github/sensitive-word
1
0
Fork 0
mirror of https://github.com/houbb/sensitive-word.git synced 2026-03-22 00:17:35 +08:00
Code Issues Packages Projects Releases Wiki Activity

黑白名单共同检测时逻辑的修改

Browse Source
This commit is contained in:
yds
2025-05-02 23:31:43 +08:00
parent 91e811b360
commit 6132261e77
7 changed files with 233 additions and 73 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
src/main/java/com/github/houbb/sensitive/word/api/ISensitiveWord.java
View File

@@ -19,7 +19,7 @@ public interface ISensitiveWord {
* @see WordValidModeEnum#FAIL_OVER 建议使用全部检测返回模式
*/
List<IWordResult> findAll(final String string,
final IWordContext context);
final IWordContext context);
/**
* 返回第一个对应的敏感词
@@ -29,22 +29,20 @@ public interface ISensitiveWord {
* @since 0.3.2
*/
IWordResult findFirst(final String string,
final IWordContext context);
final IWordContext context);
/**
* 替换所有敏感词内容
* <p>
*
* ps: 这里可以添加优化。
*
* @param target 目标字符串
* @param target 目标字符串
* @param context 上下文
* @param replace 替换策略
* @return 替换后结果
* @since 0.3.2
*/
String replace(final String target,
final IWordContext context,
final IWordReplace replace);
final IWordContext context);
/**
* 包含
@@ -54,6 +52,6 @@ public interface ISensitiveWord {
* @since 0.3.2
*/
boolean contains(final String string,
final IWordContext context);
final IWordContext context);
}

6
src/main/java/com/github/houbb/sensitive/word/bs/SensitiveWordBs.java
View File

@@ -672,12 +672,10 @@ public class SensitiveWordBs implements ISensitiveWordDestroy {
* @since 0.2.0
*/
public String replace(final String target) {
return this.replace(target,context.wordReplace());
}
public String replace(final String target, IWordReplace replace) {
return sensitiveWord.replace(target, context, replace);
return sensitiveWord.replace(target, context);
}
/**
* 获取敏感词的标签
*

8
src/main/java/com/github/houbb/sensitive/word/core/AbstractSensitiveWord.java
View File

@@ -38,8 +38,9 @@ public abstract class AbstractSensitiveWord implements ISensitiveWord {
* @return 结果
* @since 0.3.2
*/
protected String doReplace(String target, List<IWordResult> allList, IWordContext context, IWordReplace replace) {
protected String doReplace(String target, List<IWordResult> allList, IWordContext context) {
// 根据 index 直接分割
final IWordReplace replace = context.wordReplace();
// 是否需要对 allList 排序？
StringBuilder stringBuilder = new StringBuilder();
@@ -90,7 +91,7 @@ public abstract class AbstractSensitiveWord implements ISensitiveWord {
}
@Override
public String replace(String target, IWordContext context, IWordReplace replace) {
public String replace(String target, IWordContext context) {
if(StringUtil.isEmpty(target)) {
return target;
}
@@ -100,9 +101,8 @@ public abstract class AbstractSensitiveWord implements ISensitiveWord {
return target;
}
return doReplace(target, allList, context, replace);
return doReplace(target, allList, context);
}
@Override
public boolean contains(String string, IWordContext context) {
//1. 第一个存在

34
src/main/java/com/github/houbb/sensitive/word/core/SensitiveWord.java
View File

