Given a list of strings, you could concatenate these strings together into a loop, where for each string you could choose to reverse it or not. Among all the possible loops, you need to find the lexicographically biggest string after cutting the loop, which will make the looped string into a regular one.

Specifically, to find the lexicographically biggest string, you need to experience two phases:

  1. Concatenate all the strings into a loop, where you can reverse some strings or not and connect them in the same order as given.
  2. Cut and make one breakpoint in any place of the loop, which will make the looped string into a regular one starting from the character at the cutpoint.

And your job is to find the lexicographically biggest one among all the possible regular strings.

Example:

Input: "abc", "xyz"
Output: "zyxcba"
Explanation: You can get the looped string "-abcxyz-", "-abczyx-", "-cbaxyz-", "-cbazyx-",
where '-' represents the looped status.
The answer string came from the fourth looped one,
where you could cut from the middle character 'a' and get "zyxcba".

Note:

  1. The input strings will only contain lowercase letters.
  2. The total length of all the strings will not over 1,000.

给定一组字符串,对于每个字符串可以逆序或者不变,将字符串按照原始顺序拼接组成一个字符串,循环每一个字符串拼接它们,可以选择循环中的一个字符串在某一个位置切开。求能得到的字典序最大的字符串。

最终结果一定是从某个字符串在某个地方切开(包括在string[0]之前切开)。其余字符串是否需要reverse则取决于reverse之后其字典序是否会增大,如果增大了就reverse,否则就保持不变,这样才能保证最后的结果是字典序最大的。对于a字符串而言,是否需要反向,以及从哪里切开,均取决于最终形成的字符串头部的字典序大小。

解法:首先将所有的字符串都处理成字典序最大的(反向或者不反向),然后连接起来。接着遍历每个字符串,试图将它以及它的reverse从每个位置切开,并测试形成的regular字符串,一旦发现其字典序大于当前的最大值,则更新当前最大值。最后返回最大值即可。

首先将res初始化为“a”,即为字典序中最小的字符串。而在测试切开位置的过程中,只有在p1[j]或者p2[j]大于res[0]的时候才更新res,这样就可以减少冗余的字符串拼接操作,从而提高运行效率。

第一步,处理原字符串,如果strs[i].reverse > strs[i], 则strs[i] = strs[i].reverse. 可以确保mid string是最优的

第二步,对于一个string array,比如 "abc", "def", "xyz", 先生成一个mid string,mid string没有最后一个string,比如"abcdef". 然后update mid string:

mid = mid.substr(str.length()) + strs[(i+n-1) % n];
这样mid string就变成 defxyz(由abcdef,去掉abc,加上xyz),xyzabc.

然后扫原string array中的每一个string,生成相应的截取result。比如对于abc,我们有mid string = defxyz, 所以组合是:
正序:abc-defxyz, bc-defxyz-a, c-defxyz-ab, defxyz-abc
反序:cba-defxyz, ba-defxyz-c, a-defxyz-cb, defxyz-cba,
然后所有循环中挑出全局最大的.

参考 For every given string, we replace the string with the lexicographically larger string out of the original string and the reversed one. After this, we pick up every new string(chosen as the string on which the cuts will be applied), and apply a cut at all the positions of the currently picked string and form the full concantenated string keeping the rest of the newly formed strings intact. We also reverse the current string and follow the same process. While doing this, we keep a track of the largest lexicographic string found so far.

大厂题:阿里巴巴

Java:

public class Solution {
public String splitLoopedString(String[] strs) {
// 先各自翻转一下
for (int i = 0; i < strs.length; i++) {
String rev = new StringBuilder(strs[i]).reverse().toString();
if (strs[i].compareTo(rev) < 0)
strs[i] = rev;
} String res = "";
// 尝试every str作为那个被cut的
for (int i = 0; i < strs.length; i++) {
String rev = new StringBuilder(strs[i]).reverse().toString();
// 比较其 翻转 和 没翻转
for (String st: new String[] {strs[i], rev}) {
// try every k 作为分界点
for (int k = 0; k < st.length(); k++) {
// 拼接其他strs到t
StringBuilder t = new StringBuilder(st.substring(k));
for (int j = i + 1; j < strs.length; j++)
t.append(strs[j]);
for (int j = 0; j < i; j++)
t.append(strs[j]);
t.append(st.substring(0, k));
// 看t是否最大
if (t.toString().compareTo(res) > 0)
res = t.toString();
}
}
}
return res;
}
}  

