@khgtrn/lib 1.0.0

package/dist/cjs/func.js

@@ -0,0 +1,574 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.isEmpty = isEmpty;
+exports.isNumber = isNumber;
+exports.vi2en = vi2en;
+exports.crlf2lf = crlf2lf;
+exports.removeNewline = removeNewline;
+exports.shuffleArray = shuffleArray;
+exports.objectValueToArray = objectValueToArray;
+exports.groupBy = groupBy;
+exports.removeByKey = removeByKey;
+exports.removeEmptyValue = removeEmptyValue;
+exports.randomString = randomString;
+exports.byte2hex = byte2hex;
+exports.uuid7bin = uuid7bin;
+exports.uuid7 = uuid7;
+exports.base64encode = base64encode;
+exports.base64decode = base64decode;
+exports.getObjectValue = getObjectValue;
+exports.ov = ov;
+exports.toInt = toInt;
+exports.numberToRoman = numberToRoman;
+exports.romanToNumber = romanToNumber;
+exports.isRomanNumber = isRomanNumber;
+exports.base64ToBlob = base64ToBlob;
+exports.downloadFile = downloadFile;
+exports.deepClone = deepClone;
+/**
+ * Checks whether a value is considered "empty".
+ *
+ * - string: empty after trimming.
+ * - number: equal to `0`.
+ * - boolean: `false`.
+ * - `null`/`undefined`: always empty.
+ * - array: length `0`.
+ * - other objects: no own enumerable keys.
+ * - anything else (function, symbol, ...): never empty.
+ *
+ * @param value - Value to check.
+ * @returns `true` if the value is considered empty.
+ */
+function isEmpty(value) {
+    if (typeof value === "string") {
+        return value.trim() === "";
+    }
+    else if (typeof value === "number") {
+        return value === 0;
+    }
+    else if (typeof value === "boolean") {
+        return !value;
+    }
+    else if (value === undefined || value === null) {
+        return true;
+    }
+    else if (Array.isArray(value)) {
+        return value.length === 0;
+    }
+    else if (typeof value === "object") {
+        return Object.keys(value).length === 0;
+    }
+    return false;
+}
+/**
+ * Checks whether a value is a valid number (not `NaN`, not `Infinity`).
+ * @param value - Value to check.
+ * @returns Type-guard: `true` if `value` is a finite number.
+ */
+function isNumber(value) {
+    return typeof value === "number" && Number.isFinite(value);
+}
+/**
+ * Converts Vietnamese diacritics to their plain ASCII equivalents
+ * (e.g. `"Điều chỉnh"` -> `"Dieu chinh"`).
+ *
+ * @param s - Input string.
+ * @returns The string with Vietnamese diacritics removed.
+ */
+function vi2en(s) {
+    return (s
+        // Decompose accented characters into base character + combining mark
+        .normalize("NFD")
+        // Strip all combining diacritical marks
+        .replace(/[\u0300-\u036f]/g, "")
+        // đ/Đ have no combining-mark decomposition, handle them separately
+        .replace(/đ/g, "d")
+        .replace(/Đ/g, "D"));
+}
+/**
+ * Normalizes Windows-style line endings (`\r\n`) to Unix-style (`\n`).
+ * @param value - Input string.
+ * @returns The string with all `\r\n` replaced by `\n`.
+ */
+function crlf2lf(value) {
+    return value.replace(/\r\n/g, "\n");
+}
+/**
+ * Removes every newline from a string, after normalizing line endings and
+ * trimming surrounding whitespace.
+ * @param value - Input string.
+ * @returns The string with all newlines removed.
+ */
+function removeNewline(value) {
+    return crlf2lf(value).trim().replace(/\n/g, "");
+}
+/**
+ * Shuffles an array in place using the Fisher-Yates algorithm.
+ * @param array - Array to shuffle (mutated directly).
+ * @returns The same array reference, shuffled.
+ */
+function shuffleArray(array) {
+    for (let i = array.length - 1; i > 0; i--) {
+        const j = Math.floor(Math.random() * (i + 1));
+        // Swap elements i and j
+        [array[i], array[j]] = [array[j], array[i]];
+    }
+    return array;
+}
+/**
+ * Collects an object's own enumerable values into an array.
+ * @param obj - Source object. `null`/`undefined` yields an empty array.
+ * @returns Array of the object's values.
