glin-profanity 2.3.8 → 3.1.0

package/dist/index.d.cts

@@ -1,89 +1,5 @@
-/**
- * Type definitions for glin-profanity JavaScript/TypeScript package.
- * Unified API that mirrors the Python package structure.
- */
-/** Severity levels for profanity matches - unified with Python */
-declare enum SeverityLevel {
-    EXACT = 1,
-    FUZZY = 2
-}
-/** Supported languages - unified list with Python */
-type Language = 'arabic' | 'chinese' | 'czech' | 'danish' | 'english' | 'esperanto' | 'finnish' | 'french' | 'german' | 'hindi' | 'hungarian' | 'italian' | 'japanese' | 'korean' | 'norwegian' | 'persian' | 'polish' | 'portuguese' | 'russian' | 'spanish' | 'swedish' | 'thai' | 'turkish';
-/** Represents a profanity match in text - unified with Python */
-interface Match {
-    word: string;
-    index: number;
-    severity: SeverityLevel;
-    contextScore?: number;
-    reason?: string;
-    isWhitelisted?: boolean;
-}
-/** Result of profanity check operation - unified field names */
-interface CheckProfanityResult {
-    containsProfanity: boolean;
-    profaneWords: string[];
-    processedText?: string;
-    severityMap?: Record<string, SeverityLevel>;
-    matches?: Match[];
-    contextScore?: number;
-    reason?: string;
-}
-/** Configuration for context-aware filtering - unified with Python */
-interface ContextAwareConfig {
-    enableContextAware?: boolean;
-    contextWindow?: number;
-    confidenceThreshold?: number;
-    domainWhitelists?: Record<string, string[]>;
-}
-/** Main filter configuration options - unified with Python */
-interface FilterConfig extends ContextAwareConfig {
-    languages?: Language[];
-    allLanguages?: boolean;
-    caseSensitive?: boolean;
-    wordBoundaries?: boolean;
-    customWords?: string[];
-    replaceWith?: string;
-    severityLevels?: boolean;
-    ignoreWords?: string[];
-    logProfanity?: boolean;
-    allowObfuscatedMatch?: boolean;
-    fuzzyToleranceLevel?: number;
-}
-/** Result with minimum severity filtering */
-interface FilteredProfanityResult {
-    result: CheckProfanityResult;
-    filteredWords: string[];
-}
-
-declare class Filter {
-    private words;
-    private caseSensitive;
-    private wordBoundaries;
-    private replaceWith?;
-    private severityLevels;
-    private ignoreWords;
-    private logProfanity;
-    private allowObfuscatedMatch;
-    private fuzzyToleranceLevel;
-    private enableContextAware;
-    private contextWindow;
-    private confidenceThreshold;
-    private contextAnalyzer?;
-    private primaryLanguage;
-    constructor(config?: FilterConfig);
-    private debugLog;
-    private normalizeObfuscated;
-    private getRegex;
-    private isFuzzyToleranceMatch;
-    private evaluateSeverity;
-    isProfane(value: string): boolean;
-    matches(word: string): boolean;
-    checkProfanity(text: string): CheckProfanityResult;
-    checkProfanityWithMinSeverity(text: string, minSeverity?: SeverityLevel): {
-        filteredWords: string[];
-        result: CheckProfanityResult;
-    };
-}
+import { L as Language, S as SeverityLevel, C as CheckProfanityResult } from './types-BgQe4FSE.cjs';
+export { c as ContextAwareConfig, F as Filter, a as FilterConfig, b as FilteredProfanityResult, H as HybridAnalysisResult, e as MLAnalysisResult, f as MLDetectorConfig, M as Match, F as ProfanityFilter, T as ToxicityLabel, d as ToxicityPrediction } from './types-BgQe4FSE.cjs';
 
 interface ProfanityCheckerConfig {
     languages?: Language[];
@@ -93,6 +9,7 @@ interface ProfanityCheckerConfig {
     customWords?: string[];
     replaceWith?: string;
     severityLevels?: boolean;
+    ignoreWords?: string[];
     allowObfuscatedMatch?: boolean;
     fuzzyToleranceLevel?: number;
     minSeverity?: SeverityLevel;
@@ -117,4 +34,252 @@ declare const useProfanityChecker: (config?: ProfanityCheckerConfig) => {
     isWordProfane: (word: string) => boolean;
 };
 
-export { type CheckProfanityResult, type ContextAwareConfig, Filter, type FilterConfig, type FilteredProfanityResult, type Language, type Match, type ProfanityCheckResult, type ProfanityCheckerConfig, Filter as ProfanityFilter, SeverityLevel, checkProfanity, checkProfanityAsync, isWordProfane, useProfanityChecker };
+/**
+ * @fileoverview Leetspeak detection and normalization utilities.
