Java 구현 AC 로봇 전체 텍스트 검색 예
/**
* Created by zhaoyy on 2017/2/7.
*/
interface Node {
char value();
boolean exists();
boolean isRoot();
Node parent();
Node childOf(char c);
Node fail();
void setFail(Node node);
void setExists(boolean exists);
void add(Node child);
List<Node> children();
}
두 번째 단계는 두 가지 Node를 실현한다. 만약에 어휘가 모두 인쇄 가능한 ASCII 문자라면 AsciiNode를 사용하고, 그렇지 않으면 (예를 들어 한자 포함)hash표를 바탕으로 하는 MapNode를 사용한다.이 두 노드는 모두 AbstractNode에서 통합됩니다.
/**
* Created by zhaoyy on 2017/2/8.
*/
abstract class AbstractNode implements Node {
private static final char EMPTY = '\0';
private final char value;
private final Node parent;
private boolean exists;
private Node fail;
public AbstractNode(Node parent, char value) {
this.parent = parent;
this.value = value;
this.exists = false;
this.fail = null;
}
public AbstractNode() {
this(null, EMPTY);
}
private static String tab(int n) {
StringBuilder builder = new StringBuilder();
for (int i = 0; i < n; i++) {
builder.append('\t');
}
return builder.toString();
}
private static String toString(Node node, int depth) {
StringBuilder builder = new StringBuilder();
String tab = tab(depth);
Node fail = node.fail();
Node parent = node.parent();
builder
.append(tab)
.append('<')
.append(node.value())
.append(" exists=\"")
.append(node.exists())
.append('"')
.append(" parent=\"")
.append(parent == null ? "null" : parent.value())
.append('"')
.append(" fail=\"")
.append(fail == null ? "null" : fail.value())
.append("\">\r
");
for (Node child : node.children())
builder.append(toString(child, depth + 1));
builder
.append(tab)
.append("</")
.append(node.value())
.append(">\r
");
return builder.toString();
}
@Override
public char value() {
return value;
}
@Override
public boolean exists() {
return exists;
}
@Override
public boolean isRoot() {
return value == EMPTY;
}
@Override
public Node parent() {
return parent;
}
@Override
public Node fail() {
return fail;
}
@Override
public void setFail(Node node) {
this.fail = node;
}
@Override
public void setExists(boolean exists) {
this.exists = exists;
}
@Override
public String toString() {
return toString(this, 0);
}
}
/////////////////////////////////////////////////////////////////////////////////////////////
/**
* Created by zhaoyy on 2017/2/8.
*/
final class AsciiNode extends AbstractNode implements Node {
private static final char FROM = 32;
private static final char TO = 126;
private final Node[] children;
public AsciiNode() {
super();
this.children = new Node[TO - FROM + 1];
}
public AsciiNode(Node parent, char value) {
super(parent, value);
this.children = new Node[TO - FROM + 1];
}
@Override
public Node childOf(char c) {
if (c >= FROM && c <= TO)
return children[(int) c - FROM];
else return null;
}
@Override
public void add(Node child) {
int index = (int) child.value();
children[index - FROM] = child;
}
@Override
public List<Node> children() {
List<Node> nodes = new ArrayList<Node>();
for (Node child : children)
if (child != null)
nodes.add(child);
return nodes;
}
}
//////////////////////////////////////////////////////////////////////////////////////////////
/**
* Created by zhaoyy on 2017/2/8.
*/
final class MapNode extends AbstractNode implements Node {
private final Map<Character, Node> children;
public MapNode() {
super();
this.children = new HashMap<Character, Node>();
}
public MapNode(Node parent, char value) {
super(parent, value);
this.children = new HashMap<Character, Node>();
}
@Override
public Node childOf(char c) {
return children.get(c);
}
@Override
public void add(Node child) {
children.put(child.value(), child);
}
@Override
public List<Node> children() {
List<Node> nodes = new ArrayList<Node>();
nodes.addAll(children.values());
return nodes;
}
}
세 번째 단계,먼저 Node 구조기를 정의합니다.
/**
* Created by zhaoyy on 2017/2/8.
*/
public interface NodeMaker {
Node make(Node parent, char value);
Node makeRoot();
}
그리고 AC 로봇을 구축하여 생성 및 검색 방법을 실현합니다.
/**
* Created by zhaoyy on 2017/2/7.
