In this problem you have to simulate the workflow of one-thread server. There are n queries to process, the i-th
will be received at moment ti and
needs to be processed for di units
of time. All ti are
guaranteed to be distinct.

When a query appears server may react in three possible ways:

  1. If server is free and query queue is empty, then server immediately starts to process this query.
  2. If server is busy and there are less than b queries in the queue, then new query is added to the end of the queue.
  3. If server is busy and there are already b queries pending in the queue, then new query is just rejected and will never be processed.

As soon as server finished to process some query, it picks new one from the queue (if it's not empty, of course). If a new query comes at some moment x,
and the server finishes to process another query at exactly the same moment, we consider that first query is picked from the queue and only then new query appears.

For each query find the moment when the server will finish to process it or print -1 if this query will be rejected.

Input

The first line of the input contains two integers n and b (1 ≤ n, b ≤ 200 000) —
the number of queries and the maximum possible size of the query queue.

Then follow n lines with queries descriptions (in chronological order). Each description consists of two integers ti and di (1 ≤ ti, di ≤ 109),
where ti is
the moment of time when the i-th query appears and di is
the time server needs to process it. It is guaranteed that ti - 1 < ti for
all i > 1.

Output

Print the sequence of n integers e1, e2, ..., en,
where ei is
the moment the server will finish to process the i-th query (queries are numbered in the order they appear in the input) or  - 1 if
the corresponding query will be rejected.

Examples
input
5 1
2 9
4 8
10 9
15 2
19 1
output
11 19 -1 21 22
input
4 1
2 8
4 8
10 9
15 2
output
10 18 27 -1
Note

Consider the first sample.

  1. The server will start to process first query at the moment 2 and will finish to process it at the moment 11.
  2. At the moment 4 second query appears and proceeds to the queue.
  3. At the moment 10 third query appears. However, the server is still busy with query 1, b = 1 and
    there is already query 2 pending in the queue, so third query is just rejected.
  4. At the moment 11 server will finish to process first query and will take the second query from the queue.
  5. At the moment 15 fourth query appears. As the server is currently busy it proceeds to the queue.
  6. At the moment 19 two events occur simultaneously: server finishes to proceed the second query and the fifth query appears. As was said in the
    statement above, first server will finish to process the second query, then it will pick the fourth query from the queue and only then will the fifth query appear. As the queue is empty fifth query is proceed there.
  7. Server finishes to process query number 4 at the moment 21.
    Query number 5 is picked from the queue.
  8. Server finishes to process query number 5 at the moment 22.
题意:有n个任务要处理,每一次只能处理一个,待处理的任务放在任务队列中,每次一个任务来的时候,队列里的任务都处理完了,那么就开始处理这个任务;如果队列里的任务没有处理完,而且队列中的任务数量小于b,那么处理把这个任务加入队列;否则把这个任务舍去,问每一个任务执行结束后的时间。

思路:开一个队列,用来模拟任务队列。每次一个任务到来的时候,先把这个任务到来前的能处理的任务都处理完,然后看队列是不是为空并且处理完所有任务的时间小于等于这个任务到来的时间(即判断能不能直接处理这个任务),如果不能直接处理这个任务的话,判断任务队列里的任务个数是不是等于b,如果是就舍去这个任务,否则把这个任务放入队列中。

#include<iostream>
#include<stdio.h>
#include<stdlib.h>
#include<string.h>
#include<math.h>
#include<vector>
#include<map>
#include<set>
#include<queue>
#include<stack>
#include<string>
#include<bitset>
#include<algorithm>
using namespace std;
typedef long long ll;
typedef long double ldb;
#define inf 99999999
#define pi acos(-1.0)
#define maxn 200050
ll t[maxn],d[maxn];
int q[511111];
ll ans[maxn]; int main()
{
int i,j,x;
ll n,m,b,ed;
while(scanf("%lld%lld",&n,&b)!=EOF)
{
for(i=1;i<=n;i++){
scanf("%lld%lld",&t[i],&d[i]);
}
int front,rear;
front=1;rear=0;
ed=0;
for(i=1;i<=n;i++){
while(front<=rear){
x=q[front];
if(ed<=t[i]){
ed=max(ed,t[x])+d[x];
ans[x]=ed;
front++;
}
else break;
}
if(ed<=t[i] && front>rear ){
ed=t[i]+d[i];
ans[i]=ed;
}
else if(rear-front+1<b){
rear++;
q[rear]=i;
}
else ans[i]=-1;
}
while(front<=rear){
x=q[front];
front++;
ed=max(ed,t[x])+d[x];
ans[x]=ed;
}
for(i=1;i<=n;i++){
if(i==n)printf("%lld\n",ans[i]);
else printf("%lld ",ans[i]);
} }
return 0; }