+ * Converts leetspeak/1337 speak text back to standard characters for profanity detection.
+ * @module utils/leetspeak
+ */
+/**
+ * Leetspeak detection intensity levels.
+ * - `basic`: Common substitutions only (0→o, 1→i, 3→e, 4→a, 5→s)
+ * - `moderate`: Basic + symbols (@→a, $→s, !→i) and repeated chars
+ * - `aggressive`: All known substitutions including multi-char patterns
+ */
+type LeetspeakLevel = 'basic' | 'moderate' | 'aggressive';
+/**
+ * Configuration options for leetspeak normalization.
+ */
+interface LeetspeakOptions {
+    /**
+     * Detection intensity level.
+     * @default 'moderate'
+     */
+    level?: LeetspeakLevel;
+    /**
+     * Whether to collapse repeated characters (e.g., "fuuuuck" → "fuck").
+     * @default true
+     */
+    collapseRepeated?: boolean;
+    /**
+     * Maximum allowed consecutive repeated characters before collapsing.
+     * @default 2
+     */
+    maxRepeated?: number;
+    /**
+     * Whether to remove spaces between single characters (e.g., "f u c k" → "fuck").
+     * @default true
+     */
+    removeSpacedChars?: boolean;
+}
+/**
+ * Normalizes leetspeak text to standard characters.
+ *
+ * @param text - The input text containing potential leetspeak
+ * @param options - Configuration options for normalization
+ * @returns The normalized text with leetspeak characters replaced
+ *
+ * @example
+ * ```typescript
+ * import { normalizeLeetspeak } from 'glin-profanity';
+ *
+ * normalizeLeetspeak('f4ck');     // Returns: 'fack'
+ * normalizeLeetspeak('sh!t');     // Returns: 'shit'
+ * normalizeLeetspeak('b1tch');    // Returns: 'bitch'
+ * normalizeLeetspeak('@ss');      // Returns: 'ass'
+ * normalizeLeetspeak('f u c k');  // Returns: 'fuck'
+ * normalizeLeetspeak('fuuuuck');  // Returns: 'fuck'
+ * ```
+ */
+declare function normalizeLeetspeak(text: string, options?: LeetspeakOptions): string;
+/**
+ * Collapses sequences of spaced single characters into words.
+ * Handles patterns like "f u c k" → "fuck" and "s h i t" → "shit".
+ *
+ * @param text - The input text
+ * @returns Text with spaced characters collapsed
+ *
+ * @example
+ * ```typescript
+ * collapseSpacedCharacters('f u c k you');  // Returns: 'fuck you'
+ * collapseSpacedCharacters('this is s h i t'); // Returns: 'this is shit'
+ * ```
+ */
+declare function collapseSpacedCharacters(text: string): string;
+/**
+ * Collapses repeated consecutive characters beyond a threshold.
+ * Handles patterns like "fuuuuck" → "fuck" and "shiiiit" → "shit".
+ *
+ * @param text - The input text
+ * @param maxRepeated - Maximum allowed consecutive repeated characters
+ * @returns Text with repeated characters collapsed
+ *
+ * @example
+ * ```typescript
+ * collapseRepeatedCharacters('fuuuuck', 2);  // Returns: 'fuuck'
+ * collapseRepeatedCharacters('fuuuuck', 1);  // Returns: 'fuck'
+ * ```
+ */
+declare function collapseRepeatedCharacters(text: string, maxRepeated?: number): string;
+/**
+ * Detects if text contains potential leetspeak patterns.
+ * Useful for deciding whether to apply leetspeak normalization.
