mirror of
https://gitee.com/chinabugotech/hutool.git
synced 2025-04-19 03:01:48 +08:00
AC自动机实现
This commit is contained in:
Looly
parent
f78f9569e6
commit
1578a46761
@ -43,10 +43,7 @@ import java.util.HashSet;
|
||||
import java.util.LinkedList;
|
||||
import java.util.List;
|
||||
import java.util.Set;
|
||||
import java.util.function.Function;
|
||||
import java.util.function.Predicate;
|
||||
import java.util.function.Supplier;
|
||||
import java.util.function.UnaryOperator;
|
||||
import java.util.function.*;
|
||||
import java.util.regex.Matcher;
|
||||
import java.util.regex.Pattern;
|
||||
|
||||
@ -1476,9 +1473,9 @@ public class CharSequenceUtil extends StrValidator {
|
||||
* }
|
||||
* </pre>
|
||||
*
|
||||
* @param str 被处理的字符串
|
||||
* @param prefix 前缀
|
||||
* @param suffix 后缀
|
||||
* @param str 被处理的字符串
|
||||
* @param prefix 前缀
|
||||
* @param suffix 后缀
|
||||
* @param ignoreCase 是否忽略大小写
|
||||
* @return 处理后的字符串
|
||||
* @since 3.1.2
|
||||
@ -1494,17 +1491,17 @@ public class CharSequenceUtil extends StrValidator {
|
||||
|
||||
if (startWith(str2, prefix, ignoreCase)) {
|
||||
from = prefix.length();
|
||||
if(from == to){
|
||||
if (from == to) {
|
||||
// "a", "a", "a" -> ""
|
||||
return EMPTY;
|
||||
}
|
||||
}
|
||||
if (endWith(str2, suffix, ignoreCase)) {
|
||||
to -= suffix.length();
|
||||
if(from == to){
|
||||
if (from == to) {
|
||||
// "a", "a", "a" -> ""
|
||||
return EMPTY;
|
||||
} else if(to < from){
|
||||
} else if (to < from) {
|
||||
// pre去除后和suffix有重叠，如 ("aba", "ab", "ba") -> "a"
|
||||
to += suffix.length();
|
||||
}
|
||||
@ -1574,22 +1571,22 @@ public class CharSequenceUtil extends StrValidator {
|
||||
int from = 0;
|
||||
int to = str2.length();
|
||||
|
||||
if(!prefixStr.isEmpty()){
|
||||
if (!prefixStr.isEmpty()) {
|
||||
while (str2.startsWith(prefixStr, from)) {
|
||||
from += prefix.length();
|
||||
if(from == to){
|
||||
if (from == to) {
|
||||
// "a", "a", "a" -> ""
|
||||
return EMPTY;
|
||||
}
|
||||
}
|
||||
}
|
||||
if(!suffixStr.isEmpty()){
|
||||
if (!suffixStr.isEmpty()) {
|
||||
while (str2.startsWith(suffixStr, to - suffixStr.length())) {
|
||||
to -= suffixStr.length();
|
||||
if(from == to){
|
||||
if (from == to) {
|
||||
// "a", "a", "a" -> ""
|
||||
return EMPTY;
|
||||
}else if(to < from){
|
||||
} else if (to < from) {
|
||||
// pre去除后和suffix有重叠，如 ("aba", "ab", "ba") -> "a"
|
||||
to += suffixStr.length();
|
||||
break;
|
||||
@ -1730,7 +1727,7 @@ public class CharSequenceUtil extends StrValidator {
|
||||
|
||||
final StringBuilder sb = new StringBuilder();
|
||||
final int subLen = toIndex - fromIndex;
|
||||
str.toString().codePoints().skip(fromIndex).limit(subLen).forEach(v -> sb.append(Character.toChars(v)));
|
||||
str.codePoints().skip(fromIndex).limit(subLen).forEach(v -> sb.append(Character.toChars(v)));
|
||||
return sb.toString();
|
||||
}
|
||||
|
||||
@ -4131,4 +4128,28 @@ public class CharSequenceUtil extends StrValidator {
|
||||
}
|
||||
return (isCodePoint ? str.codePoints() : str.chars()).toArray();
|
||||
}
|
||||
|
||||
/**
|
||||
* 遍历字符串的每个字符，并处理
|
||||
*
|
||||
* @param str 字符串
|
||||
* @param consumer 字符处理
|
||||
*/
|
||||
public static void forEach(final CharSequence str, final Consumer<Character> consumer) {
|
||||
forEach(str, false, (cInt)-> consumer.accept((char) cInt));
|
||||
}
|
||||
|
||||
/**
|
||||
* 遍历字符串的每个字符，并处理
|
||||
*
|
||||
* @param str 字符串
|
||||
* @param isCodePoint 是否为Unicode码点（即支持emoji等多char字符）
