Obfuscation Detection

import { Filter } from 'glin-profanity';

// Enable obfuscation detection
const filter = new Filter({
  allowObfuscatedMatch: true,
  languages: ['english']
  // wordBoundaries automatically set to false
});

// Test various obfuscation patterns
console.log(filter.isProfane('sh1t'));        // true - number substitution
console.log(filter.isProfane('f*ck'));        // true - asterisk removal
console.log(filter.isProfane('d@mn'));        // true - symbol substitution  
console.log(filter.isProfane('a$$hole'));     // true - dollar sign substitution
console.log(filter.isProfane('sh!t'));        // true - exclamation substitution
console.log(filter.isProfane('daaaamn'));     // true - repeated characters

// Detailed analysis
const result = filter.checkProfanity('This sh1t is f*cking annoying!');
console.log(result.containsProfanity);        // true
console.log(result.profaneWords);             // ['sh1t', 'f*cking']
console.log(result.processedText);            // 'This **** is ****ing annoying!'

import { Filter } from 'glin-profanity';

// Advanced obfuscation with fuzzy matching
const advancedFilter = new Filter({
  allowObfuscatedMatch: true,
  fuzzyToleranceLevel: 0.6,    // Lower threshold for more aggressive matching
  severityLevels: true,        // Track EXACT vs FUZZY matches
  replaceWith: '[CENSORED]',   // Custom replacement text
  languages: ['english', 'spanish']
});

// Test complex obfuscation patterns
const testCases = [
  'This is bull$h1t!',         // Mixed substitution
  'F*cking @$$holes!',         // Multiple patterns
  'Daaaamn th@t $ucks!',       // Repeated + substitution
  'Such a b1tch move',         // Number + partial match
  'Holy $h!t that was bad'     // Multiple symbol substitution
];

testCases.forEach(text => {
  const result = advancedFilter.checkProfanity(text);
  console.log(`Text: "${text}"`);
  console.log(`- Contains profanity: ${result.containsProfanity}`);
  console.log(`- Detected words: ${result.profaneWords.join(', ')}`);
  console.log(`- Severity map:`, result.severityMap);
  console.log('---');
});

from glin_profanity import Filter

# Enable obfuscation detection (note snake_case)
filter_instance = Filter({
    "allow_obfuscated_match": True,
    "languages": ["english"]
    # word_boundaries automatically set to False
})

# Test identical patterns to JavaScript
print(filter_instance.is_profane('sh1t'))        # True - number substitution
print(filter_instance.is_profane('f*ck'))        # True - asterisk removal
print(filter_instance.is_profane('d@mn'))        # True - symbol substitution
print(filter_instance.is_profane('a$$hole'))     # True - dollar substitution
print(filter_instance.is_profane('sh!t'))        # True - exclamation substitution
print(filter_instance.is_profane('daaaamn'))     # True - repeated characters

# Detailed analysis with snake_case results
result = filter_instance.check_profanity('This sh1t is f*cking annoying!')
print(result["contains_profanity"])              # True
print(result["profane_words"])                   # ['sh1t', 'f*cking']
print(result["processed_text"])                  # 'This **** is ****ing annoying!'

from glin_profanity import Filter

# Verify identical behavior between JS and Python
def test_obfuscation_parity():
    # Same configuration as JavaScript examples
    py_filter = Filter({
        "allow_obfuscated_match": True,
        "fuzzy_tolerance_level": 0.6,
        "severity_levels": True,
        "languages": ["english"]
    })
    
    # Test cases from cross-language parity tests
    test_cases = [
        "This is d*mn annoying",      # Asterisk substitution
        "This is d4mn bad",           # Number substitution (not in standard mapping)
        "This is d@mn terrible",      # Symbol substitution
        "This is daaaammmn bad",      # Repeated characters
        "What the f*ck is this?"      # Real-world usage
    ]
    
    for text in test_cases:
        result = py_filter.check_profanity(text)
        print(f'Text: "{text}"')
        print(f'- Detected: {result["contains_profanity"]}')
        print(f'- Words: {result["profane_words"]}')
        if "severity_map" in result:
            print(f'- Severity: {result["severity_map"]}')
        print('---')

test_obfuscation_parity()

const filter = new Filter({
  allowObfuscatedMatch: false,  // Default
  wordBoundaries: true          // Default
});

// Only detects whole words
console.log(filter.isProfane('damn'));       // true - whole word
console.log(filter.isProfane('damnit'));     // false - part of larger word
console.log(filter.isProfane('goddamn'));    // false - part of compound word

const filter = new Filter({
  allowObfuscatedMatch: true,   // Enables obfuscation
  // wordBoundaries: false      // Automatically disabled
});

// Detects partial matches and obfuscated patterns
console.log(filter.isProfane('damn'));       // true - whole word
console.log(filter.isProfane('d@mn'));       // true - obfuscated
console.log(filter.isProfane('damnit'));     // true - partial match
console.log(filter.isProfane('goddamn'));    // true - partial match
console.log(filter.isProfane('godd@mn'));    // true - partial + obfuscated