+ *
+ * @param text - The input text to analyze
+ * @returns True if leetspeak patterns are detected
+ *
+ * @example
+ * ```typescript
+ * containsLeetspeak('hello');     // Returns: false
+ * containsLeetspeak('h3ll0');     // Returns: true
+ * containsLeetspeak('f4ck');      // Returns: true
+ * containsLeetspeak('@ss');       // Returns: true
+ * ```
+ */
+declare function containsLeetspeak(text: string): boolean;
+/**
+ * Creates a normalized variant generator for a word.
+ * Generates all possible leetspeak variants of a dictionary word.
+ *
+ * @param word - The base word to generate variants for
+ * @param level - The leetspeak level to use for variant generation
+ * @returns Array of possible leetspeak variants
+ *
+ * @example
+ * ```typescript
+ * generateLeetspeakVariants('ass');
+ * // Returns: ['ass', '@ss', 'a$$', '@$$', '4ss', '4$$', ...]
+ * ```
+ */
+declare function generateLeetspeakVariants(word: string, level?: LeetspeakLevel): string[];
+
+/**
+ * @fileoverview Unicode normalization utilities for profanity detection.
+ * Handles homoglyphs, full-width characters, diacritics, and other Unicode tricks.
+ * @module utils/unicode
+ */
+/**
+ * Configuration options for Unicode normalization.
+ */
+interface UnicodeNormalizationOptions {
+    /**
+     * Apply NFKD normalization to decompose characters.
+     * @default true
+     */
+    nfkd?: boolean;
+    /**
+     * Convert homoglyphs (lookalike characters) to ASCII.
+     * @default true
+     */
+    homoglyphs?: boolean;
+    /**
+     * Convert full-width characters to half-width.
+     * @default true
+     */
+    fullWidth?: boolean;
+    /**
+     * Remove diacritical marks (accents, umlauts, etc.).
+     * @default true
+     */
+    removeDiacritics?: boolean;
+    /**
+     * Remove zero-width characters (ZWJ, ZWNJ, etc.).
+     * @default true
+     */
+    removeZeroWidth?: boolean;
+}
+/**
+ * Normalizes Unicode text for consistent profanity detection.
+ * Handles various Unicode tricks used to evade filters.
+ *
+ * @param text - The input text containing potential Unicode obfuscation
+ * @param options - Configuration options for normalization
+ * @returns The normalized text
+ *
+ * @example
+ * ```typescript
+ * import { normalizeUnicode } from 'glin-profanity';
+ *
+ * normalizeUnicode('fυck');     // Returns: 'fuck' (Greek upsilon → u)
+ * normalizeUnicode('fＵck');    // Returns: 'fuck' (full-width U → u)
+ * normalizeUnicode('fück');     // Returns: 'fuck' (ü → u)
+ * normalizeUnicode('fùck');     // Returns: 'fuck' (ù → u)
+ * normalizeUnicode('f​uck');    // Returns: 'fuck' (removes zero-width space)
+ * ```
+ */
+declare function normalizeUnicode(text: string, options?: UnicodeNormalizationOptions): string;
+/**
+ * Removes zero-width and invisible characters from text.
+ *
+ * @param text - The input text
+ * @returns Text with zero-width characters removed
+ */
+declare function removeZeroWidthCharacters(text: string): string;
+/**
+ * Converts full-width ASCII characters to half-width.
+ * Full-width characters (U+FF01 to U+FF5E) are used in CJK text
+ * but can also be used to evade filters.
+ *
+ * @param text - The input text
+ * @returns Text with full-width characters converted
+ *
+ * @example
+ * ```typescript
+ * convertFullWidth('ＡＢＣ');  // Returns: 'ABC'
+ * convertFullWidth('ｆｕｃｋ'); // Returns: 'fuck'
+ * ```
+ */
+declare function convertFullWidth(text: string): string;
+/**
+ * Converts homoglyph characters to their ASCII equivalents.
+ *
+ * @param text - The input text
+ * @returns Text with homoglyphs converted
+ */
+declare function convertHomoglyphs(text: string): string;
+/**
+ * Applies NFKD normalization and optionally removes diacritical marks.
+ * NFKD decomposes characters into base characters and combining marks.