+ */
+function objectValueToArray(obj) {
+    var arr = [];
+    for (var i in obj)
+        arr.push(obj[i]);
+    return arr;
+}
+/**
+ * Groups list items by the key returned from `fn`, similar to Lodash's
+ * `groupBy`. Keys are compared by their JSON representation, so they can be
+ * primitives or plain objects.
+ *
+ * @param list - Array-like list of items to group. `null`/`undefined` yields an empty array.
+ * @param fn - Maps an item to the value used to group it.
+ * @returns Array of groups, each an array of items sharing the same key.
+ */
+function groupBy(list, fn) {
+    if (!list)
+        return [];
+    var groups = {};
+    for (var i = 0; i < list.length; i++) {
+        var group = JSON.stringify(fn(list[i]));
+        if (group in groups) {
+            groups[group].push(list[i]);
+        }
+        else {
+            groups[group] = [list[i]];
+        }
+    }
+    return objectValueToArray(groups);
+}
+/**
+ * Remove keys from object or array
+ * @param objectOrArray Object or array
+ * @param keys Keys to remove
+ * @returns Object or array with keys removed
+ */
+function removeByKey(objectOrArray, keys) {
+    if (Array.isArray(objectOrArray)) {
+        return objectOrArray.map((item) => typeof item === "object" && item !== null ? removeByKey(item, keys) : item);
+    }
+    if (typeof objectOrArray === "object" && objectOrArray !== null) {
+        const entries = Object.entries(objectOrArray)
+            .filter(([key]) => !keys.includes(key))
+            .map(([key, value]) => {
+            if (Array.isArray(value) || (typeof value === "object" && value !== null)) {
+                return [key, removeByKey(value, keys)];
+            }
+            return [key, value];
+        });
+        return Object.fromEntries(entries);
+    }
+    // Primitives are returned as-is
+    return objectOrArray;
+}
+/**
+ * Remove empty values from object or array
+ * @param objectOrArray Object or array
+ * @param options Options:
+ * - removeNull: Remove null values. Default: true
+ * - removeUndefined: Remove undefined values. Default: true
+ * - removeEmptyString: Remove empty string values. Default: true
+ * @returns Object or array with empty values removed
+ */
+function removeEmptyValue(objectOrArray, options = {}) {
+    const { removeNull = true, removeUndefined = true, removeEmptyString = true } = options;
+    const shouldRemove = (value) => (removeNull && value === null) ||
+        (removeUndefined && value === undefined) ||
+        (removeEmptyString && value === "");
+    if (Array.isArray(objectOrArray)) {
+        const result = objectOrArray
+            .map((item) => {
+            if (Array.isArray(item) || (typeof item === "object" && item !== null)) {
+                return removeEmptyValue(item, {
+                    removeNull,
+                    removeUndefined,
+                    removeEmptyString,
+                });
+            }
+            return item;
+        })
+            .filter((item) => !shouldRemove(item));
+        return result;
+    }
+    if (typeof objectOrArray === "object" && objectOrArray !== null) {
+        const entries = Object.entries(objectOrArray)
+            .map(([key, value]) => {
+            if (Array.isArray(value) || (typeof value === "object" && value !== null)) {
+                return [
+                    key,
+                    removeEmptyValue(value, {
+                        removeNull,
+                        removeUndefined,
+                        removeEmptyString,
+                    }),
+                ];
+            }
+            return [key, value];
+        })
+            .filter(([, value]) => !shouldRemove(value));
+        return Object.fromEntries(entries);
+    }
+    return objectOrArray;
+}
+/**
+ * Generates a random string of the given length, optionally guaranteeing at
+ * least one character from each requested category ("upper", "lower",
+ * "number", "specific").
+ *
+ * @param length - Desired length of the resulting string.
+ * @param opt - Character categories that must each appear at least once.
+ * `"specific"` only counts as a required category when `specificChars` is
+ * non-empty; otherwise it's silently ignored (as if not requested at all).
+ * @param specificChars - Custom character set used for the `"specific"` category.
+ * @returns A random string of exactly `length` characters. If `opt` is empty
+ * (or every requested category ends up unusable), falls back to alphanumeric
+ * characters rather than returning an empty string.
+ * @throws {Error} If `length` is smaller than the number of required categories,
+ * since the "at least one of each" guarantee couldn't fit otherwise.
