ 105) — the number to add. Type 2 means copying a prefix of length li to the end ci times, in this case the line further contains two integers li, ci(1 ≤ li ≤ 105, 1 ≤ ci ≤ 104), li is the length of the prefix, ci is the number of copyings. It is guaranteed that the length of prefix li is never larger than the current length of the sequence.
The next line contains integer n (1 ≤ n ≤ 105) — the number of elements Sereja is interested in. The next line contains the numbers of elements of the final sequence Sereja is interested in. The numbers are given in the strictly increasing order. It is guaranteed that all numbers are strictly larger than zero and do not exceed the length of the resulting sequence. Consider the elements of the final sequence numbered starting from 1 from the beginning to the end of the sequence.
Please, do not use the %lld specifier to read or write 64-bit integers in С++. It is preferred to use the cin, cout streams or the %I64dspecifier.
Print the elements that Sereja is interested in, in the order in which their numbers occur in the input.
6 1 1 1 2 2 2 1 1 3 2 5 2 1 4 16 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
1 2 1 2 3 1 2 1 2 3 1 2 1 2 3 4
题目大意是:对于两种不同的操作,操作1表示直接将一个数加到数组后面,操作2表示对前面长度为L的数串进行复制。
看到题目数据,刚开始想着直接模拟生成数组不就行了。。然后就RE。。数组开大以后爆内存了。。
无奈,只好去看题解:根据题目意思可知,复制的长度最长不超过10^5,也就是说,10^5后面的数字一定可以在前面找到的。所以可以记录前10^5的数组,对于后面的数一定可以在前面找到,所以可以记录下每次操作以后的数字串长度,对要找的数字的位置进行二分查找。找到其位置后,如果是属于操作1的,则直接是这个数。如果是操作2的,就继续往前面找。
#include<stdio.h> #include<string.h> #define M 100005 int type[M],m,a[M],b[M]; __int64 len[M],pos; int finds(__int64 pos){ int l=1,r=m,mid; while(l<r){ mid=(l+r)/2; if(len[mid]<pos) l=mid+1; else r=mid; } return l; } int solve(__int64 pos) { int k=finds(pos); if(type[k]==1)return a[k]; pos=pos-len[k-1]; pos=pos%a[k]; if(pos==0)pos=a[k]; return solve(pos); } int main() { int n,c,i,j; scanf("%d",&m); len[0]=0; for(i=1;i<=m;i++) { scanf("%d",&type[i]); if(type[i]==1){ scanf("%d",&a[i]); len[i]=len[i-1]+1; } else if(type[i]==2){ scanf("%d%d",&a[i],&b[i]); len[i]=len[i-1]+a[i]*b[i]; } } scanf("%d",&n); while(n--) { scanf("%I64d",&pos); printf("%d ",solve(pos)); } printf("\n"); return 0; }