+ *
+ * @param text - The input text
+ * @param removeDiacritics - Whether to remove diacritical marks
+ * @returns Normalized text
+ *
+ * @example
+ * ```typescript
+ * normalizeNFKD('fück', true);   // Returns: 'fuck'
+ * normalizeNFKD('café', true);   // Returns: 'cafe'
+ * normalizeNFKD('naïve', true);  // Returns: 'naive'
+ * ```
+ */
+declare function normalizeNFKD(text: string, removeDiacritics?: boolean): string;
+/**
+ * Detects if text contains potential Unicode obfuscation.
+ * Useful for deciding whether to apply Unicode normalization.
+ *
+ * @param text - The input text to analyze
+ * @returns True if Unicode obfuscation patterns are detected
+ *
+ * @example
+ * ```typescript
+ * containsUnicodeObfuscation('hello');     // Returns: false
+ * containsUnicodeObfuscation('fυck');      // Returns: true (Greek letter)
+ * containsUnicodeObfuscation('f​uck');     // Returns: true (zero-width)
+ * ```
+ */
+declare function containsUnicodeObfuscation(text: string): boolean;
+/**
+ * Gets the character set being used in text.
+ * Helps identify mixed-script attacks (e.g., mixing Latin and Cyrillic).
+ *
+ * @param text - The input text
+ * @returns Object with detected character set information
+ */
+declare function detectCharacterSets(text: string): {
+    hasLatin: boolean;
+    hasCyrillic: boolean;
+    hasGreek: boolean;
+    hasFullWidth: boolean;
+    hasMixed: boolean;
+};
+
+export { CheckProfanityResult, Language, type LeetspeakLevel, type LeetspeakOptions, type ProfanityCheckResult, type ProfanityCheckerConfig, SeverityLevel, type UnicodeNormalizationOptions, checkProfanity, checkProfanityAsync, collapseRepeatedCharacters, collapseSpacedCharacters, containsLeetspeak, containsUnicodeObfuscation, convertFullWidth, convertHomoglyphs, detectCharacterSets, generateLeetspeakVariants, isWordProfane, normalizeLeetspeak, normalizeNFKD, normalizeUnicode, removeZeroWidthCharacters, useProfanityChecker };

package/dist/index.d.ts

@@ -1,89 +1,5 @@
-/**
- * Type definitions for glin-profanity JavaScript/TypeScript package.
- * Unified API that mirrors the Python package structure.
- */
-/** Severity levels for profanity matches - unified with Python */
-declare enum SeverityLevel {
-    EXACT = 1,
-    FUZZY = 2
-}
-/** Supported languages - unified list with Python */
-type Language = 'arabic' | 'chinese' | 'czech' | 'danish' | 'english' | 'esperanto' | 'finnish' | 'french' | 'german' | 'hindi' | 'hungarian' | 'italian' | 'japanese' | 'korean' | 'norwegian' | 'persian' | 'polish' | 'portuguese' | 'russian' | 'spanish' | 'swedish' | 'thai' | 'turkish';
-/** Represents a profanity match in text - unified with Python */
-interface Match {
-    word: string;
-    index: number;
-    severity: SeverityLevel;
-    contextScore?: number;
-    reason?: string;
-    isWhitelisted?: boolean;
-}
-/** Result of profanity check operation - unified field names */
-interface CheckProfanityResult {
-    containsProfanity: boolean;
-    profaneWords: string[];
-    processedText?: string;
-    severityMap?: Record<string, SeverityLevel>;
-    matches?: Match[];
-    contextScore?: number;
-    reason?: string;
-}
-/** Configuration for context-aware filtering - unified with Python */
-interface ContextAwareConfig {
-    enableContextAware?: boolean;
-    contextWindow?: number;
-    confidenceThreshold?: number;
-    domainWhitelists?: Record<string, string[]>;
-}
-/** Main filter configuration options - unified with Python */
-interface FilterConfig extends ContextAwareConfig {
-    languages?: Language[];
-    allLanguages?: boolean;
-    caseSensitive?: boolean;
-    wordBoundaries?: boolean;
-    customWords?: string[];
