字典树(Trie)—— 存储字符串作为key的数据结构。因为Trie存储字符串时特殊的结构,它和BST、HashTable等相比具有独特的优势(如前缀匹配等)。
满足条件
每个节点只存储一个字母;
位于同一分支上的字母组成字符串(尾节点需要单独标记)。
实现方式
Trie有三种实现方式:
数组 —— 字母有限,每个节点可以使用固定大小的数组作为child节点的索引;
BST —— 使用BST作为child节点的索引,与数组相比可以节约一定的空间;
HashTable —— 使用HashTable作为child节点的索引,具有常量时间的search *** 作。
Operation
public void add(String w); // 将字符串w添加进Trie中存储
public List<String> allString(); // 返回Trie中存储的所有字符串
public String longestPrefixOf(String w); // 返回Trie中和w匹配的最长的前缀
public List<String> stringsWithPrefix(String prefix); // 返回所有带有前缀prefix的字符串
java代码
以BST为例存储Trie的child节点(图片空间有限,只选取了前几个节点)
Trie.java
import java.util.ArrayList;
import java.util.List;
public class Trie {
private Node root;
Trie() {
root = new Node( '0', null);
}
public void add(String w) {
/* 将字符串w添加进Trie中存储
*/
if (root.getNext() == null) {
root.setNext(w.charAt(0));
}
addHelper(root.getNext(), w, 0);
}
private void addHelper(BST curLevel, String w, int curChar) {
char c = w.charAt(curChar);
curChar += 1;
if (!curLevel.contain(c)){
curLevel.add(c);
}
Node nextNode = curLevel.get(c);
if (curChar >= w.length()) {
nextNode.setEnd();
return;
}
if (nextNode.getNext() == null) {
nextNode.setNext(w.charAt(curChar));
}
addHelper(nextNode.getNext(), w, curChar);
}
public List<String> allString() {
/* 返回Trie中存储的所有字符串
*/
List<String> res = new ArrayList<>();
addStringHelper(root.getNodes(), res, "");
return res;
}
private void addStringHelper(List<Node> nodes, List<String> res, String prev) {
if (nodes == null) {
return;
} else {
for (Node node : nodes) {
char c = node.getKey();
if (node.getIsEnd()) {
res.add(prev + c);
}
addStringHelper(node.getNodes(), res, prev + c);
}
}
}
public String longestPrefixOf(String w) {
/* 返回Trie中和w匹配的最长的前缀
*/
return longestHelper(root.getNodes(), w, 0, "");
}
private String longestHelper(List<Node> nodes, String w, int curChar, String res) {
if (nodes != null) {
for (Node node : nodes) {
char c = node.getKey();
if (c == w.charAt(curChar)) {
return longestHelper(node.getNodes(), w, curChar + 1, res + c);
}
}
}
return res;
}
public List<String> stringsWithPrefix(String prefix) {
/* 返回所有带有前缀prefix的字符串
*/
List<String> res = new ArrayList<>();
stringPrefixHelper(root.getNodes(), prefix, 0, prefix, res);
return res;
}
private void stringPrefixHelper(List<Node> nodes, String w, int curChar, String prev, List<String> res) {
if (nodes == null) {
return;
} else {
for (Node node : nodes) {
/* 有两种情况
* 此时curChar没有到w的末端,继续在同样的node中search
* curChar到达了w的末端,node也是字符串的结尾,将其放入res中
*/
if (curChar < w.length()) {
if (node.getKey() == w.charAt(curChar)) {
stringPrefixHelper(node.getNodes(), w, curChar + 1, prev, res);
}
} else {
String newPrev = prev + node.getKey();
if (node.getIsEnd()) {
res.add(newPrev);
stringPrefixHelper(node.getNodes(), w, curChar, newPrev, res);
}
stringPrefixHelper(node.getNodes(), w, curChar, newPrev, res);
}
}
}
}
}
Node.java
import java.util.List;
public class Node {
private char key;
private boolean isEnd;
private BST next; // 使用BST存储child
Node(char key, BST next) {
this.key = key;
this.next = next;
this.isEnd = false;
}
public char getKey() {
return key;
}
public BST getNext() {
return next;
}
public List<Node> getNodes() {
if (next == null) {
return null;
} else {
return next.getAllNodes();
}
}
public boolean getIsEnd() {
return this.isEnd;
}
public void setEnd() {
this.isEnd = true;
}
public void setNext(char c) {
next = new BST(c);
}
}
BST.java
import java.util.ArrayList;
import java.util.List;
public class BST {
private BSTNode root;
BST(char key) {
root = new BSTNode(key);
}
private class BSTNode {
private char key;
private Node childNode;
private BSTNode left;
private BSTNode right;
BSTNode(char key) {
this.key = key;
this.childNode = new Node(key, null);
this.left = null;
this.right = null;
}
}
public List<Node> getAllNodes() {
/* 返回BST中存储的所有Nodes
*/
List<Node> res = new ArrayList<>();
getAllNodesHelper(root, res);
return res;
}
private void getAllNodesHelper(BSTNode curNode, List<Node> res) {
if (curNode == null) {
return;
} else {
res.add(curNode.childNode);
getAllNodesHelper(curNode.left, res);
getAllNodesHelper(curNode.right, res);
}
}
public Node get(char c) {
/* 返回当前BST中和c具有相同key的BSTNode存储的节点;
*/
return getHelper(root, c);
}
private Node getHelper(BSTNode curNode, char c) {
if (curNode == null) {
return null;
} else if (curNode.key == c) {
return curNode.childNode;
} else if (curNode.key < c) {
return getHelper(curNode.right, c);
} else {
return getHelper(curNode.left, c);
}
}
public void add(char c) {
/* 将c添加进BST中
*/
root = addHelper(root, c);
}
private BSTNode addHelper(BSTNode curNode, char c) {
if (curNode == null) {
return new BSTNode(c);
} else if (curNode.key < c) {
curNode.right = addHelper(curNode.right, c);
} else {
curNode.left = addHelper(curNode.left, c);
}
return curNode;
}
public boolean contain(char c) {
/* BST中是否存在和c相同key的节点
*/
return containHelper(root, c);
}
private boolean containHelper(BSTNode curNode, char c) {
if (curNode == null) {
return false;
} else if (curNode.key == c) {
return true;
} else if (curNode.key < c) {
return containHelper(curNode.right, c);
} else {
return containHelper(curNode.left, c);
}
}
}
Test.java
public class Test {
public static void main(String[] args) {
/* 对Trie进行测试,测试样例有限,所以构造的Trie也可能存在问题
*/
Trie t = new Trie();
t.add("same");
t.add("saple");
t.add("apple");
t.add("test");
t.add("saddddd");
System.out.println(t.allString());
System.out.println(t.longestPrefixOf("sample"));
System.out.println(t.stringsWithPrefix("s"));
}
}
欢迎分享,转载请注明来源:内存溢出
评论列表(0条)