deep-profanity-filter/dist/input_preprocessor.d.ts

A thorough profanity filter that considers most common circumventions. Works with your custom list of blocked and whitelisted words and phrases. Identifies and/or replaces bad words. Works with *wildcards* at *start and/or end* of words.

/**
 * Replaces all emojis in a text that feature a letter with
 * normal latin characters.
 * Example: "🇬⭕ 🔛" turns into "go on" or
 * "🅿🇺®️💰🇪" turns into "purse".
 * Note: This does NOT replace random emojis used to represent
 * or mask letters, such as '🌸' representing an 'O'.
 *
 * Suggested usage: `textToLatin(unEmoji(inputText))`
 * @param {string} inputText - The text from which you would like to remove all text based emojis.
 * @returns the input text, with all letter based emojis transformed to become text.
 */
export declare function unEmoji(inputText: string): string;
/**
 * Removes most common accents from characters.
 * Example: The text "Z̵̡̭͝ả̶̬̘̈́l̶̜͗g̵̜̲͒́o̶̞̅̊" becomes "Zalgo",
 * the text "à-côtés" becomes "a-cotes",
 * non-latin characters stay non-latin, e.g. "ᑕⓞ֑ο̤͕𝕃ܑׅ" becomes "ᑕⓞο𝕃".
 * @param {string} inputText - The text for which you wish to have all
 * accents removed.
 * @returns the input text, stripped of all accents.
 */
export declare function removeTextAccents(inputText: string): string;
/**
 * Converts a text of fancy unicode font to latin alphabet characters.
 * This translation happens based on "visual appearance" of the letters,
 * so if you do this to text that is written in a language of non-latin
 * alphabet, you will get weird outputs.
 *
 * Disclaimer: This may at times mistranslate messages, and the list of
 * characters that get converted is most likely not complete, although
 * it is very thoroughly assembled. It will remove most common accents,
 * and returns a latin string in lower case letters. Any characters that
 * could not be mapped to latin characters will still appear in the string.
 *
 * Example:
 * "ᵺⓘ꯱ ₮Ꮛ乂Շ" would turn into "this text" or "Z̵̡̭͝ả̶̬̘̈́l̶̜͗g̵̜̲͒́o̶̞̅̊" turns into "zalgo", or
 * "ᑕⓞ֑ο̤͕𝕃ܑׅ" turns into "cool".
 *
 * Suggested usage: `textToLatin(unEmoji(inputText))`
 * @param {string} inputText - The text which you would like to convert to latin
 * @returns the input text, with foreign or special alphabet letters translated
 * to latin lower case characters
 */
export declare function textToLatin(inputText: string): string;
/**
 * For any given input text, reduce any repeating characters to a given maximum amount of repetitions.
 *
 * As an example, the input string: `"heeellllooooo"` becomes: `"heelloo"` if that number is 2, or
 * `"heeelllooo"` if that number is 3, or `helo` if that number is 1.
 *
 * For English, it is recommended to not use values lower than 2. If this preprocessing is used, make
 * sure that the bad words and allowed terms also feature at most the same number of repeated characters.
 * I.e. if using this with the number "2", there is no use of putting words like "helllo" on the lists.
 * @param inputText - The text from which to remove repeat characters.
 * @param maxAllowedCharsInSequence - The maximum number of characters in sequence (such as "aaa", "bbb",
 * ...) that are allowed to remain in the input string. `Recommended: 2 or 3`, depending on the language
 * of your input text.
 * @returns The input text with all repeat characters that occur more than the max amount in sequence
 * removed.
 * @throws If `maxAllowedCharsInSequence` is not an integer (such as 1.5) or if it is 0 or less.
 */
export declare function reduceRepeatCharacters(inputText: string, maxAllowedCharsInSequence?: number): string;