-    replaceWith?: string;
-    severityLevels?: boolean;
-    ignoreWords?: string[];
-    logProfanity?: boolean;
-    allowObfuscatedMatch?: boolean;
-    fuzzyToleranceLevel?: number;
-}
-/** Result with minimum severity filtering */
-interface FilteredProfanityResult {
-    result: CheckProfanityResult;
-    filteredWords: string[];
-}
-
-declare class Filter {
-    private words;
-    private caseSensitive;
-    private wordBoundaries;
-    private replaceWith?;
-    private severityLevels;
-    private ignoreWords;
-    private logProfanity;
-    private allowObfuscatedMatch;
-    private fuzzyToleranceLevel;
-    private enableContextAware;
-    private contextWindow;
-    private confidenceThreshold;
-    private contextAnalyzer?;
-    private primaryLanguage;
-    constructor(config?: FilterConfig);
-    private debugLog;
-    private normalizeObfuscated;
-    private getRegex;
-    private isFuzzyToleranceMatch;
-    private evaluateSeverity;
-    isProfane(value: string): boolean;
-    matches(word: string): boolean;
-    checkProfanity(text: string): CheckProfanityResult;
-    checkProfanityWithMinSeverity(text: string, minSeverity?: SeverityLevel): {
-        filteredWords: string[];
-        result: CheckProfanityResult;
-    };
-}
+import { L as Language, S as SeverityLevel, C as CheckProfanityResult } from './types-BgQe4FSE.js';
+export { c as ContextAwareConfig, F as Filter, a as FilterConfig, b as FilteredProfanityResult, H as HybridAnalysisResult, e as MLAnalysisResult, f as MLDetectorConfig, M as Match, F as ProfanityFilter, T as ToxicityLabel, d as ToxicityPrediction } from './types-BgQe4FSE.js';
 
 interface ProfanityCheckerConfig {
     languages?: Language[];
@@ -93,6 +9,7 @@ interface ProfanityCheckerConfig {
     customWords?: string[];
     replaceWith?: string;
     severityLevels?: boolean;
+    ignoreWords?: string[];
     allowObfuscatedMatch?: boolean;
     fuzzyToleranceLevel?: number;
     minSeverity?: SeverityLevel;
@@ -117,4 +34,252 @@ declare const useProfanityChecker: (config?: ProfanityCheckerConfig) => {
     isWordProfane: (word: string) => boolean;
 };
 
-export { type CheckProfanityResult, type ContextAwareConfig, Filter, type FilterConfig, type FilteredProfanityResult, type Language, type Match, type ProfanityCheckResult, type ProfanityCheckerConfig, Filter as ProfanityFilter, SeverityLevel, checkProfanity, checkProfanityAsync, isWordProfane, useProfanityChecker };
+/**
+ * @fileoverview Leetspeak detection and normalization utilities.
+ * Converts leetspeak/1337 speak text back to standard characters for profanity detection.
+ * @module utils/leetspeak
+ */
+/**
+ * Leetspeak detection intensity levels.
+ * - `basic`: Common substitutions only (0→o, 1→i, 3→e, 4→a, 5→s)
+ * - `moderate`: Basic + symbols (@→a, $→s, !→i) and repeated chars
+ * - `aggressive`: All known substitutions including multi-char patterns
+ */
+type LeetspeakLevel = 'basic' | 'moderate' | 'aggressive';
+/**
+ * Configuration options for leetspeak normalization.
+ */
+interface LeetspeakOptions {
+    /**
+     * Detection intensity level.
+     * @default 'moderate'
+     */
+    level?: LeetspeakLevel;
+    /**
+     * Whether to collapse repeated characters (e.g., "fuuuuck" → "fuck").
+     * @default true
+     */
+    collapseRepeated?: boolean;
+    /**
+     * Maximum allowed consecutive repeated characters before collapsing.
+     * @default 2
+     */
+    maxRepeated?: number;
+    /**
+     * Whether to remove spaces between single characters (e.g., "f u c k" → "fuck").
+     * @default true
+     */
+    removeSpacedChars?: boolean;
+}
+/**
+ * Normalizes leetspeak text to standard characters.