@@ -37,7 +37,7 @@ public class SensitiveWord extends AbstractSensitiveWord {
@Override
protected IWordResult doFindFirst(String string, IWordContext context) {
List<IWordResult> wordResults = innerSensitiveWords(string, WordValidModeEnum.FAIL_FAST, context);
if(!CollectionUtil.isEmpty(wordResults)){
if (!CollectionUtil.isEmpty(wordResults)) {
return wordResults.get(0);
}
return null;
@@ -47,14 +47,14 @@ public class SensitiveWord extends AbstractSensitiveWord {
/**
* 获取敏感词列表
*
* @param text 文本
* @param text 文本
* @param modeEnum 模式
* @return 结果列表
* @since 0.0.1
*/
private List<IWordResult> innerSensitiveWords(final String text,
final WordValidModeEnum modeEnum,
final IWordContext context) {
final WordValidModeEnum modeEnum,
final IWordContext context) {
//1. 是否存在敏感词，如果比存在，直接返回空列表
final IWordCheck sensitiveCheck = context.sensitiveCheck();
List<IWordResult> resultList = Guavas.newArrayList();
@@ -74,38 +74,32 @@ public class SensitiveWord extends AbstractSensitiveWord {
// v0.21.0 白名单跳过
WordCheckResult checkResult = sensitiveCheck.sensitiveCheck(i, checkContext);
int wordLengthAllow = checkResult.wordLengthResult().wordAllowLen();
if(wordLengthAllow > 0) {
i += wordLengthAllow-1;
continue;
}
int wordLengthDeny = checkResult.wordLengthResult().wordDenyLen();
// 命中
final WordLengthResult wordLengthResult = checkResult.wordLengthResult();
int wordLength = wordLengthResult.wordDenyLen();
if (wordLength > 0) {
//如果命中的白名单长度小于黑名单，则直接对黑名单的敏感词进行保存
if (wordLengthAllow < wordLengthDeny) {
// 保存敏感词
WordResult wordResult = WordResult.newInstance()
.startIndex(i)
.endIndex(i+wordLength)
.endIndex(i + wordLengthDeny)
.type(checkResult.type())
.word(wordLengthResult.wordDeny());
.word(checkResult.wordLengthResult().wordDeny());
//v0.13.0 添加判断
if(wordResultCondition.match(wordResult, text, modeEnum, context)) {
if (wordResultCondition.match(wordResult, text, modeEnum, context)) {
resultList.add(wordResult);
// 快速返回
if (WordValidModeEnum.FAIL_FAST.equals(modeEnum)) {
break;
}
}
// 增加 i 的步长
// 为什么要-1，因为默认就会自增1
// TODO: 这里可以根据字符串匹配算法优化。
i += wordLength - 1;
i += wordLengthDeny - 1;
} else {
//如果命中的白名单长度大于黑名单长度，则跳过白名单个字符
i += Math.max(0, wordLengthAllow - 1);
}
}

24
src/main/java/com/github/houbb/sensitive/word/support/check/WordCheckWord.java
View File

@@ -51,7 +51,7 @@ public class WordCheckWord extends AbstractWordCheck {
int tempLen = 0;
int maxWhite = 0;
int maxBlack = 0;
int skipLen=0;
int skipLen = 0;
for (int i = beginIndex; i < rawChars.length; i++) {
if (wordCharIgnore.ignore(i, rawChars, innerContext) && tempLen != 0) {
@@ -67,24 +67,18 @@ public class WordCheckWord extends AbstractWordCheck {
WordContainsTypeEnum wordContainsTypeEnumDeny = wordData.contains(stringBuilder, innerContext);
if (WordContainsTypeEnum.CONTAINS_END.equals(wordContainsTypeEnumAllow)) {
maxWhite += tempLen;
if (!failFast) {
//此处将tempLen设为0，为了防止重复累加
tempLen = 0;
}else{
maxWhite = tempLen;
if (failFast) {
//为falFast模式，主动设为notFound退出循环
wordContainsTypeEnumAllow=WordContainsTypeEnum.NOT_FOUND;
wordContainsTypeEnumAllow = WordContainsTypeEnum.NOT_FOUND;
}
}
if (WordContainsTypeEnum.CONTAINS_END.equals(wordContainsTypeEnumDeny)) {
maxBlack += tempLen;
if (!failFast) {
//此处将tempLen设为0，为了防止重复累加
tempLen = 0;
}else{
maxBlack = tempLen;
if (failFast) {
//为falFast模式，主动设为notFound退出循环
wordContainsTypeEnumDeny=WordContainsTypeEnum.NOT_FOUND;
wordContainsTypeEnumDeny = WordContainsTypeEnum.NOT_FOUND;
}
}
@@ -95,8 +89,8 @@ public class WordCheckWord extends AbstractWordCheck {
}
String string = stringBuilder.toString();
String wordAllow = string.substring(0, Math.max(0,maxWhite - skipLen));
String wordDeny = string.substring(0, Math.max(0,maxBlack - skipLen));
String wordAllow = string.substring(0, Math.max(0, maxWhite - skipLen));
String wordDeny = string.substring(0, Math.max(0, maxBlack - skipLen));
return WordLengthResult.newInstance()

20
src/test/java/com/github/houbb/sensitive/word/bs/SensitiveWordBsReplaceTest.java
View File

@@ -1,20 +0,0 @@
package com.github.houbb.sensitive.word.bs;
import com.github.houbb.sensitive.word.api.IWordReplace;
import com.github.houbb.sensitive.word.replace.MyWordReplace;
import org.junit.Assert;
import org.junit.Test;
public class SensitiveWordBsReplaceTest {
@Test
public void defineReplaceTest() {
final String text = "五星红旗迎风飘扬，毛主席的画像屹立在天安门前。";
SensitiveWordBs sensitiveWordBs = SensitiveWordBs.newInstance().init();
IWordReplace replace = new MyWordReplace();
String result = sensitiveWordBs.replace(text, replace);
Assert.assertEquals("国家旗帜迎风飘扬，教员的画像屹立在***前。", result);
}
}