|
||||
* @param consumer 字符处理
|
||||
*/
|
||||
public static void forEach(final CharSequence str, final boolean isCodePoint, final IntConsumer consumer) {
|
||||
if (null == str) {
|
||||
return;
|
||||
}
|
||||
(isCodePoint ? str.codePoints() : str.chars()).forEach(consumer);
|
||||
}
|
||||
}
|
||||
|
||||
@ -1,51 +1,56 @@
|
||||
package org.dromara.hutool.core.text.finder;
|
||||
|
||||
import org.dromara.hutool.core.text.StrUtil;
|
||||
|
||||
import java.util.*;
|
||||
|
||||
/**
|
||||
* 多字符串查询器 底层思路 使用 AC 自动机实现
|
||||
*
|
||||
* @author newshiJ
|
||||
* @date 2024/8/2 上午10:07
|
||||
*/
|
||||
public class MultiStrFinder {
|
||||
|
||||
// 字符索引
|
||||
protected final Map<Character,Integer> charIndex = new HashMap<>();
|
||||
/**
|
||||
* 创建多字符串查询器
|
||||
* @param source 字符串集合
|
||||
* @return 多字符串查询器
|
||||
*/
|
||||
public static MultiStrFinder of(final Collection<String> source) {
|
||||
return new MultiStrFinder(source);
|
||||
}
|
||||
|
||||
// 字符索引
|
||||
protected final Map<Character, Integer> charIndexMap = new HashMap<>();
|
||||
// 全部字符数量
|
||||
protected final int allCharSize;
|
||||
|
||||
// 根节点
|
||||
protected final Node root;
|
||||
|
||||
// 全部节点数量
|
||||
int nodeSize;
|
||||
|
||||
/**
|
||||
* 构建多字符串查询器
|
||||
* @param source
|
||||
*
|
||||
* @param source 字符串集合
|
||||
*/
|
||||
public MultiStrFinder(Collection<String> source){
|
||||
public MultiStrFinder(final Collection<String> source) {
|
||||
// 待匹配的字符串
|
||||
final Set<String> stringSet = new HashSet<>();
|
||||
|
||||
// 所有字符
|
||||
final Set<Character> charSet = new HashSet<>();
|
||||
for (String string : source) {
|
||||
for (final String string : source) {
|
||||
stringSet.add(string);
|
||||
char[] charArray = string.toCharArray();
|
||||
for (char c : charArray) {
|
||||
charSet.add(c);
|
||||
}
|
||||
StrUtil.forEach(string, charSet::add);
|
||||
}
|
||||
allCharSize = charSet.size();
|
||||
int index = 0;
|
||||
for (Character c : charSet) {
|
||||
charIndex.put(c,index);
|
||||
for (final Character c : charSet) {
|
||||
charIndexMap.put(c,index);
|
||||
index ++;
|
||||
}
|
||||
|
||||
root = Node.createRoot(allCharSize);
|
||||
this.root = Node.createRoot(index);
|
||||
|
||||
buildPrefixTree(stringSet);
|
||||
buildFail();
|
||||
@ -53,19 +58,18 @@ public class MultiStrFinder {
|
||||
|
||||
/**
|
||||
* 构建前缀树
|
||||
*
|
||||
* @param stringSst 待匹配的字符串
|
||||
*/
|
||||
protected void buildPrefixTree(Collection<String> stringSst){
|
||||
protected void buildPrefixTree(final Collection<String> stringSst) {
|
||||
// 节点编号 根节点已经是0了 所以从 1开始编号
|
||||
int nodeIndex = 1;
|
||||
for (String string : stringSst) {
|
||||
for (final String string : stringSst) {
|
||||
Node node = root;
|
||||
char[] charArray = string.toCharArray();
|
||||
for (int i = 0; i < charArray.length; i++) {
|
||||
char c = charArray[i];
|
||||
boolean addValue = node.addValue(c, nodeIndex, charIndex);
|
||||
if(addValue){
|
||||
nodeIndex ++;
|
||||
for (final char c : string.toCharArray()) {
|
||||
final boolean addValue = node.addValue(c, nodeIndex, charIndexMap);
|
||||
if (addValue) {
|
||||
nodeIndex++;
|
||||
}
|
||||
node = node.directRouter[getIndex(c)];
|
||||
}
|
||||
@ -78,11 +82,11 @@ public class MultiStrFinder {
|
||||
* 构建 fail指针过程
|
||||
* 构建 directRouter 直接访问路由表 减少跳fail次数 直接跳 router 边
|
||||
*/
|
||||
protected void buildFail(){
|
||||
LinkedList<Node> nodeQueue = new LinkedList<>();
