n2words 1.24.0 → 3.0.0

package/lib/languages/ko.js

@@ -1,83 +1,263 @@
-import GreedyScaleLanguage from '../classes/greedy-scale-language.js'
+/**
+ * Korean language converter - Functional Implementation
+ *
+ * A performance-optimized number-to-words converter using precomputed lookup tables.
+ *
+ * Key features:
+ * - Myriad-based (만) grouping - 4 digits
+ * - Implicit '일' (one) omission before scale words
+ * - Space separation after 만+ scales
+ * - Hangul numerals
+ */
+
+import { parseNumericValue } from '../utils/parse-numeric.js'
+
+// ============================================================================
+// Vocabulary (module-level constants)
+// ============================================================================
+
+const ONES = ['', '일', '이', '삼', '사', '오', '육', '칠', '팔', '구']
+
+const TEN = '십'
+const HUNDRED = '백'
+const THOUSAND = '천'
+
+const ZERO = '영'
+const NEGATIVE = '마이너스'
+const DECIMAL_SEP = '점'
+
+// Myriad scale words (powers of 10,000)
+// 만 (10^4), 억 (10^8), 조 (10^12), 경 (10^16), etc.
+const SCALES = ['만', '억', '조', '경', '해', '자', '양']
+
+// ============================================================================
+// Precomputed Lookup Tables (built once at module load)
+// ============================================================================
+
+/**
+ * Builds segment word for 0-9999 (4-digit myriad segment).
+ * Korean omits "일" before 십, 백, 천.
+ */
+function buildSegment (n) {
+  if (n === 0) return ''
+
+  const ones = n % 10
+  const tens = Math.floor(n / 10) % 10
+  const hundreds = Math.floor(n / 100) % 10
+  const thousands = Math.floor(n / 1000)
+
+  let result = ''
+
+  // Thousands
+  if (thousands > 0) {
+    if (thousands === 1) {
+      result += THOUSAND
+    } else {
+      result += ONES[thousands] + THOUSAND
+    }
+  }
+
+  // Hundreds
+  if (hundreds > 0) {
+    if (hundreds === 1) {
+      result += HUNDRED
+    } else {
+      result += ONES[hundreds] + HUNDRED
+    }
+  }
+
+  // Tens
+  if (tens > 0) {
+    if (tens === 1) {
+      result += TEN
+    } else {
+      result += ONES[tens] + TEN
+    }
+  }
+
+  // Ones
+  if (ones > 0) {
+    result += ONES[ones]
+  }
+
+  return result
+}
+
+// Precompute all 10000 segment words (0-9999) for myriad grouping
+const SEGMENTS = new Array(10000)
+
+for (let i = 0; i < 10000; i++) {
+  SEGMENTS[i] = buildSegment(i)
+}
+
+// ============================================================================
+// Conversion Functions
+// ============================================================================
+
+/**
+ * Converts a non-negative integer to Korean words.
+ *
+ * @param {bigint} n - Non-negative integer to convert
+ * @returns {string} Korean words
+ */
+function integerToWords (n) {
+  if (n === 0n) return ZERO
+
+  // Fast path: numbers < 10000 (direct lookup)
+  if (n < 10000n) {
+    return SEGMENTS[Number(n)]
+  }
+
+  // For numbers >= 10000, use myriad decomposition
+  return buildLargeNumberWords(n)
+}
+
+/**
+ * Builds words for numbers >= 10000.
+ * Uses myriad (만) grouping - 4 digits per segment.
+ *
+ * @param {bigint} n - Number >= 10000
+ * @returns {string} Korean words
+ */
+function buildLargeNumberWords (n) {
+  const numStr = n.toString()
+  const len = numStr.length
+
+  // Build segments of 4 digits from right to left
+  const segments = []
+  const segmentSize = 4
+
+  const remainderLen = len % segmentSize
+  let pos = 0
+  if (remainderLen > 0) {
+    segments.push(Number(numStr.slice(0, remainderLen)))
+    pos = remainderLen
+  }
+  while (pos < len) {
+    segments.push(Number(numStr.slice(pos, pos + segmentSize)))
+    pos += segmentSize
+  }
+
+  // Convert segments to words
+  const parts = []
+  let scaleIndex = segments.length - 1
+
+  for (let i = 0; i < segments.length; i++) {
+    const segment = segments[i]
+
+    if (segment !== 0) {
+      if (scaleIndex === 0) {
+        // Units segment (no scale word)
+        parts.push({ word: SEGMENTS[segment], isScale: false })
+      } else {
+        // Segment with scale word
+        const scaleWord = SCALES[scaleIndex - 1]
+
+        // Korean omits segment when it's 1 before scale words
+        if (segment === 1) {
+          parts.push({ word: scaleWord, isScale: true })
+        } else {
+          parts.push({ word: SEGMENTS[segment], isScale: false })
+          parts.push({ word: scaleWord, isScale: true })
+        }
+      }
+    }
+
+    scaleIndex--
+  }
+
+  // Join with Korean spacing rules
+  return joinKoreanParts(parts)
+}
 