+ */
+function randomString(length, opt = [], specificChars = "") {
+    const upperChars = "ABCDEFGHIJKLMNOPQRSTUVWXYZ";
+    const lowerChars = "abcdefghijklmnopqrstuvwxyz";
+    const numbers = "0123456789";
+    const requiredCategories = new Set(opt);
+    // Without specificChars, "specific" can't be honored, so treat it as unset
+    if (!specificChars) {
+        requiredCategories.delete("specific");
+    }
+    if (length < requiredCategories.size) {
+        throw new Error(`length (${length}) is too small to include one of each required category: ${[...requiredCategories].join(", ")}`);
+    }
+    let result = [];
+    let allChars = "";
+    const randomOneChar = (chars) => chars.charAt(Math.floor(Math.random() * chars.length));
+    if (requiredCategories.has("upper")) {
+        allChars += upperChars;
+        // Guarantee at least one uppercase letter
+        result.push(randomOneChar(upperChars));
+    }
+    if (requiredCategories.has("lower")) {
+        allChars += lowerChars;
+        // Guarantee at least one lowercase letter
+        result.push(randomOneChar(lowerChars));
+    }
+    if (requiredCategories.has("number")) {
+        allChars += numbers;
+        // Guarantee at least one digit
+        result.push(randomOneChar(numbers));
+    }
+    if (requiredCategories.has("specific")) {
+        allChars += specificChars;
+        // Guarantee at least one character from the custom set
+        result.push(randomOneChar(specificChars));
+    }
+    // If no category applied (e.g. opt is empty, or only "specific" was
+    // requested without specificChars), fall back to alphanumeric so the
+    // result isn't silently empty.
+    if (!allChars) {
+        allChars = upperChars + lowerChars + numbers;
+    }
+    // Fill up the remaining characters to reach the requested length
+    for (let i = result.length; i < length; i++) {
+        result.push(randomOneChar(allChars));
+    }
+    // Shuffle so the guaranteed characters aren't always at the front
+    return shuffleArray(result).join("");
+}
+/**
+ * Formats a byte value (0-255) as a zero-padded, 2-digit lowercase hex string.
+ * Intended for internal use with well-formed byte values (e.g. from a
+ * `Uint8Array`); values outside 0-255 are not validated.
+ *
+ * @param b - Byte value, expected in the 0-255 range.
+ * @returns 2-character hex string, e.g. `"0f"`.
+ */
+function byte2hex(b) {
+    return ("0" + b.toString(16)).slice(-2);
+}
+/**
+ * Generates the raw 16 bytes of a UUIDv7 (timestamp + random, RFC 4122
+ * variant). Relies on the Web Crypto `crypto.getRandomValues` API, available
+ * in browsers and modern Node.js.
+ *
+ * @returns 16-byte `Uint8Array` encoding a UUIDv7.
+ */
+function uuid7bin() {
+    var bytes = new Uint8Array(16);
+    var now = Date.now();
+    // 48-bit timestamp (ms). Division (rather than a left shift) is used so
+    // values above the 32-bit range bitwise operators support are still split
+    // correctly; `&` truncates the fractional part left over from the division.
+    bytes[0] = (now / 0x10000000000) & 0xff;
+    bytes[1] = (now / 0x100000000) & 0xff;
+    bytes[2] = (now / 0x1000000) & 0xff;
+    bytes[3] = (now / 0x10000) & 0xff;
+    bytes[4] = (now / 0x100) & 0xff;
+    bytes[5] = now & 0xff;
+    // random 10 bytes
+    var rnd = new Uint8Array(10);
+    crypto.getRandomValues(rnd);
+    for (var i = 0; i < 10; i++) {
+        bytes[i + 6] = rnd[i];
+    }
+    // version 7
+    bytes[6] = (bytes[6] & 0x0f) | 0x70;
+    // variant RFC4122
+    bytes[8] = (bytes[8] & 0x3f) | 0x80;
+    return bytes;
+}
+/**
+ * Generates a UUIDv7 string (e.g. `"01890a5d-ac96-774b-bcce-b302099a8057"`).
+ * @returns UUIDv7 in the standard 8-4-4-4-12 hex string format.
+ */
+function uuid7() {
+    var bin = uuid7bin();
+    var hex = [];
+    for (var i = 0; i < bin.length; i++) {
+        hex.push(byte2hex(bin[i]));
+    }
+    return [
+        hex.slice(0, 4).join(""),
+        hex.slice(4, 6).join(""),
+        hex.slice(6, 8).join(""),
+        hex.slice(8, 10).join(""),
+        hex.slice(10).join(""),
+    ].join("-");
+}
+/**
+ * Encodes a string to base64 (UTF-8 bytes).
+ * @param str - String to encode.
+ * @returns Base64-encoded string.