|
||||
protected void buildFail() {
|
||||
final LinkedList<Node> nodeQueue = new LinkedList<>();
|
||||
for (int i = 0; i < root.directRouter.length; i++) {
|
||||
Node nextNode = root.directRouter[i];
|
||||
if(nextNode == null){
|
||||
final Node nextNode = root.directRouter[i];
|
||||
if (nextNode == null) {
|
||||
root.directRouter[i] = root;
|
||||
continue;
|
||||
}
|
||||
@ -91,13 +95,13 @@ public class MultiStrFinder {
|
||||
}
|
||||
|
||||
// 进行广度优先遍历
|
||||
while (!nodeQueue.isEmpty()){
|
||||
Node parent = nodeQueue.removeFirst();
|
||||
while (!nodeQueue.isEmpty()) {
|
||||
final Node parent = nodeQueue.removeFirst();
|
||||
// 因为 使用了 charIndex 进行字符到下标的映射 i 可以直接认为就是对应字符 char
|
||||
for (int i = 0; i < parent.directRouter.length; i++) {
|
||||
Node child = parent.directRouter[i];
|
||||
final Node child = parent.directRouter[i];
|
||||
// child 为 null 表示没有子节点
|
||||
if(child == null){
|
||||
if (child == null) {
|
||||
parent.directRouter[i] = parent.fail.directRouter[i];
|
||||
continue;
|
||||
}
|
||||
@ -110,27 +114,28 @@ public class MultiStrFinder {
|
||||
|
||||
/**
|
||||
* 查询匹配的字符串
|
||||
*
|
||||
* @param text 返回每个匹配的 字符串 value是字符首字母地址
|
||||
* @return
|
||||
* @return 匹配结果
|
||||
*/
|
||||
public Map<String,List<Integer>> findMatch(String text){
|
||||
public Map<String, List<Integer>> findMatch(final String text) {
|
||||
// 节点经过次数 放在方法内部声明变量 希望可以一个构建对象 进行多次匹配
|
||||
HashMap<String, List<Integer>> resultMap = new HashMap<>();
|
||||
final HashMap<String, List<Integer>> resultMap = new HashMap<>();
|
||||
|
||||
char[] chars = text.toCharArray();
|
||||
final char[] chars = text.toCharArray();
|
||||
Node currentNode = root;
|
||||
for (int i = 0; i < chars.length; i++) {
|
||||
char c = chars[i];
|
||||
Integer index = charIndex.get(c);
|
||||
final char c = chars[i];
|
||||
final Integer index = charIndexMap.get(c);
|
||||
// 找不到字符索引 认为一定不在匹配字符中存在 直接从根节点开始重新计算
|
||||
if(index == null){
|
||||
if (index == null) {
|
||||
currentNode = root;
|
||||
continue;
|
||||
}
|
||||
// 进入下一跳 可能是正常下一跳 也可能是fail加上后的 下一跳
|
||||
currentNode = currentNode.directRouter[index];
|
||||
// 判断是否尾部节点 是尾节点 说明已经匹配到了完整的字符串 将匹配结果写入返回对象
|
||||
if(currentNode.isEnd){
|
||||
if (currentNode.isEnd) {
|
||||
resultMap.computeIfAbsent(currentNode.tagetString, k -> new ArrayList<>())
|
||||
.add(i - currentNode.tagetString.length() + 1);
|
||||
}
|
||||
@ -143,22 +148,19 @@ public class MultiStrFinder {
|
||||
|
||||
/**
|
||||
* 获取字符 下标
|
||||
* @param c
|
||||
* @return
|
||||
*
|
||||
* @param c 字符
|
||||
* @return 下标
|
||||
*/
|
||||
protected int getIndex(char c){
|
||||
Integer i = charIndex.get(c);
|
||||
if(i == null){
|
||||
protected int getIndex(final char c) {
|
||||
final Integer i = charIndexMap.get(c);
|
||||
if (i == null) {
|
||||
return -1;
|
||||
}
|
||||
return i;
|
||||
}
|
||||
|
||||
|
||||
public static MultiStrFinder create(Collection<String> source){
|
||||
return new MultiStrFinder(source);
|
||||
}
|
||||
|
||||
/**
|
||||
* AC 自动机节点
|
||||
*/
|
||||
@ -189,19 +191,21 @@ public class MultiStrFinder {
|
||||
// fail指针来源
|
||||
public List<Node> failPre = new ArrayList<>();
|
||||
|
||||
public Node(){}
|
||||
public Node() {
|
||||
}
|
||||
|
||||
/**
|
||||
* 新增子节点
|
||||
* @param c 字符
|
||||
*
|
||||
* @param c 字符
|
||||
* @param nodeIndex 节点编号
|
||||
* @param charIndex 字符索引
|
||||
* @return 如果已经存在子节点 false 新增 ture
|
||||
*/
|
||||