 /**
- * Korean language converter.
+ * Joins parts with Korean spacing rules.
+ * - Concatenate without spaces within segments
+ * - Space after scale words before next number
  *
- * Features:
- * - Space-separated for large numbers (>= 만/10,000)
- * - Concatenated for smaller numbers
- * - Omits '일' (1) before multipliers
+ * @param {Array} parts - Parts with isScale metadata
+ * @returns {string} Joined string
  */
-export class Korean extends GreedyScaleLanguage {
-  negativeWord = '마이너스'
-  decimalSeparatorWord = '점'
-  zeroWord = '영'
-  scaleWordPairs = [
-    [10_000_000_000_000_000_000_000_000_000n, '양'],
-    [1_000_000_000_000_000_000_000_000n, '자'],
-    [100_000_000_000_000_000_000n, '해'],
-    [10_000_000_000_000_000n, '경'],
-    [1_000_000_000_000n, '조'],
-    [100_000_000n, '억'],
-    [10_000n, '만'],
-    [1000n, '천'],
-    [100n, '백'],
-    [10n, '십'],
-    [9n, '구'],
-    [8n, '팔'],
-    [7n, '칠'],
-    [6n, '육'],
-    [5n, '오'],
-    [4n, '사'],
-    [3n, '삼'],
-    [2n, '이'],
-    [1n, '일'],
-    [0n, '영']
-  ]
-
-  /**
-   * Merges two adjacent word-number pairs according to Korean grammar rules.
-   *
-   * Korean-specific rules:
-   * - Omits '일' (1) before multipliers <= 만 (10,000)
-   * - Concatenates without space for small numbers (< 만)
-   * - Separates with space for large numbers (>= 만)
-   * - Multiplies when right > left, adds when left > right
-   *
-   * @param {Object} leftPair The left operand as `{ word: number }`.
-   * @param {Object} rightPair The right operand as `{ word: number }`.
-   * @returns {Object} Merged pair with combined word and resulting number.
-   */
-  mergeScales (leftPair, rightPair) {
-    const leftWord = Object.keys(leftPair)[0]
-    const rightWord = Object.keys(rightPair)[0]
-    const leftNumber = Object.values(leftPair)[0] // BigInt
-    const rightNumber = Object.values(rightPair)[0] // BigInt
-
-    // Implicit "일": omit 1 before multipliers up to 만 (10,000)
-    if (leftNumber === 1n && rightNumber <= 10_000n) return rightPair
-    // Concatenate (no space) for small numbers less than 만
-    if (leftNumber < 10_000n && leftNumber > rightNumber) return { [`${leftWord}${rightWord}`]: leftNumber + rightNumber }
-    // Space-separate for large numbers (>= 만) when adding
-    if (leftNumber >= 10_000n && leftNumber > rightNumber) return { [`${leftWord} ${rightWord}`]: leftNumber + rightNumber }
-    // Multiply for all scale combinations
-    return { [`${leftWord}${rightWord}`]: leftNumber * rightNumber }
+function joinKoreanParts (parts) {
+  if (parts.length === 0) return ZERO
+  if (parts.length === 1) return parts[0].word
+
+  const result = []
+
+  for (let i = 0; i < parts.length; i++) {
+    const part = parts[i]
+    const prevPart = i > 0 ? parts[i - 1] : null
+
+    // Add space after scale words before next number
+    if (prevPart && prevPart.isScale && !part.isScale) {
+      result.push(' ')
+    }
+
+    result.push(part.word)
   }
+
+  return result.join('')
 }
 
