# 「力扣」第 127 题:单词接龙(困难)
# 视频讲解
这道题在 题解 (opens new window) 和 B 站 (opens new window) 可以收看视频讲解,可以点击下面的视频右上角「去 bilibili 观看」,选择快速播放,获得更好的观看体验。
# 题目描述
字典 wordList
中从单词 beginWord
和 endWord
的 转换序列 是一个按下述规格形成的序列:
- 序列中第一个单词是
beginWord
。 - 序列中最后一个单词是
endWord
。 - 每次转换只能改变一个字母。
- 转换过程中的中间单词必须是字典
wordList
中的单词。
给你两个单词 beginWord
和 endWord
和一个字典 wordList
,找到从 beginWord
到 endWord
的 最短转换序列 中的 单词数目 。如果不存在这样的转换序列,返回 0。
示例 1:
输入:beginWord = "hit", endWord = "cog", wordList = ["hot","dot","dog","lot","log","cog"]
输出:5
解释:一个最短转换序列是 "hit" -> "hot" -> "dot" -> "dog" -> "cog", 返回它的长度 5。
示例 2:
输入:beginWord = "hit", endWord = "cog", wordList = ["hot","dot","dog","lot","log"]
输出:0
解释:endWord "cog" 不在字典中,所以无法进行转换。
提示:
1 <= beginWord.length <= 10
endWord.length == beginWord.length
1 <= wordList.length <= 5000
wordList[i].length == beginWord.length
beginWord
、endWord
和wordList[i]
由小写英文字母组成beginWord != endWord
wordList
中的所有字符串 互不相同
# 思路分析
# 一句话题解
无向图中两个顶点之间的最短路径的长度,可以通过广度优先遍历得到;
为什么 BFS 得到的路径最短?可以把起点和终点所在的路径拉直来看,两点之间线段最短;
已知目标顶点的情况下,可以分别从起点和目标顶点(终点)执行广度优先遍历,直到遍历的部分有交集,这是双向广度优先遍历的思想。
# 方法一:广度优先遍历
分析题意:
- 「转换」意即:两个单词对应位置只有一个字符不同,例如 "hit" 与 "hot",这种转换是可以逆向的,因此,根据题目给出的单词列表,可以构建出一个无向(无权)图;
- 如果一开始就构建图,每一个单词都需要和除它以外的另外的单词进行比较,复杂度是
,这里 是单词列表的长度; - 为此,我们在遍历一开始,把所有的单词列表放进一个哈希表中,然后在遍历的时候构建图,每一次得到在单词列表里可以转换的单词,复杂度是
,借助哈希表,找到邻居与 无关; - 使用 BFS 进行遍历,需要的辅助数据结构是:
- 队列;
visited
集合。说明:可以直接在wordSet
(由wordList
放进集合中得到)里做删除。但更好的做法是新开一个哈希表,遍历过的字符串放进哈希表里。这种做法具有普遍意义。绝大多数在线测评系统和应用场景都不会在意空间开销。
Java 代码:
参考代码 1:
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashSet;
import java.util.LinkedList;
import java.util.List;
import java.util.Queue;
import java.util.Set;
public class Solution {
public int ladderLength(String beginWord, String endWord, List<String> wordList) {
// 先将 wordList 放到哈希表里,便于判断某个单词是否在 wordList 里
Set<String> wordSet = new HashSet<>(wordList);
if (wordSet.size() == 0 || !wordSet.contains(endWord)) {
return 0;
}
wordSet.remove(beginWord);
// 图的广度优先遍历,必须使用的队列和表示是否访问过的 visited (数组,哈希表)
Queue<String> queue = new LinkedList<>();
queue.offer(beginWord);
Set<String> visited = new HashSet<>();
visited.add(beginWord);
int wordLen = beginWord.length();
// 包含起点,因此初始化的时候步数为 1
int step = 1;
while (!queue.isEmpty()) {
int currentSize = queue.size();
for (int i = 0; i < currentSize; i++) {
// 依次遍历当前队列中的单词
String word = queue.poll();
char[] charArray = word.toCharArray();
// 修改每一个字符
for (int j = 0; j < wordLen; j++) {
// 一轮以后应该重置,否则结果不正确
char originChar = charArray[j];
for (char k = 'a'; k <= 'z'; k++) {
if (k == originChar) {
continue;
}
charArray[j] = k;
String nextWord = String.valueOf(charArray);
if (wordSet.contains(nextWord)) {
if (nextWord.equals(endWord)) {
return step + 1;
}
if (!visited.contains(nextWord)) {
queue.add(nextWord);
// 注意:添加到队列以后,必须马上标记为已经访问
visited.add(nextWord);
}
}
}
// 恢复
charArray[j] = originChar;
}
}
step++;
}
return 0;
}
public static void main(String[] args) {
String beginWord = "hit";
String endWord = "cog";
List<String> wordList = new ArrayList<>();
String[] wordListArray = new String[]{"hot", "dot", "dog", "lot", "log", "cog"};
Collections.addAll(wordList, wordListArray);
Solution solution = new Solution();
int res = solution.ladderLength(beginWord, endWord, wordList);
System.out.println(res);
}
}
Python 代码:
from typing import List
from collections import deque
class Solution:
def ladderLength(self, beginWord: str, endWord: str, wordList: List[str]) -> int:
word_set = set(wordList)
if len(word_set) == 0 or endWord not in word_set:
return 0
if beginWord in word_set:
word_set.remove(beginWord)
queue = deque()
queue.append(beginWord)
visited = set(beginWord)
visited.add(beginWord)
word_len = len(beginWord)
step = 1
while queue:
current_size = len(queue)
for i in range(current_size):
word = queue.popleft()
word_list = list(word)
for j in range(word_len):
origin_char = word_list[j]
for k in range(26):
word_list[j] = chr(ord('a') + k)
next_word = ''.join(word_list)
if next_word in word_set:
if next_word == endWord:
return step + 1
if next_word not in visited:
queue.append(next_word)
word_set.remove(next_word)
word_list[j] = origin_char
step += 1
return 0