+ *
+ * @param text - The input text containing potential leetspeak
+ * @param options - Configuration options for normalization
+ * @returns The normalized text with leetspeak characters replaced
+ *
+ * @example
+ * ```typescript
+ * import { normalizeLeetspeak } from 'glin-profanity';
+ *
+ * normalizeLeetspeak('f4ck');     // Returns: 'fack'
+ * normalizeLeetspeak('sh!t');     // Returns: 'shit'
+ * normalizeLeetspeak('b1tch');    // Returns: 'bitch'
+ * normalizeLeetspeak('@ss');      // Returns: 'ass'
+ * normalizeLeetspeak('f u c k');  // Returns: 'fuck'
+ * normalizeLeetspeak('fuuuuck');  // Returns: 'fuck'
+ * ```
+ */
+declare function normalizeLeetspeak(text: string, options?: LeetspeakOptions): string;
+/**
+ * Collapses sequences of spaced single characters into words.
+ * Handles patterns like "f u c k" → "fuck" and "s h i t" → "shit".
+ *
+ * @param text - The input text
+ * @returns Text with spaced characters collapsed
+ *
+ * @example
+ * ```typescript
+ * collapseSpacedCharacters('f u c k you');  // Returns: 'fuck you'
+ * collapseSpacedCharacters('this is s h i t'); // Returns: 'this is shit'
+ * ```
+ */
+declare function collapseSpacedCharacters(text: string): string;
+/**
+ * Collapses repeated consecutive characters beyond a threshold.
+ * Handles patterns like "fuuuuck" → "fuck" and "shiiiit" → "shit".
+ *
+ * @param text - The input text
+ * @param maxRepeated - Maximum allowed consecutive repeated characters
+ * @returns Text with repeated characters collapsed
+ *
+ * @example
+ * ```typescript
+ * collapseRepeatedCharacters('fuuuuck', 2);  // Returns: 'fuuck'
+ * collapseRepeatedCharacters('fuuuuck', 1);  // Returns: 'fuck'
+ * ```
+ */
+declare function collapseRepeatedCharacters(text: string, maxRepeated?: number): string;
+/**
+ * Detects if text contains potential leetspeak patterns.
+ * Useful for deciding whether to apply leetspeak normalization.
+ *
+ * @param text - The input text to analyze
+ * @returns True if leetspeak patterns are detected
+ *
+ * @example
+ * ```typescript
+ * containsLeetspeak('hello');     // Returns: false
+ * containsLeetspeak('h3ll0');     // Returns: true
+ * containsLeetspeak('f4ck');      // Returns: true
+ * containsLeetspeak('@ss');       // Returns: true
+ * ```
+ */
+declare function containsLeetspeak(text: string): boolean;
+/**
+ * Creates a normalized variant generator for a word.
+ * Generates all possible leetspeak variants of a dictionary word.
+ *
+ * @param word - The base word to generate variants for
+ * @param level - The leetspeak level to use for variant generation
+ * @returns Array of possible leetspeak variants
+ *
+ * @example
+ * ```typescript
+ * generateLeetspeakVariants('ass');
+ * // Returns: ['ass', '@ss', 'a$$', '@$$', '4ss', '4$$', ...]
+ * ```
+ */
+declare function generateLeetspeakVariants(word: string, level?: LeetspeakLevel): string[];
+
+/**
+ * @fileoverview Unicode normalization utilities for profanity detection.
+ * Handles homoglyphs, full-width characters, diacritics, and other Unicode tricks.
+ * @module utils/unicode
+ */
+/**
+ * Configuration options for Unicode normalization.
+ */
+interface UnicodeNormalizationOptions {
+    /**
+     * Apply NFKD normalization to decompose characters.
+     * @default true
+     */
+    nfkd?: boolean;
+    /**
+     * Convert homoglyphs (lookalike characters) to ASCII.
+     * @default true
+     */
+    homoglyphs?: boolean;
+    /**
+     * Convert full-width characters to half-width.
+     * @default true
+     */
+    fullWidth?: boolean;
+    /**
+     * Remove diacritical marks (accents, umlauts, etc.).
+     * @default true
+     */
+    removeDiacritics?: boolean;
+    /**
+     * Remove zero-width characters (ZWJ, ZWNJ, etc.).
+     * @default true
+     */
+    removeZeroWidth?: boolean;
+}
+/**
+ * Normalizes Unicode text for consistent profanity detection.
+ * Handles various Unicode tricks used to evade filters.