+ */
+function base64encode(str) {
+    const bytes = new TextEncoder().encode(str);
+    // Spreading the whole byte array into String.fromCharCode blows the call
+    // stack for large inputs, so build the binary string in bounded chunks.
+    const chunkSize = 0x8000;
+    let binary = "";
+    for (let i = 0; i < bytes.length; i += chunkSize) {
+        binary += String.fromCharCode(...bytes.subarray(i, i + chunkSize));
+    }
+    return btoa(binary);
+}
+/**
+ * Decodes a base64 string back to its original (UTF-8) string.
+ * @param base64 - Base64-encoded string.
+ * @returns Decoded string.
+ */
+function base64decode(base64) {
+    return new TextDecoder().decode(Uint8Array.from(atob(base64), (c) => c.charCodeAt(0)));
+}
+/**
+ * Reads a nested property from an object using a dot-separated path.
+ * Example: `getObjectValue(user, 'abc.def')` returns `user.abc.def`.
+ *
+ * Equivalent to optional chaining (`object?.abc?.def`), for use on
+ * TypeScript versions below 3.7 (e.g. Angular versions below 9). On newer
+ * TypeScript versions, prefer optional chaining directly instead.
+ *
+ * @param object - Source object.
+ * @param path - Dot-separated property path.
+ * @returns The resolved value, or `null`/`undefined` if any segment is missing.
+ * @global
+ */
+function getObjectValue(object, path) {
+    if (!object)
+        return null;
+    return path
+        .split(".")
+        .reduce((acc, part) => acc && acc[part], object);
+}
+/**
+ * Short alias for {@link getObjectValue}.
+ * @param object - Source object.
+ * @param path - Dot-separated property path.
+ * @returns The resolved value, or `null`/`undefined` if any segment is missing.
+ */
+function ov(object, path) {
+    return getObjectValue(object, path);
+}
+/**
+ * Converts a value to an integer, similar to `parseInt`/`Number` but with an
+ * explicit fallback for empty values (see {@link isEmpty}).
+ *
+ * @param value - Value to convert.
+ * @param defaultValue - Value returned when `value` is empty. Defaults to `0`.
+ * @returns The parsed integer, `defaultValue` if `value` is empty, or `NaN`
+ * if `value` is a non-numeric, non-empty string.
+ */
+function toInt(value, defaultValue = 0) {
+    if (isEmpty(value))
+        return defaultValue;
+    return typeof value === "string" ? Number.parseInt(value.trim()) : Number(value);
+}
+/**
+ * Converts a positive integer to a Roman numeral. Only supports the standard
+ * range (1-3999); larger numbers produce a non-standard repeating "M" prefix.
+ *
+ * @param num - Number to convert.
+ * @returns Roman numeral string, or `""` if `num` is zero or negative.
+ */
+function numberToRoman(num) {
+    if (num <= 0)
+        return "";
+    const romanMap = [
+        { value: 1000, numeral: "M" },
+        { value: 900, numeral: "CM" },
+        { value: 500, numeral: "D" },
+        { value: 400, numeral: "CD" },
+        { value: 100, numeral: "C" },
+        { value: 90, numeral: "XC" },
+        { value: 50, numeral: "L" },
+        { value: 40, numeral: "XL" },
+        { value: 10, numeral: "X" },
+        { value: 9, numeral: "IX" },
+        { value: 5, numeral: "V" },
+        { value: 4, numeral: "IV" },
+        { value: 1, numeral: "I" },
+    ];
+    let result = "";
+    for (const item of romanMap) {
+        while (num >= item.value) {
+            result += item.numeral;
+            num -= item.value;
+        }
+    }
+    return result;
+}
+/**
+ * Converts a Roman numeral string to a number. Does not validate that the
+ * input is a well-formed Roman numeral (use {@link isRomanNumber} first if
+ * that matters) — unrecognized characters make the result `NaN`.
+ *
+ * @param roman - Roman numeral string (case-sensitive, uppercase letters).
+ * @returns The numeric value, or `NaN` if `roman` contains unrecognized characters.
+ */
+function romanToNumber(roman) {
+    const map = {
+        I: 1,
+        V: 5,
+        X: 10,
+        L: 50,
+        C: 100,
+        D: 500,
+        M: 1000,
+    };
+    let result = 0;
+    for (let i = 0; i < roman.length; i++) {
+        const current = map[roman.charAt(i)];
+        const next = map[roman.charAt(i + 1)];
+        if (next && current < next) {
+            result -= current;
+        }
+        else {
+            result += current;
+        }
+    }
+    return result;
+}
+/**
+ * Checks whether a string is a well-formed Roman numeral (1-3999), matching
+ * the standard subtractive notation. Case-insensitive.
