atpota.to
The 1st decentralized social network for sharing when you're on the toilet. Post a "flush" today! Powered by the AT Protocol.
1// List of banned terms (this is a sample list that can be modified as needed)
2// These words will be filtered from all posts in the application
3const BANNED_WORDS: string[] = [
4 // Generic offensive terms
5 'slur',
6
7 // Hate speech related
8 'racist', 'bigot', 'bigotry', 'homophobic', 'transphobic',
9
10 // Profanity
11 'fuck', 'damn', 'ass', 'asshole', 'bitch',
12
13 // Violence
14 'kill', 'murder', 'attack', 'violence', 'harm', 'hurt',
15
16 // Discrimination terms
17 'retard', 'retarded', 'idiot', 'stupid', 'dumb',
18
19 // Sexual content
20 'penis', 'vagina', 'dick', 'cock', 'pussy', 'sex',
21 'masturbate', 'orgasm', 'horny', 'erection',
22 'blowjob', 'handjob',
23
24 // Spam-related
25 'viagra', 'cialis', 'enlarge', 'cryptocurrency', 'bitcoin', 'ethereum',
26 'make money', 'get rich', 'earn fast', 'pyramid', 'scheme',
27
28 // Links and promotion
29 'discord.gg', 'telegram.me'
30];
31
32// Special regexes for detecting slurs - adapted from https://github.com/Blank-Cheque/Slurs
33/* eslint-disable no-misleading-character-class */
34const EXPLICIT_SLUR_REGEXES = [
35 /\bc[hH][iIl1][nN][kKsS]?\b/, // Anti-Asian slur
36 /\bc[oO]{2}[nN][sS]?\b/, // Anti-Black slur
37 /\bf[aA][gG]{1,2}([oOeE][tT]?|[iIyY][nNeE]?)?s?\b/, // Anti-LGBTQ+ slur
38 /\bk[iIyY][kK][eE][sS]?\b/, // Anti-Semitic slur
39 /\bn[iIl1oO][gG]{2}([aAeE][rR]?|[lL][eE][tT]|[nNoO][gG])?s?\b/, // Anti-Black slur
40 /\bn[iIl1oO][gG]{2}[aAeE][sS]\b/, // Anti-Black slur variation
41 /\bt[rR][aA][nN][nN][iIyY][eE]?[sS]?\b/, // Anti-transgender slur
42];
43
44/**
45 * Checks if a text contains any banned words
46 * @param text The text to check
47 * @returns True if the text contains banned words, false otherwise
48 */
49export function containsBannedWords(text: string): boolean {
50 if (!text) return false;
51
52 // Normalize text by removing common obfuscation techniques
53 let normalizedText = text.toLowerCase()
54 .replace(/0/g, 'o') // Replace numbers with letters they resemble
55 .replace(/1/g, 'i')
56 .replace(/3/g, 'e')
57 .replace(/4/g, 'a')
58 .replace(/5/g, 's')
59 .replace(/\$/g, 's') // Replace symbols with letters they resemble
60 .replace(/@/g, 'a')
61 .replace(/!/g, 'i')
62 .replace(/\*/g, '') // Remove common censorship characters
63 .replace(/\./g, '')
64 .replace(/-/g, '')
65 .replace(/_/g, '')
66 .replace(/\s+/g, ' '); // Normalize whitespace
67
68 // Check explicit slur regexes (specialized pattern matching)
69 for (const regex of EXPLICIT_SLUR_REGEXES) {
70 if (regex.test(text)) {
71 return true;
72 }
73 }
74
75 // Check for exact matches and partial matches in the banned words list
76 return BANNED_WORDS.some(word => {
77 // Check for exact word match with word boundaries
78 const exactRegex = new RegExp(`\\b${word}\\b`, 'i');
79 if (exactRegex.test(normalizedText)) return true;
80
81 // Check for intentional letter spacing like "s e x"
82 const spacedWord = word.split('').join('\\s*');
83 const spacedRegex = new RegExp(`\\b${spacedWord}\\b`, 'i');
84 if (spacedRegex.test(normalizedText)) return true;
85
86 // For shorter words (4 letters or less), also check for substring matches
87 // This helps catch compound words that contain banned terms
88 if (word.length <= 4) {
89 const substringRegex = new RegExp(word, 'i');
90 return substringRegex.test(normalizedText);
91 }
92
93 return false;
94 });
95}
96
97/**
98 * Sanitizes text by removing or replacing banned words
99 * @param text The text to sanitize
100 * @returns Sanitized text with banned words replaced by asterisks
101 */
102export function sanitizeText(text: string): string {
103 if (!text) return text;
104
105 let sanitized = text;
106
107 // First pass: replace exact word matches
108 BANNED_WORDS.forEach(word => {
109 const exactRegex = new RegExp(`\\b${word}\\b`, 'gi');
110 sanitized = sanitized.replace(exactRegex, '*'.repeat(word.length));
111 });
112
113 // Second pass: look for spaced out words (e.g., "s e x")
114 BANNED_WORDS.forEach(word => {
115 if (word.length > 2) {
116 const spacedWord = word.split('').join('\\s*');
117 const spacedRegex = new RegExp(`\\b${spacedWord}\\b`, 'gi');
118
119 // Use a callback to replace with the right number of asterisks
120 sanitized = sanitized.replace(spacedRegex, (match) => {
121 return '*'.repeat(match.replace(/\s+/g, '').length);
122 });
123 }
124 });
125
126 // Third pass: for shorter words, also check substrings in larger words
127 BANNED_WORDS.filter(word => word.length <= 4).forEach(word => {
128 // This regex finds the word as a substring but not at word boundaries
129 const substringRegex = new RegExp(`(?<!\\w)${word}(?!\\w)`, 'gi');
130 sanitized = sanitized.replace(substringRegex, '*'.repeat(word.length));
131 });
132
133 return sanitized;
134}
135
136/**
137 * Specialized function to check for explicit slurs using the advanced regex patterns
138 * @param text The text to check
139 * @returns True if the text contains any explicit slurs
140 */
141export function containsExplicitSlurs(text: string): boolean {
142 if (!text) return false;
143
144 return EXPLICIT_SLUR_REGEXES.some(regex => regex.test(text));
145}
146
147/**
148 * Checks if a character is one of our allowed emojis
149 * @param emoji The emoji character to check
150 * @returns True if the emoji is in our allowed list
151 */
152export function isAllowedEmoji(emoji: string): boolean {
153 const ALLOWED_EMOJIS = [
154 '🚽', '🧻', '💩', '💨', '🚾', '🧼', '🪠', '🚻', '🩸', '💧', '💦', '😌',
155 '😣', '🤢', '🤮', '🥴', '😮💨', '😳', '😵', '🌾', '🍦', '📱', '📖', '💭',
156 '1️⃣', '2️⃣', '🟡', '🟤'
157 ];
158
159 return ALLOWED_EMOJIS.includes(emoji);
160}