Java: DFS

public class Solution {
String res = "";
public String splitLoopedString(String[] strs) {
dfs(strs, "", 0, strs.length);
return res;
}
public void dfs(String[] strs, String s, int i, int n) {
if (i < n) {
dfs(strs, s + strs[i], i + 1, n);
dfs(strs, s + new StringBuffer(strs[i]).reverse().toString(), i + 1, n);
} else {
for (int j = 0; j < s.length(); j++) {
String t = s.substring(j) + s.substring(0, j);
if (t.compareTo(res) > 0)
res = t;
}
}
}
}

Java: BFS

public class Solution {

    public String splitLoopedString(String[] strs) {
Queue < String > queue = new LinkedList < > ();
String res = "";
int i = 0, j = 0;
queue.add("");
while (i < strs.length) {
String t = queue.remove();
queue.add(t + strs[i]);
queue.add(t + new StringBuffer(strs[i]).reverse().toString());
j++;
if (j == 1 << i) {
i++;
j = 0;
}
}
while (!queue.isEmpty()) {
String t = queue.remove();
for (int k = 0; k < t.length(); k++) {
String t1 = t.substring(k) + t.substring(0, k);
if (t1.compareTo(res) > 0)
res = t1;
}
}
return res;
}
}

Java: Optimized Solution 

public class Solution {
public String splitLoopedString(String[] strs) {
for (int i = 0; i < strs.length; i++) {
String rev = new StringBuilder(strs[i]).reverse().toString();
if (strs[i].compareTo(rev) < 0)
strs[i] = rev;
}
String res = "";
for (int i = 0; i < strs.length; i++) {
String rev = new StringBuilder(strs[i]).reverse().toString();
for (String st: new String[] {strs[i], rev}) {
for (int k = 0; k < st.length(); k++) {
StringBuilder t = new StringBuilder(st.substring(k));
for (int j = i + 1; j < strs.length; j++)
t.append(strs[j]);
for (int j = 0; j < i; j++)
t.append(strs[j]);
t.append(st.substring(0, k));
if (t.toString().compareTo(res) > 0)
res = t.toString();
}
}
}
return res;
}
}  

Python:

class Solution(object):
def splitLoopedString(self, strs):
"""
:type strs: List[str]
:rtype: str
"""
strs = [max(s, s[::-1]) for s in strs]
ans = ''
for i, st in enumerate(strs):
left, right = ''.join(strs[:i]), ''.join(strs[i+1:])
for s in (st, st[::-1]):
for j in range(len(s)):
ans = max(ans, s[j:] + right + left + s[:j])
return ans  

C++:

class Solution {
public:
string splitLoopedString(vector<string>& strs) {
if(strs.empty()) return "";
else if(strs.size() == 1) return max(strs[0], string(strs[0].rbegin(), strs[0].rend()));
string all = "";
int n = strs.size();
for(int i=0; i<n; i++){
string temp = string(strs[i].rbegin(), strs[i].rend());
if(temp > strs[i]) strs[i] = temp;
}
for(int i=0; i<n-1; i++){
all += strs[i];
}
string result = all + strs[n-1];
for(int i=0; i<n; i++){
string str = strs[i], rev = string(strs[i].rbegin(), strs[i].rend());
all = all.substr(str.length()) + strs[(i+n-1) % n];
for(int j=0; j<=str.length(); j++){
string s1 = str.substr(j) + all + str.substr(0, j), s2 = rev.substr(j) + all + rev.substr(0, j);
if(s1 >= s2 && s1 > result) result = s1;
else if(s2 >= s1 && s2 > result) result = s2;
}
}
return result;
}
};  

C++:

class Solution {
public:
string splitLoopedString(vector<string>& strs) {
if (strs.empty()) return "";
string s = "", res = "a";
int n = strs.size(), cur = 0;
for (string str : strs) {
string t = string(str.rbegin(), str.rend());
s += str > t ? str : t;
}
for (int i = 0; i < n; ++i) {
string t1 = strs[i], t2 = string(t1.rbegin(), t1.rend());
string mid = s.substr(cur + t1.size()) + s.substr(0, cur);
for (int j = 0; j < strs[i].size(); ++j) {
if (t1[j] >= res[0]) res = max(res, t1.substr(j) + mid + t1.substr(0, j));
if (t2[j] >= res[0]) res = max(res, t2.substr(j) + mid + t2.substr(0, j));
}
cur += strs[i].size();
}
return res;
}
};

  

All LeetCode Questions List 题目汇总

[LeetCode] 555. Split Concatenated Strings 分割串联字符串的更多相关文章

  1. [LeetCode] Split Concatenated Strings 分割串联字符串

    Given a list of strings, you could concatenate these strings together into a loop, where for each st ...