codeforces644B. Processing Queries (模拟)的更多相关文章

  1. CROC 2016 - Qualification B. Processing Queries 模拟

    B. Processing Queries 题目连接: http://www.codeforces.com/contest/644/problem/B Description In this prob ...

  2. Code Forces 644B Processing Queries

    B. Processing Queries time limit per test5 seconds memory limit per test256 megabytes inputstandard ...

  3. Codeforces Round #515 (Div. 3) C. Books Queries (模拟)

    题意:有一个一维的书架,\(L\)表示在最左端放一本书,\(R\)表示在最右端放一本书,\(?\)表示从左数或从右数,最少数多少次才能得到要找的书. 题解:我们开一个稍微大一点的数组,从它的中间开始模 ...

  4. 代码本色 用编程模拟自然系统 (Daniel Shiffman 著)

    https://cdnjs.cloudflare.com/ajax/libs/p5.js/0.5.7/p5.js http://www.box2d.org http://www.jbox2d.org ...

  5. Programming Entity Framework 翻译(1)-目录

    1. Introducing the ADO.NET Entity Framework ado.net entity framework 介绍 1 The Entity Relationship Mo ...

  6. 【mysql】关于临时表

    mysql官方的介绍 In some cases, the server creates internal temporary tables while processing queries. Suc ...

  7. open_table与opened_table

    好多人在调优Mysql的时候,总是对open_tables和opend_tables两个参数分别不清. 网上好多解释都是这样的:open_tables:当前打开表的数量opened_tables:当前 ...

  8. 【MySQL】查询使用临时表

    MySQL查询产生临时表的分析 官网说明的地址:http://dev.mysql.com/doc/refman/5.5/en/internal-temporary-tables.html 参考:htt ...

  9. mysql --The MEMORY Storage Engine--官方文档

    原文地址:http://dev.mysql.com/doc/refman/5.7/en/memory-storage-engine.html The MEMORY storage engine (fo ...

随机推荐

  1. Zap简单使用

    前言 zap 是 uber 开源的一个日志记录的包, uber 在 go 的领域建树颇多, zap 更是优秀, 相比于自带的 log ,他有更多的功能, 当然, 最显眼的还是他很快, 本文介绍 zap ...

  2. Docker学习笔记之Dockerfile

    Dockerfile的编写格式为<命令><形式参数>,命令不区分大小写,但一般使用大写字母.Docker会依据Dockerfile文件中编写的命令顺序依次执行命令.Docker ...

  3. alter column和modify column

    5.6中,发现其实alter column 和更改modify column 步骤是一样的 mysql> create table xs(name varchar(12),age int def ...

  4. 【Linux】md5sum 生产所有文件的md5值,并对照目标文件是否相同

    现在加入有很多很多文件需要测试md5,想看下是否都传输成功了,如何批量生成文件的md5并且逐条对照呢? 下面来简单介绍下 md5sum这个命令有一个选项"-c" 这个选项的意思是c ...

  5. ctfhub技能树—信息泄露—svn泄露

    打开靶机 查看页面信息 使用dvcs-ripper工具进行处理 ./rip-svn.pl -v -u http://challenge-3b6d43d72718eefb.sandbox.ctfhub. ...

  6. 避免用using包装DbContext【翻译】

    EF和EF Core 的DbContext类实现IDisposable接口.因此,很多最佳编程实践中都建议你将它们放在一个using()块中.不幸的是,至少在Web应用程序中,这样做通常不是一个好主意 ...

  7. RocketMq消息 demo

    参考 https://blog.csdn.net/asdf08442a/article/details/54882769 整理出来的测试 demo 1.produce 生产者 1 package co ...

  8. 面试常问的ArrayQueue底层实现

    public class ArrayQueue<T> extends AbstractList<T>{ //定义必要的属性,容量.数组.头指针.尾指针 private int ...

  9. CentOS对接GlusterFS

    存储节点部署示例环境,仅供参考 主机名 IP 系统 gfs01 10.10.10.13 CentOS 7.4.1708 gfs02 10.10.10.14 CentOS 7.4.1708 一.Glus ...

  10. C++ Primer Plus读书笔记(五)循环和关系表达式

    1.前缀运算符的优先级是从右到左,例如: 1 *++p; //这句话的含义是先对p进行++,再取* 2.循环 c++11新增了一种循环方式,有点和python类似,如下: 1 array<int ...