if __name__ == '__main__':
beginWord = "hit"
endWord = "cog"
wordList = ["hot", "dot", "dog", "lot", "log", "cog"]
solution = Solution()
res = solution.ladderLength(beginWord, endWord, wordList)
print(res)
# 方法二:双向 BFS
- 已知目标顶点的情况下,可以分别从起点和目标顶点(终点)执行广度优先遍历,直到遍历的部分有交集。这种方式搜索的单词数量会更小一些;
- 更合理的做法是,每次从单词数量小的集合开始扩散;
- 这里
beginVisited
和endVisited
交替使用,等价于单向 BFS 里使用队列,每次扩散都要加到总的visited
里。
参考代码 2:
Java 代码:
import java.util.ArrayList;
import java.util.Collections;
import java.util.HashSet;
import java.util.List;
import java.util.Set;
public class Solution {
public int ladderLength(String beginWord, String endWord, List<String> wordList) {
// 先将 wordList 放到哈希表里,便于判断某个单词是否在 wordList 里
Set<String> wordSet = new HashSet<>(wordList);
if (wordSet.size() == 0 || !wordSet.contains(endWord)) {
return 0;
}
// 标准写法,总的 visited 数组
Set<String> visited = new HashSet<>();
// 分别用左边和右边扩散的哈希表代替单向 BFS 里的队列
Set<String> beginVisited = new HashSet<>();
beginVisited.add(beginWord);
Set<String> endVisited = new HashSet<>();
endVisited.add(endWord);
int len = beginWord.length();
int step = 1;
// 简化成 while (!beginVisited.isEmpty()) 亦可
while (!beginVisited.isEmpty() && !endVisited.isEmpty()) {
// 打开以方便调试
// System.out.println("beginVisited => " + beginVisited);
// System.out.println(" endVisited => " + endVisited + "\n");
// 优先选择小的哈希表进行扩散,考虑到的情况更少
if (beginVisited.size() > endVisited.size()) {
Set<String> temp = beginVisited;
beginVisited = endVisited;
endVisited = temp;
}
// 逻辑到这里,保证 beginVisited 是相对较小的集合
// nextLevelVisited 在扩散完成以后,会成为新的 beginVisited
Set<String> nextLevelVisited = new HashSet<>();
for (String word : beginVisited) {
char[] charArray = word.toCharArray();
for (int i = 0; i < len; i++) {
char currentChar = charArray[i];
for (char c = 'a'; c <= 'z'; c++) {
if (charArray[i] == c) {
continue;
}
charArray[i] = c;
String nextWord = String.valueOf(charArray);
if (wordSet.contains(nextWord)) {
if (endVisited.contains(nextWord)) {
return step + 1;
}
if (!visited.contains(nextWord)) {
nextLevelVisited.add(nextWord);
visited.add(nextWord);
}
}
}
// 恢复,下次再用
charArray[i] = currentChar;
}
}
// 这一行代表表示从 begin 这一侧向外扩散了一层
beginVisited = nextLevelVisited;
step++;
}
return 0;
}
public static void main(String[] args) {
List<String> wordList = new ArrayList<>();
String[] words = {"hot", "dot", "dog", "lot", "log", "cog"};
Collections.addAll(wordList, words);
Solution solution = new Solution();
String beginWord = "hit";
String endWord = "cog";
int ladderLength = solution.ladderLength(beginWord, endWord, wordList);
System.out.println(String.format("从 %s 到 %s 的最短转换序列的长度:%d。", beginWord, endWord, ladderLength));
}
}
Python 代码:
from typing import List
from collections import deque
class Solution:
def ladderLength(self, beginWord: str, endWord: str, wordList: List[str]) -> int:
word_set = set(wordList)
if len(word_set) == 0 or endWord not in word_set:
return 0
if beginWord in word_set:
word_set.remove(beginWord)
visited = set()
visited.add(beginWord)
visited.add(endWord)
begin_visited = set()
begin_visited.add(beginWord)
end_visited = set()
end_visited.add(endWord)
word_len = len(beginWord)
step = 1
# 简化成 while begin_visited 亦可
while begin_visited and end_visited:
# 打开帮助调试
# print(begin_visited)
# print(end_visited)
if len(begin_visited) > len(end_visited):
begin_visited, end_visited = end_visited, begin_visited
next_level_visited = set()
for word in begin_visited:
word_list = list(word)
for j in range(word_len):
origin_char = word_list[j]
for k in range(26):
word_list[j] = chr(ord('a') + k)
next_word = ''.join(word_list)
if next_word in word_set:
if next_word in end_visited:
return step + 1
if next_word not in visited:
next_level_visited.add(next_word)
visited.add(next_word)
word_list[j] = origin_char
begin_visited = next_level_visited
step += 1
return 0
if __name__ == '__main__':
beginWord = "hit"
endWord = "cog"
wordList = ["hot", "dot", "dog", "lot", "log", "cog"]
solution = Solution()
res = solution.ladderLength(beginWord, endWord, wordList)
print(res)
作者:liweiwei1419 链接:https://suanfa8.com/breadth-first-search/solutions-3/0127-word-ladder 来源:算法吧 著作权归作者所有。商业转载请联系作者获得授权,非商业转载请注明出处。