*/
public final class WordTable {
private final Node root;
private WordTable(Collection<? extends CharSequence> words, NodeMaker maker) {
Node root = buildTrie(words, maker);
setFailNode(root);
this.root = root;
}
public static WordTable compile(Collection<? extends CharSequence> words) {
if (words == null || words.isEmpty())
throw new IllegalArgumentException();
final NodeMaker maker;
if (isAscii(words))
maker = new NodeMaker() {
@Override
public Node make(Node parent, char value) {
return new AsciiNode(parent, value);
}
@Override
public Node makeRoot() {
return new AsciiNode();
}
};
else maker = new NodeMaker() {
@Override
public Node make(Node parent, char value) {
return new MapNode(parent, value);
}
@Override
public Node makeRoot() {
return new MapNode();
}
};
return new WordTable(words, maker);
}
private static boolean isAscii(Collection<? extends CharSequence> words) {
for (CharSequence word : words) {
int len = word.length();
for (int i = 0; i < len; i++) {
int c = (int) word.charAt(i);
if (c < 32 || c > 126)
return false;
}
}
return true;
}
private static Node buildTrie(Collection<? extends CharSequence> sequences, NodeMaker maker) {
Node root = maker.makeRoot();
for (CharSequence sequence : sequences) {
int len = sequence.length();
Node current = root;
for (int i = 0; i < len; i++) {
char c = sequence.charAt(i);
Node node = current.childOf(c);
if (node == null) {
node = maker.make(current, c);
current.add(node);
}
current = node;
if (i == len - 1)
node.setExists(true);
}
}
return root;
}
private static void setFailNode(final Node root) {
root.setFail(null);
Queue<Node> queue = new LinkedList<Node>();
queue.add(root);
while (!queue.isEmpty()) {
Node parent = queue.poll();
Node temp;
for (Node child : parent.children()) {
if (parent.isRoot())
child.setFail(root);
else {
temp = parent.fail();
while (temp != null) {
Node node = temp.childOf(child.value());
if (node != null) {
child.setFail(node);
break;
}
temp = temp.fail();
}
if (temp == null)
child.setFail(root);
}
queue.add(child);
}
}
}
public boolean findAnyIn(CharSequence cs) {
int len = cs.length();
Node node = root;
for (int i = 0; i < len; i++) {
Node next = node.childOf(cs.charAt(i));
if (next == null) {
next = node.fail();
if (next == null) {
node = root;
continue;
}
}
if (next.exists())
return true;
}
return false;
}
public List<MatchInfo> search(CharSequence cs) {
if (cs == null || cs.length() == 0)
return Collections.emptyList();
List<MatchInfo> result = new ArrayList<MatchInfo>();
int len = cs.length();
Node node = root;
for (int i = 0; i < len; i++) {
Node next = node.childOf(cs.charAt(i));
if (next == null) {
next = node.fail();
if (next == null) {
node = root;
continue;
}
}
if (next.exists()) {
MatchInfo info = new MatchInfo(i, next);
result.add(info);
node = root;
continue;
}
node = next;
}
return result;
}
@Override
public String toString() {
return root.toString();
}
}
검색 결과를 저장할 엔티티를 정의합니다.
/**
* Created by zhaoyy on 2017/2/7.
*/
public final class MatchInfo {
private final int index;
private final String word;
public MatchInfo(int index, String word) {
this.index = index;
this.word = word;
}
public MatchInfo(int index, Node node) {
StringBuilder builder = new StringBuilder();
while (node != null) {
if (!node.isRoot())
builder.append(node.value());
node = node.parent();
}
String word = builder.reverse().toString();
this.index = index + 1 - word.length();
this.word = word;
}
public int getIndex() {
return index;
}
public String getWord() {
return word;
}
@Override
public String toString() {
return index + ":" + word;
}
}
4단계, Demo 호출:
public static void main(String[] args) {
List<String> list = Arrays.asList("say", "her", "he", "she", "shr", "alone");
WordTable table = WordTable.compile(list);
System.out.println(table);
System.out.println(table.search("1shesaynothingabouthislivinghimalone"));
}
다음은 출력 결과입니다.
< exists="false" parent="null" fail="null">
<s exists="false" parent=" " fail=" ">
<a exists="false" parent="s" fail="a">
<y exists="true" parent="a" fail=" ">
</y>
</a>
<h exists="false" parent="s" fail="h">
<e exists="true" parent="h" fail="e">
</e>
<r exists="true" parent="h" fail=" ">
</r>
</h>
</s>
<h exists="false" parent=" " fail=" ">
<e exists="true" parent="h" fail=" ">
<r exists="true" parent="e" fail=" ">
</r>
</e>
</h>
<a exists="false" parent=" " fail=" ">
<l exists="false" parent="a" fail=" ">
<o exists="false" parent="l" fail=" ">
<n exists="false" parent="o" fail=" ">
<e exists="true" parent="n" fail=" ">
</e>
</n>
</o>
</l>
</a>
</ >
[1:she, 4:say, 31:alone]
이상은 본문의 전체 내용입니다. 여러분의 학습에 도움이 되고 저희를 많이 응원해 주십시오.
이 내용에 흥미가 있습니까?
현재 기사가 여러분의 문제를 해결하지 못하는 경우 AI 엔진은 머신러닝 분석(스마트 모델이 방금 만들어져 부정확한 경우가 있을 수 있음)을 통해 가장 유사한 기사를 추천합니다:
38. Java의 Leetcode 솔루션텍스트를 자유롭게 공유하거나 복사할 수 있습니다.하지만 이 문서의 URL은 참조 URL로 남겨 두십시오.
CC BY-SA 2.5, CC BY-SA 3.0 및 CC BY-SA 4.0에 따라 라이센스가 부여됩니다.