+ *
+ * @param value - String to validate.
+ * @returns `true` if `value` is a valid Roman numeral.
+ */
+function isRomanNumber(value) {
+    if (!value)
+        return false;
+    const romanRegex = /^(M{0,4}(CM|CD|D?C{0,3})(XC|XL|L?X{0,3})(IX|IV|V?I{0,3}))$/;
+    return romanRegex.test(value.toUpperCase());
+}
+/**
+ * Converts a base64 string to a `Blob`. Browser-only (relies on `atob` and
+ * `Blob`).
+ *
+ * @param base64 - Base64-encoded content.
+ * @param mimeType - MIME type to assign to the resulting `Blob`.
+ * @param sliceSize - Chunk size (in decoded characters) used while building
+ * the byte arrays, to avoid excessive memory allocation for large inputs.
+ * @returns A `Blob` containing the decoded bytes.
+ */
+function base64ToBlob(base64, mimeType, sliceSize = 512) {
+    const byteCharacters = atob(base64);
+    const bytes = [];
+    for (let offset = 0; offset < byteCharacters.length; offset += sliceSize) {
+        const slice = byteCharacters.slice(offset, offset + sliceSize);
+        const byteNumbers = new Array(slice.length);
+        for (let i = 0; i < slice.length; i++) {
+            byteNumbers[i] = slice.charCodeAt(i);
+        }
+        const byte = new Uint8Array(byteNumbers);
+        bytes.push(byte);
+    }
+    return new Blob(bytes, { type: mimeType });
+}
+/**
+ * Triggers a browser download (or opens in a tab) for base64-encoded file
+ * content. Browser-only (relies on `document`, `URL.createObjectURL`).
+ *
+ * @param fileName - Suggested file name for the download.
+ * @param mimeType - MIME type/subtype (a bare subtype like `"pdf"` is
+ * expanded to `"application/pdf"`).
+ * @param base64Content - Base64-encoded file content.
+ * @param action - `"download"` saves the file, `"open"` opens it in the same
+ * tab, `"open_blank"` opens it in a new tab. Defaults to `"download"`.
+ */
+function downloadFile(fileName, mimeType, base64Content, action = "download") {
+    if (!mimeType.startsWith("application/")) {
+        mimeType = `application/${mimeType}`;
+    }
+    const blob = base64ToBlob(base64Content, mimeType);
+    const blobUrl = URL.createObjectURL(blob);
+    let link = document.createElement("a");
+    document.body.appendChild(link);
+    link.download = fileName;
+    if (action === "open_blank") {
+        link.target = "_blank";
+    }
+    else if (action === "open") {
+        link.target = "_self";
+    }
+    link.href = blobUrl;
+    link.click();
+    document.body.removeChild(link);
+}
+/**
+ * Recursively clones plain objects, arrays and `Date` instances.
+ *
+ * Note: `Date` has no own enumerable properties, so a plain for-in copy
+ * would silently turn every `Date` into an empty `{}` — it's special-cased
+ * here. Other special object types (`Map`, `Set`, `RegExp`, ...) are not
+ * handled and will also be cloned as plain `{}`.
+ *
+ * @param obj - Value to clone.
+ * @returns A deep copy of `obj` (primitives are returned as-is).
+ */
+function deepClone(obj) {
+    if (obj === null || typeof obj !== "object") {
+        return obj;
+    }
+    if (obj instanceof Date) {
+        return new Date(obj.getTime());
+    }
+    let clonedObj = Array.isArray(obj) ? [] : {};
+    for (let key in obj) {
+        if (obj.hasOwnProperty(key)) {
+            clonedObj[key] = deepClone(obj[key]);
+        }
+    }
+    return clonedObj;
+}

package/dist/cjs/index.d.ts

@@ -0,0 +1,4 @@
+export * from "./base-enum";
+export * from "./func";
+export * from "./number-to-words";
+export * from "./round";

package/dist/cjs/index.js

@@ -0,0 +1,20 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __exportStar = (this && this.__exportStar) || function(m, exports) {
+    for (var p in m) if (p !== "default" && !Object.prototype.hasOwnProperty.call(exports, p)) __createBinding(exports, m, p);
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+__exportStar(require("./base-enum"), exports);
+__exportStar(require("./func"), exports);
+__exportStar(require("./number-to-words"), exports);
+__exportStar(require("./round"), exports);

package/dist/cjs/number-to-words/helpers.d.ts

@@ -0,0 +1,17 @@
+import { NumberToWordsDefinition } from "./types";
+/** Spells out a non-negative integer using the given locale's rules. */
+export declare function convertIntegerPart(value: number, definition: NumberToWordsDefinition, num: number | string, locale: string): string;
+/** Spells out a fractional part one digit at a time, e.g. "05" -> "không năm". */
+export declare function convertFractionalPart(fractionalDigits: string, definition: NumberToWordsDefinition): string;
+/**
+ * Validates `num` and normalizes it to a plain (non-exponential, unsigned)
+ * numeric string with a `-` prefix kept only to signal the sign, e.g.
