n2words 1.24.0 → 3.0.0

package/lib/languages/lv.js

@@ -1,153 +1,321 @@
-import SlavicLanguage from '../classes/slavic-language.js'
+/**
+ * Latvian language converter - Functional Implementation
+ *
+ * A performance-optimized number-to-words converter using precomputed lookup tables.
+ *
+ * Key features:
+ * - Two-form pluralization (singular for 1 except 11, plural otherwise)
+ * - Gender agreement (masculine/feminine for numbers < 1000)
+ * - Special hundreds forms (simts/simti/simtu)
+ * - Omit "one" before scale words
+ * - Long scale naming
+ */
+
+import { parseNumericValue } from '../utils/parse-numeric.js'
+import { validateOptions } from '../utils/validate-options.js'
+
+// ============================================================================
+// Vocabulary (module-level constants)
+// ============================================================================
+
+const ONES_MASC = ['', 'viens', 'divi', 'trīs', 'četri', 'pieci', 'seši', 'septiņi', 'astoņi', 'deviņi']
+const ONES_FEM = ['', 'viena', 'divas', 'trīs', 'četras', 'piecas', 'sešas', 'septiņas', 'astoņas', 'deviņas']
+
+const TEENS = ['desmit', 'vienpadsmit', 'divpadsmit', 'trīspadsmit', 'četrpadsmit', 'piecpadsmit', 'sešpadsmit', 'septiņpadsmit', 'astoņpadsmit', 'deviņpadsmit']
+const TENS = ['', '', 'divdesmit', 'trīsdesmit', 'četrdesmit', 'piecdesmit', 'sešdesmit', 'septiņdesmit', 'astoņdesmit', 'deviņdesmit']
+
+// Hundreds: simts (100, 110-199), simti (200-999), simtu (101-109)
+const HUNDRED_SINGULAR = 'simts'
+const HUNDRED_PLURAL = 'simti'
+const HUNDRED_GENITIVE = 'simtu'
+
+const ZERO = 'nulle'
+const NEGATIVE = 'mīnus'
+const DECIMAL_SEP = 'komats'
+
+// Scale words: [singular, plural, genitive]
+const SCALE_FORMS = [
+  ['tūkstotis', 'tūkstoši', 'tūkstošu'],
+  ['miljons', 'miljoni', 'miljonu'],
+  ['miljards', 'miljardi', 'miljardu'],
+  ['triljons', 'triljoni', 'triljonu'],
+  ['kvadriljons', 'kvadriljoni', 'kvadriljonu'],
+  ['kvintiljons', 'kvintiljoni', 'kvintiljonu'],
+  ['sikstiljons', 'sikstiljoni', 'sikstiljonu'],
+  ['septiljons', 'septiljoni', 'septiljonu'],
+  ['oktiljons', 'oktiljoni', 'oktiljonu']
+]
+
+// ============================================================================
+// Precomputed Lookup Tables (built once at module load)
+// ============================================================================
 
 /**
- * @typedef {Object} SlavicOptions
- * @property {boolean} [feminine=false] Use feminine forms for numbers.
+ * Builds segment word for 0-999 (masculine form).
+ * Does NOT include special handling for segment=1 (omitOneBeforeScale).
+ * That's handled at join time.
  */
+function buildSegment (n) {
+  if (n === 0) return ''
+
+  const ones = n % 10
+  const tens = Math.floor(n / 10) % 10
+  const hundreds = Math.floor(n / 100)
+
+  const parts = []
+
+  // Hundreds - Latvian has special forms
+  if (hundreds > 0) {
+    if (hundreds === 1 && tens === 0 && ones > 0) {
+      // 101-109: use genitive form "simtu"
+      parts.push(HUNDRED_GENITIVE)
+    } else if (hundreds > 1) {
+      // 200-999: use plural "simti"
+      parts.push(ONES_MASC[hundreds])
+      parts.push(HUNDRED_PLURAL)
+    } else {
+      // 100, 110-199: use singular "simts"
+      parts.push(HUNDRED_SINGULAR)
+    }
+  }
+
+  // Tens
+  if (tens > 1) {
+    parts.push(TENS[tens])
+  }
+
+  // Teens or ones
+  if (tens === 1) {
+    parts.push(TEENS[ones])
+  } else if (ones > 0) {
+    parts.push(ONES_MASC[ones])
+  }
+
+  return parts.join(' ')
+}
 