200
src/test/java/com/github/houbb/sensitive/word/bs/SensitiveWordFailFastTest.java
View File

@@ -1,10 +1,12 @@
package com.github.houbb.sensitive.word.bs;
import com.github.houbb.sensitive.word.api.IWordAllow;
import com.github.houbb.sensitive.word.api.IWordDeny;
import org.junit.Assert;
import org.junit.Test;
import java.util.Arrays;
import java.util.Collections;
import java.util.List;
public class SensitiveWordFailFastTest {
@@ -20,12 +22,108 @@ public class SensitiveWordFailFastTest {
}
}).init();
String text = "我在我的家里玩我的世界";
SensitiveWordBs bs1 = SensitiveWordBs.newInstance()
.failFastWordPattern(true)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("操你妈");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("你");
}
})
.init();
//黑长白短，且初始下标一致
SensitiveWordBs bs2 = SensitiveWordBs.newInstance()
.failFastWordPattern(true)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("大傻逼");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("大");
}
})
.init();
//白长黑短，且白和黑初始下标不再一起
SensitiveWordBs bs3 = SensitiveWordBs.newInstance()
.failFastWordPattern(true)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("口交");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("地铁口交易");
}
})
.init();
//白长黑短，且白和黑初始下标在一起
SensitiveWordBs bs4 = SensitiveWordBs.newInstance()
.failFastWordPattern(true)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("龟孙");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("龟孙可");
}
})
.init();
String text = "我在我的家里玩我的世界";
List<String> textList = bs.findAll(text);
Assert.assertEquals(Arrays.asList("我的", "我的"), textList);
String text1 = "操你妈";
List<String> textList1 = bs1.findAll(text1);
Assert.assertEquals(Collections.singletonList("操你妈"), textList1);
String text2 = "大傻逼";
List<String> textList2 = bs2.findAll(text2);
Assert.assertEquals(Collections.singletonList("大傻逼"), textList2);
String text3 = "地铁口交易";
List<String> textList3 = bs3.findAll(text3);
Assert.assertTrue("Expected empty list", textList3.isEmpty());
String text4 = "龟孙可";
List<String> textList4 = bs4.findAll(text4);
Assert.assertTrue("Expected empty list", textList4.isEmpty());
}
@Test
public void fallOverTest() {
SensitiveWordBs bs = SensitiveWordBs.newInstance()
@@ -37,11 +135,109 @@ public class SensitiveWordFailFastTest {
}
}).init();
String text = "我在我的家里玩我的世界";
//黑长白短，且初始下标不一致
SensitiveWordBs bs1 = SensitiveWordBs.newInstance()
.failFastWordPattern(false)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("操你妈");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("你");
}
})
.init();
//黑长白短，且初始下标一致
SensitiveWordBs bs2 = SensitiveWordBs.newInstance()
.failFastWordPattern(false)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("大傻逼");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("大");
}
})
.init();
//白长黑短，且白和黑初始下标不再一起
SensitiveWordBs bs3 = SensitiveWordBs.newInstance()
.failFastWordPattern(false)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("口交");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("地铁口交易");
}
})
.init();
//白长黑短，且白和黑初始下标在一起
SensitiveWordBs bs4 = SensitiveWordBs.newInstance()
.failFastWordPattern(false)
.wordDeny(new IWordDeny() {
@Override
public List<String> deny() {
return Collections.singletonList("龟孙");
}
})
.wordAllow(new IWordAllow() {
@Override
public List<String> allow() {
return Collections.singletonList("龟孙可");
}
})
.init();
String text = "我在我的家里玩我的世界";
List<String> textList = bs.findAll(text);
Assert.assertEquals(Arrays.asList("我的", "我的世界"), textList);
String text1 = "操你妈";
List<String> textList1 = bs1.findAll(text1);
Assert.assertEquals(Collections.singletonList("操你妈"), textList1);
String text2 = "大傻逼";
List<String> textList2 = bs2.findAll(text2);
Assert.assertEquals(Collections.singletonList("大傻逼"), textList2);
String text3 = "地铁口交易";
List<String> textList3 = bs3.findAll(text3);
Assert.assertTrue("Expected empty list", textList3.isEmpty());
String text4 = "龟孙可";
List<String> textList4 = bs4.findAll(text4);
Assert.assertTrue("Expected empty list", textList4.isEmpty());
}
}