+ * `123` -> `"123"`, `"-1.50"` -> `"-1.50"`.
+ *
+ * A `number` is only accepted when it's an integer — a non-integer `number`
+ * has already gone through IEEE-754 float conversion by the time this
+ * function sees it (trailing zeros lost, rounding artifacts like
+ * `0.1 + 0.2`), so decimals must be passed as a string instead, which is
+ * read back digit-for-digit exactly as written.
+ */
+export declare function normalizeNumericInput(num: number | string): string;

package/dist/cjs/number-to-words/helpers.js

@@ -0,0 +1,72 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.convertIntegerPart = convertIntegerPart;
+exports.convertFractionalPart = convertFractionalPart;
+exports.normalizeNumericInput = normalizeNumericInput;
+/** Splits a non-negative integer into groups of 3 digits, most significant first. */
+function splitIntoGroups(num) {
+    if (num === 0)
+        return [0];
+    const groups = [];
+    let n = num;
+    while (n > 0) {
+        groups.unshift(n % 1000);
+        n = Math.floor(n / 1000);
+    }
+    return groups;
+}
+/** Spells out a non-negative integer using the given locale's rules. */
+function convertIntegerPart(value, definition, num, locale) {
+    if (value === 0)
+        return definition.zero;
+    const groups = splitIntoGroups(value);
+    const maxScaleIndex = groups.length - 1;
+    if (maxScaleIndex >= definition.scaleWords.length) {
+        throw new Error(`numberToWords: ${num} exceeds the supported range for locale "${locale}"`);
+    }
+    const segments = [];
+    let isLeadingGroup = true;
+    groups.forEach((group, i) => {
+        if (group === 0)
+            return;
+        const scaleIndex = maxScaleIndex - i;
+        const scaleWord = definition.scaleWords[scaleIndex];
+        const groupWords = definition.convertGroup(group, isLeadingGroup);
+        segments.push(scaleWord ? `${groupWords} ${scaleWord}` : groupWords);
+        isLeadingGroup = false;
+    });
+    return segments.join(" ");
+}
+/** Spells out a fractional part one digit at a time, e.g. "05" -> "không năm". */
+function convertFractionalPart(fractionalDigits, definition) {
+    return [...fractionalDigits].map((d) => definition.digits[Number(d)]).join(" ");
+}
+const numericStringPattern = /^-?\d+(\.\d+)?$/;
+/**
+ * Validates `num` and normalizes it to a plain (non-exponential, unsigned)
+ * numeric string with a `-` prefix kept only to signal the sign, e.g.
+ * `123` -> `"123"`, `"-1.50"` -> `"-1.50"`.
+ *
+ * A `number` is only accepted when it's an integer — a non-integer `number`
+ * has already gone through IEEE-754 float conversion by the time this
+ * function sees it (trailing zeros lost, rounding artifacts like
+ * `0.1 + 0.2`), so decimals must be passed as a string instead, which is
+ * read back digit-for-digit exactly as written.
+ */
+function normalizeNumericInput(num) {
+    if (typeof num === "number") {
+        if (!Number.isFinite(num)) {
+            throw new Error(`numberToWords: ${num} is not a finite number`);
+        }
+        if (!Number.isInteger(num)) {
+            throw new Error(`numberToWords: ${num} is a non-integer number, which may have lost precision as a float ` +
+                `(trailing zeros, rounding) — pass it as a string instead, e.g. "${num}"`);
+        }
+        return num.toString();
+    }
+    const trimmed = num.trim();
+    if (!numericStringPattern.test(trimmed)) {
+        throw new Error(`numberToWords: "${num}" is not a valid numeric string`);
+    }
+    return trimmed;
+}