+ *
+ * @param text - The input text containing potential Unicode obfuscation
+ * @param options - Configuration options for normalization
+ * @returns The normalized text
+ *
+ * @example
+ * ```typescript
+ * import { normalizeUnicode } from 'glin-profanity';
+ *
+ * normalizeUnicode('fυck');     // Returns: 'fuck' (Greek upsilon → u)
+ * normalizeUnicode('fＵck');    // Returns: 'fuck' (full-width U → u)
+ * normalizeUnicode('fück');     // Returns: 'fuck' (ü → u)
+ * normalizeUnicode('fùck');     // Returns: 'fuck' (ù → u)
+ * normalizeUnicode('f​uck');    // Returns: 'fuck' (removes zero-width space)
+ * ```
+ */
+declare function normalizeUnicode(text: string, options?: UnicodeNormalizationOptions): string;
+/**
+ * Removes zero-width and invisible characters from text.
+ *
+ * @param text - The input text
+ * @returns Text with zero-width characters removed
+ */
+declare function removeZeroWidthCharacters(text: string): string;
+/**
+ * Converts full-width ASCII characters to half-width.
+ * Full-width characters (U+FF01 to U+FF5E) are used in CJK text
+ * but can also be used to evade filters.
+ *
+ * @param text - The input text
+ * @returns Text with full-width characters converted
+ *
+ * @example
+ * ```typescript
+ * convertFullWidth('ＡＢＣ');  // Returns: 'ABC'
+ * convertFullWidth('ｆｕｃｋ'); // Returns: 'fuck'
+ * ```
+ */
+declare function convertFullWidth(text: string): string;
+/**
+ * Converts homoglyph characters to their ASCII equivalents.
+ *
+ * @param text - The input text
+ * @returns Text with homoglyphs converted
+ */
+declare function convertHomoglyphs(text: string): string;
+/**
+ * Applies NFKD normalization and optionally removes diacritical marks.
+ * NFKD decomposes characters into base characters and combining marks.
+ *
+ * @param text - The input text
+ * @param removeDiacritics - Whether to remove diacritical marks
+ * @returns Normalized text
+ *
+ * @example
+ * ```typescript
+ * normalizeNFKD('fück', true);   // Returns: 'fuck'
+ * normalizeNFKD('café', true);   // Returns: 'cafe'
+ * normalizeNFKD('naïve', true);  // Returns: 'naive'
+ * ```
+ */
+declare function normalizeNFKD(text: string, removeDiacritics?: boolean): string;
+/**
+ * Detects if text contains potential Unicode obfuscation.
+ * Useful for deciding whether to apply Unicode normalization.
+ *
+ * @param text - The input text to analyze
+ * @returns True if Unicode obfuscation patterns are detected
+ *
+ * @example
+ * ```typescript
+ * containsUnicodeObfuscation('hello');     // Returns: false
+ * containsUnicodeObfuscation('fυck');      // Returns: true (Greek letter)
+ * containsUnicodeObfuscation('f​uck');     // Returns: true (zero-width)
+ * ```
+ */
+declare function containsUnicodeObfuscation(text: string): boolean;
+/**
+ * Gets the character set being used in text.
+ * Helps identify mixed-script attacks (e.g., mixing Latin and Cyrillic).
+ *
+ * @param text - The input text
+ * @returns Object with detected character set information
+ */
+declare function detectCharacterSets(text: string): {
+    hasLatin: boolean;
+    hasCyrillic: boolean;
+    hasGreek: boolean;
+    hasFullWidth: boolean;
+    hasMixed: boolean;
+};
+
+export { CheckProfanityResult, Language, type LeetspeakLevel, type LeetspeakOptions, type ProfanityCheckResult, type ProfanityCheckerConfig, SeverityLevel, type UnicodeNormalizationOptions, checkProfanity, checkProfanityAsync, collapseRepeatedCharacters, collapseSpacedCharacters, containsLeetspeak, containsUnicodeObfuscation, convertFullWidth, convertHomoglyphs, detectCharacterSets, generateLeetspeakVariants, isWordProfane, normalizeLeetspeak, normalizeNFKD, normalizeUnicode, removeZeroWidthCharacters, useProfanityChecker };