public boolean addValue(char c, int nodeIndex ,Map<Character,Integer> charIndex){
|
||||
Integer index = charIndex.get(c);
|
||||
public boolean addValue(final char c, final int nodeIndex, final Map<Character, Integer> charIndex) {
|
||||
final Integer index = charIndex.get(c);
|
||||
Node node = directRouter[index];
|
||||
if(node != null){
|
||||
if (node != null) {
|
||||
return false;
|
||||
}
|
||||
node = new Node();
|
||||
@ -214,22 +218,24 @@ public class MultiStrFinder {
|
||||
|
||||
/**
|
||||
* 标记当前节点为 字符串尾节点
|
||||
*
|
||||
* @param string
|
||||
*/
|
||||
public void setEnd(String string){
|
||||
public void setEnd(final String string) {
|
||||
tagetString = string;
|
||||
isEnd = true;
|
||||
}
|
||||
|
||||
/**
|
||||
* 获取下一跳
|
||||
* @param c 字符
|
||||
*
|
||||
* @param c 字符
|
||||
* @param charIndex 字符索引
|
||||
* @return
|
||||
*/
|
||||
public Node getNext(char c,Map<Character,Integer> charIndex){
|
||||
Integer index = charIndex.get(c);
|
||||
if(index == null){
|
||||
public Node getNext(final char c, final Map<Character, Integer> charIndex) {
|
||||
final Integer index = charIndex.get(c);
|
||||
if (index == null) {
|
||||
return null;
|
||||
}
|
||||
return directRouter[index];
|
||||
@ -237,11 +243,12 @@ public class MultiStrFinder {
|
||||
|
||||
/**
|
||||
* 构建根节点
|
||||
*
|
||||
* @param allCharSize 全部字符数量
|
||||
* @return
|
||||
*/
|
||||
public static Node createRoot(int allCharSize){
|
||||
Node node = new Node();
|
||||
public static Node createRoot(final int allCharSize) {
|
||||
final Node node = new Node();
|
||||
node.nodeIndex = 0;
|
||||
node.fail = node;
|
||||
node.directRouter = new Node[allCharSize];
|
||||
|
||||
@ -14,9 +14,9 @@ import java.util.*;
|
||||
* 3、"abc", "bc" 会优先替换"abc"
|
||||
*
|
||||
* @author newshiJ
|
||||
* @date 2024/8/2 下午3:41
|
||||
*/
|
||||
public class HighMultiReplacerV2 extends StrReplacer {
|
||||
private static final long serialVersionUID = 1L;
|
||||
|
||||
private final AhoCorasickAutomaton ahoCorasickAutomaton;
|
||||
|
||||
@ -49,19 +49,17 @@ public class HighMultiReplacerV2 extends StrReplacer {
|
||||
protected static class AhoCorasickAutomaton extends MultiStrFinder{
|
||||
protected final Map<String,String> replaceMap;
|
||||
|
||||
public AhoCorasickAutomaton(Map<String,String> replaceMap){
|
||||
public AhoCorasickAutomaton(final Map<String,String> replaceMap){
|
||||
super(replaceMap.keySet());
|
||||
this.replaceMap = replaceMap;
|
||||
}
|
||||
|
||||
|
||||
public void replace(final CharSequence text, final StringBuilder stringBuilder){
|
||||
Node currentNode = root;
|
||||
// 临时字符串存储空间
|
||||
StringBuilder temp = new StringBuilder();
|
||||
final StringBuilder temp = new StringBuilder();
|
||||
for (int i = 0; i < text.length(); i++) {
|
||||
char ch = text.charAt(i);
|
||||
Integer index = charIndex.get(ch);
|
||||
final char ch = text.charAt(i);
|
||||
final Integer index = charIndexMap.get(ch);
|
||||
// 下一个字符在候选转换字符串中都不存在 ch字符一定不会被替换
|
||||
if(index < 0){
|
||||
// 临时缓存空间中的数据写入到输出的 StringBuilder
|
||||
@ -94,7 +92,7 @@ public class HighMultiReplacerV2 extends StrReplacer {
|
||||
|
||||
// 表示匹配到 现在进行字符串替换工作
|
||||
if(currentNode.isEnd){
|
||||
int length = currentNode.tagetString.length();
|
||||
final int length = currentNode.tagetString.length();
|
||||
// 先清理匹配到的字符 最后一个字符未加入临时空间
|
||||
temp.delete(temp.length() - length + 1,length - 1);
|
||||
if(temp.length() > 0){
|
||||
|
||||
Loading…
x
Reference in New Issue
Block a user