 /**
- * Converts a number to Korean cardinal (written) form.
+ * Converts decimal digits to Korean words.
  *
- * @param {number|string|bigint} value The number to convert.
- * @param {Object} [options] Conversion options (see Korean class options).
- * @returns {string} The number expressed in Korean words.
- * @throws {TypeError} If value is NaN or invalid type.
- * @throws {Error} If value is an invalid number string.
+ * @param {string} decimalPart - Decimal digits (without the point)
+ * @returns {string} Korean words for decimal part (space-separated)
+ */
+function decimalPartToWords (decimalPart) {
+  const parts = []
+
+  // Handle leading zeros
+  let i = 0
+  while (i < decimalPart.length && decimalPart[i] === '0') {
+    parts.push(ZERO)
+    i++
+  }
+
+  // Convert remainder as a single number
+  const remainder = decimalPart.slice(i)
+  if (remainder) {
+    parts.push(integerToWords(BigInt(remainder)))
+  }
+
+  return parts.join(' ')
+}
+
+/**
+ * Converts a numeric value to Korean words.
+ *
+ * @param {number | string | bigint} value - The numeric value to convert
+ * @returns {string} The number in Korean words
+ * @throws {TypeError} If value is not a valid numeric type
+ * @throws {Error} If value is not a valid number format
  *
  * @example
- * convertToWords(42); // '사십이'
- * convertToWords(10001); // '만 일'
+ * toWords(21)         // '이십일'
+ * toWords(10000)      // '만'
+ * toWords(1000000)    // '백만'
  */
-export default function convertToWords (value, options = {}) {
-  return new Korean(options).convertToWords(value)
+function toWords (value) {
+  const { isNegative, integerPart, decimalPart } = parseNumericValue(value)
+
+  const parts = []
+
+  if (isNegative) {
+    parts.push(NEGATIVE)
+  }
+
+  parts.push(integerToWords(integerPart))
+
+  if (decimalPart) {
+    parts.push(DECIMAL_SEP)
+    parts.push(decimalPartToWords(decimalPart))
+  }
+
+  return parts.join(' ')
 }
+
+// ============================================================================
+// Public API
+// ============================================================================
+
+export { toWords }

package/lib/languages/lt.d.ts

@@ -0,0 +1,18 @@
+/**
+ * Converts a numeric value to Lithuanian words.
+ *
+ * @param {number | string | bigint} value - The numeric value to convert
+ * @param {Object} [options] - Conversion options
+ * @param {string} [options.gender='masculine'] - Gender for numbers < 1000
+ * @returns {string} The number in Lithuanian words
+ * @throws {TypeError} If value is not a valid numeric type
+ * @throws {Error} If value is not a valid number format
+ *
+ * @example
+ * toWords(42)                          // 'keturiasdešimt du'
+ * toWords(1, { gender: 'feminine' })   // 'viena'
+ * toWords(1000000)                     // 'vienas milijonas'
+ */
+export function toWords(value: number | string | bigint, options?: {
+    gender?: string | undefined;
+}): string;