@zsnout/ithkuil/parse/transform.js

A set of tools which can generate and parse romanized Ithkuil text and which can generate Ithkuil script from text and JSON data.

import { deepFreezeAndNullPrototype } from "../generate/index.js";
export const STRESSED_TO_UNSTRESSED_VOWEL_MAP = 
/* @__PURE__ */ deepFreezeAndNullPrototype({
    á: "a",
    é: "e",
    í: "i",
    ó: "o",
    ú: "u",
    â: "ä",
    ê: "ë",
    ô: "ö",
    û: "ü",
});
// Taken from https://github.com/ngoriyasjil/IthkuilGloss/blob/181241b89c962d83b999a669c298366b07df53b9/src/ithkuil/iv/gloss/Constants.kt#L27C4-L27C4
const LETTER_SUBSTITUTIONS = /* @__PURE__ */ deepFreezeAndNullPrototype({
    "​": "",
    // The previous line is keyed with the Unicode Byte Order Mark
    "’": "'",
    ʼ: "'",
    "‘": "'",
    á: "á",
    ä: "ä",
    â: "â",
    é: "é",
    ë: "ë",
    ê: "ê",
    ì: "i",
    ı: "i",
    ì: "i",
    í: "í",
    ó: "ó",
    ö: "ö",
    ô: "ô",
    ù: "u",
    ù: "u",
    ú: "ú",
    ü: "ü",
    û: "û",
    č: "č",
    ç: "ç",
    ṭ: "ţ",
    ŧ: "ţ",
    ț: "ţ",
    ţ: "ţ",
    ṭ: "ţ",
    ḍ: "ḑ",
    đ: "ḑ",
    ḍ: "ḑ",
    ḑ: "ḑ",
    ł: "ļ",
    ḷ: "ļ",
    ḷ: "ļ",
    ļ: "ļ",
    š: "š",
    ž: "ž",
    ż: "ż",
    ẓ: "ż",
    ẓ: "ż",
    ṇ: "ň",
    ň: "ň",
    ņ: "ň",
    ṇ: "ň",
    ṛ: "ř",
    ř: "ř",
    ŗ: "ř",
    r͕: "ř",
    ŗ: "ř",
    ṛ: "ř",
});
const LETTER_SUBSTITUTION_REGEX = 
// The first element of this character class is the Unicode Byte Order Mark
/[​’ʼ‘ìıùṭŧțḍđłḷẓṇṛŗ]|á|ä|â|é|ë|ê|ì|í|ó|ö|ô|ù|ú|ü|û|č|ç|ţ|ṭ|ḍ|ḑ|ḷ|ļ|š|ž|ż|ẓ|ň|ņ|ṇ|ř|ŗ|r͕|ṛ/gu;
/**
 * Transforms a word by normalizing spelling and parsing and removing stress
 * markings.
 *
 * @param word The word to be transformed.
 * @returns An object containing information about the transformed word.
 */
export function transformWord(word) {
    const original = word;
    word = word
        .toLowerCase()
        .replace(LETTER_SUBSTITUTION_REGEX, (x) => LETTER_SUBSTITUTIONS[x]);
    if (word.startsWith("'")) {
        word = word.slice(1);
    }
    const syllables = word.match(/[aáeéëêoóuú][ií]|[aáeéëêiíoó][uú]|[aeiouäëöüáéíóúâêôû]/g);
    let stress;
    if (syllables == null) {
        stress = "zerosyllabic";
    }
    else if (syllables.length == 1) {
        stress = "monosyllabic";
    }
    else {
        const stressed = syllables.map((syllable) => /[áéíóúâêôû]/.test(syllable));
        const index = stressed.findIndex((x) => x);
        const lastIndex = stressed.findLastIndex((x) => x);
        if (index != lastIndex) {
            throw new Error("Two syllables are marked as stressed.");
        }
        if (index == -1) {
            stress = "penultimate";
        }
        else {
            const value = stressed.length - index;
            if (value == 1) {
                stress = "ultimate";
            }
            else if (value == 2) {
                stress = "penultimate";
            }
            else if (value == 3) {
                stress = "antepenultimate";
            }
            else {
                throw new Error("Invalid stress in '" + word + "'.");
            }
        }
    }
    return Object.freeze({
        original,
        word: word.replace(/[áéíóúâêôû]/g, (x) => STRESSED_TO_UNSTRESSED_VOWEL_MAP[x]),
        stress,
    });
}
/**
 * Transforms a word by normalizing spelling. Does not remove stress markings.
 *
 * @param word The word to be transformed.
 * @returns An object containing information about the transformed word.
 */
export function transformWordButLeaveStressMarkings(word) {
    const original = word;
    word = word
        .toLowerCase()
        .replace(LETTER_SUBSTITUTION_REGEX, (x) => LETTER_SUBSTITUTIONS[x]);
    if (word.startsWith("'")) {
        word = word.slice(1);
    }
    return Object.freeze({
        original,
        word,
    });
}