  2. 【LeetCode】555. Split Concatenated Strings 解题报告(C++)

    作者: 负雪明烛 id: fuxuemingzhu 个人博客:http://fuxuemingzhu.cn/ 目录 题目描述 题目大意 解题方法 遍历 日期 题目地址:https://leetcode ...

  3. LeetCode Split Concatenated Strings

    原题链接在这里:https://leetcode.com/problems/split-concatenated-strings/description/ 题目: Given a list of st ...

  4. [LeetCode] Encode and Decode Strings 加码解码字符串

    Design an algorithm to encode a list of strings to a string. The encoded string is then sent over th ...

  5. [LeetCode] 249. Group Shifted Strings 分组偏移字符串

    Given a string, we can "shift" each of its letter to its successive letter, for example: & ...

  6. LeetCode OJ:Isomorphic Strings(同构字符串)

    Given two strings s and t, determine if they are isomorphic. Two strings are isomorphic if the chara ...

  7. C语言实现split以某个字符分割一个字符串

    方式一: 使用strtok # include <string.h> # include <stdio.h> void split(char *src,const char * ...

  8. split 将字符串分割成字符串数组

    list_name = list_name.split(","); split() 方法用于把一个字符串分割成字符串数组. 语法 stringObject.split(separa ...

  9. java:字符串的split方法,使用多个分隔符,分割一个字符串

    java语言中,多个分隔符,分割一个字符串: String[] tmpAuthors=tempAuthorStr.split(";|,|:|,"); 可以在线测试:java代码 在 ...

随机推荐

  1. P4560 [IOI2014]Wall 砖墙

    题目描述 给定一个长度为 nn且初始值全为 00的序列.你需要支持以下两种操作: Add L, R, hL,R,h:将序列 [L, R][L,R]内所有值小于 hh的元素都赋为 hh,此时不改变高度大 ...

  2. 数据库jdbc链接:mysql, oracle, postgresql

    #db mysql#jdbc.driver=com.mysql.jdbc.Driver#jdbc.url=jdbc:mysql://localhost:3306/mysql?&useUnico ...

  3. oracle数据库(四)

    子查询与高级查询 我们在检索数据库的时候,需要将多个表关联起来进行查询,最常用的有子查询.连接查询和集合查询,子查询可以从另外一个表获取数据,连接查询可以指定多个表的连接方式,集合查询可以将两个或者多 ...

  4. django-全文解锁和搜索引擎

    安装和配置 全文检索安装 pip install django-haystack==2.5.1 # 2.7.0只支持django1.11以上版本 搜索引擎安装 pip install whoosh 安 ...

  5. 域权限维持:Ptk(Pass The Ticket)

    Kerberos协议传送门:https://www.cnblogs.com/zpchcbd/p/11707302.html 1.黄金票据 2.白银票据

  6. 六.深浅copy

    先问问大家,什么是拷贝?拷贝是音译的词,其实他是从copy这个英文单词音译过来的,那什么是copy? copy其实就是复制一份,也就是所谓的抄一份.深浅copy其实就是完全复制一份,和部分复制一份的意 ...

  7. oracle 游标使用详解

    -- 声明游标:CURSOR cursor_name IS select_statement--For 循环游标--(1)定义游标--(2)定义游标变量--(3)使用for循环来使用这个游标decla ...

  8. BZOJ 3553: [Shoi2014]三叉神经树 LCT

    犯傻了,想到了如果是 0->1 的话就找最深的非 1 编号,是 1 -> 0 的话就找最深的非 0 编号. 但是没有想到这个东西可以直接维护. 假设不考虑叶子节点,那么如果当前点的值是 1 ...

  9. proxysql 学习一 proxysql docker 运行试用

    proxysql 是一个比较强大的mysql proxy 服务,支持动态mysql 实例调整,查询重写,查询cache,监控,数据镜像,读写分离 以及ha,最近已经发布了2.0 ,很值得试用下 环境准 ...

  10. 使用singer tap-postgres 同步数据到pg

    singer 是一个很不错的开源etl 解决方案,以下演示一个简单的数据从pg 同步到pg 很简单就是使用tap-postgres + target-postgres 环境准备 对于测试的环境的数据库 ...