gretzky / cratylus

language processing functions

cratylus

Language processing tools. These can be used alongside i18n setups (like i18next) or on their own.

Cratylus is a dialog by Plato that debates the arbitrariness of language.

Usage

Install the package with the package manager of your choice.

// or yarn or pnpm
npm install @gretzky/cratylus

Levenshtein Distance

The Levenshtein distance is the minimum number of single-character edits required to change one word into another.

levenshtein(a: string, b: string, calcRatio: boolean): number

Returns the levenshtein distance calculation iteratively over 2 matrix rows.

levenshtein("crack", "whack"); // 2
levenshtein("8675309", "1234567"); // 6

Plurality

There are 2 enums and 2 functions for using plurals based off of the Arabic language set of 5 potential plural forms:

• None = 0
• One = 1
• Two = 2,
• Few = 3-5
• Many = 5-99
• Other = 100+

PluralForm and PluralGuard

Each enum is keyed by the 5 plural forms.

PluralForm returns the string name of the plural (e.g. none = none).

PluralGuard returns an arbitrary number that represents the plural (e.g. few = 3). This is more useful if you just need to determine the plural form and the actual number doesn't matter.

getPluralForm(count: number): PluralForm

Returns the string type of the plural based on the passed number.

getPluralForm(0); // "none"
getPluralForm(1); // "one"
getPluralForm(2); // "two"
getPluralForm(4); // "few"
getPluralForm(15); // "many"
getPluralForm(66); // "many"
getPluralForm(369); // "other"

Sentence

buildSentencePrimitive(stringArr: string[]): string

Returns a sentence built from an array of strings. Any punctuation in the array is adhered to, but if there's no end punctuation a period is added. Correct sentence casing is also added.

Consider an array to be 1 sentence.

const s1 = ["this", "is", "a", "sentence"];
const s2 = ["this is", "a sentence"];
const s3 = ["tHiS iS", "a SeNtEnCe"];
const s4 = ["this is", "a sentence", "!"];
const s5 = ["this", "is", "a sentence", "?"];

buildSentence(s1); // "This is a sentence."
buildSentence(s2); // "This is a sentence."
buildSentence(s3); // "This is a sentence."
buildSentence(s4); // "This is a sentence!"
buildSentence(s5); // "This is a sentence?"

String Formatting

TextCasing

Enum that represents 5 different ways of casing a string.

// capitalizes the first letter of a string, regardless of whether or not it has other words
TextCasing.capitalize; // e.g. Foo bar
// uppercases a string
TextCasing.upper; // e.g. FOO BAR
// lowercases a string
TextCasing.lower; // e.g. foo bar
// title cases a string, capitalizing the first letter of each individual word in the string
TextCasing.title; // e.g. Foo Bar
// sentence cases a string - very similar to capitalize but the regex is adjusted slightly
TextCasing.sentence; // e.g. Foo bar

stringFormat(value: string, format: TextCasing): string

Returns a string formatted to a given text casing.

const str = "never gonna give you up";

stringFormat(str, TextCasing.capitalize); // "Never gonna give you up"
stringFormat(str, TextCasing.sentence); // "Never gonna give you up"
stringFormat(str, TextCasing.title); // "Never Gonna Give You Up"
stringFormat(str, TextCasing.lower); // "never gonna give you up"
stringFormat(str, TextCasing.upper); // "NEVER GONNA GIVE YOU UP"

removeDiacritics(str: string): string

Finds and replaces diacritic characters in a string (e.g. Ä => A).

removeDiacritics("Älesund"); // Alesund