 /**
- * Latvian language converter.
- *
- * Implements Latvian number words using the Slavic language pattern:
- * - Latvian number words (viens, divi, trīs, četri, pieci...)
- * - Latvian-specific pluralization patterns
- * - Baltic grammatical structure
- * - Simplified gender handling compared to Lithuanian
- *
- * Key Features:
- * - Three-form pluralization system shared across Slavic/Baltic languages
- *   * Form 1 (singular): 1 (e.g., "tūkstotis")
- *   * Form 2 (few): 2-4, 22-24, 32-34... excluding teens (e.g., "tūkstoši")
- *   * Form 3 (many): all other numbers (e.g., "tūkstošu")
- * - Chunk-based decomposition (splits into groups of 3 digits: ones, thousands, millions, etc.)
- * - Large number handling via thousands[] array with indexed [singular, few, many] forms
- *
- * Features:
- * - Latvian diacritical marks (ī, ā, ē, ū, etc.)
- * - Three-form pluralization (adapted for Latvian)
- * - Baltic number naming conventions
- * - Compound number formation (divdesmit, trīsdesmit)
- *
- * Inherits from SlavicLanguage for pluralization algorithms.
+ * Builds segment word for 0-999 (feminine form - only differs in ones).
  */
-export class Latvian extends SlavicLanguage {
-  negativeWord = 'mīnus'
-  decimalSeparatorWord = 'komats'
-  zeroWord = 'nulle'
-  ones = {
-    1: 'viens',
-    2: 'divi',
-    3: 'trīs',
-    4: 'četri',
-    5: 'pieci',
-    6: 'seši',
-    7: 'septiņi',
-    8: 'astoņi',
-    9: 'deviņi'
+function buildSegmentFeminine (n) {
+  if (n === 0) return ''
+
+  const ones = n % 10
+  const tens = Math.floor(n / 10) % 10
+  const hundreds = Math.floor(n / 100)
+
+  const parts = []
+
+  // Hundreds - always masculine
+  if (hundreds > 0) {
+    if (hundreds === 1 && tens === 0 && ones > 0) {
+      parts.push(HUNDRED_GENITIVE)
+    } else if (hundreds > 1) {
+      parts.push(ONES_MASC[hundreds])
+      parts.push(HUNDRED_PLURAL)
+    } else {
+      parts.push(HUNDRED_SINGULAR)
+    }
   }
 
-  tens = {
-    0: 'desmit',
-    1: 'vienpadsmit',
-    2: 'divpadsmit',
-    3: 'trīspadsmit',
-    4: 'četrpadsmit',
-    5: 'piecpadsmit',
-    6: 'sešpadsmit',
-    7: 'septiņpadsmit',
-    8: 'astoņpadsmit',
-    9: 'deviņpadsmit'
+  // Tens
+  if (tens > 1) {
+    parts.push(TENS[tens])
   }
 
-  twenties = {
-    2: 'divdesmit',
-    3: 'trīsdesmit',
-    4: 'četrdesmit',
-    5: 'piecdesmit',
-    6: 'sešdesmit',
-    7: 'septiņdesmit',
-    8: 'astoņdesmit',
-    9: 'deviņdesmit'
+  // Teens or ones - feminine for ones only
+  if (tens === 1) {
+    parts.push(TEENS[ones])
+  } else if (ones > 0) {
+    parts.push(ONES_FEM[ones])
   }
 
-  hundreds = ['simts', 'simti', 'simtu']
-
-  thousands = {
-    1: ['tūkstotis', 'tūkstoši', 'tūkstošu'],
-    2: ['miljons', 'miljoni', 'miljonu'],
-    3: ['miljards', 'miljardi', 'miljardu'],
-    4: ['triljons', 'triljoni', 'triljonu'],
-    5: ['kvadriljons', 'kvadriljoni', 'kvadriljonu'],
-    6: ['kvintiljons', 'kvintiljoni', 'kvintiljonu'],
-    7: ['sikstiljons', 'sikstiljoni', 'sikstiljonu'],
-    8: ['septiljons', 'septiljoni', 'septiljonu'],
-    9: ['oktiljons', 'oktiljoni', 'oktiljonu'],
-    10: ['nontiljons', 'nontiljoni', 'nontiljonu']
+  return parts.join(' ')
+}
+
+// Precompute all 1000 segment words (0-999)
+const SEGMENTS_MASC = new Array(1000)
+const SEGMENTS_FEM = new Array(1000)
+
+for (let i = 0; i < 1000; i++) {
+  SEGMENTS_MASC[i] = buildSegment(i)
+  SEGMENTS_FEM[i] = buildSegmentFeminine(i)
+}
+
+// ============================================================================
+// Pluralization
+// ============================================================================
+
+/**
+ * Latvian pluralization - simpler than Slavic.
+ * Singular: ends in 1 (except 11)
+ * Plural: everything else
+ *
+ * @param {number} n - The segment value
+ * @param {string[]} forms - [singular, plural, genitive]
+ * @returns {string} The appropriate form
+ */
+function pluralize (n, forms) {
+  if (n === 0) return forms[2]
+
+  const lastDigit = n % 10
+  const lastTwoDigits = n % 100
+
+  if (lastDigit === 1 && lastTwoDigits !== 11) {
+    return forms[0]
   }
 
-  pluralize (n, forms) {
-    if (n === 0n) {
-      return forms[2]
-    }
+  return forms[1]
+}
 
-    const lastDigit = n % 10n
-    const lastTwoDigits = n % 100n
+// ============================================================================
+// Conversion Functions
+// ============================================================================
 
-    if (lastDigit === 1n && lastTwoDigits !== 11n) {
-      return forms[0]
-    }
+/**
+ * Converts a non-negative integer to Latvian words.
+ *
+ * @param {bigint} n - Non-negative integer to convert
+ * @param {Object} options - Conversion options
+ * @returns {string} Latvian words
+ */
+function integerToWords (n, options = {}) {
+  if (n === 0n) return ZERO
 
-    return forms[1]
+  // Fast path: numbers < 1000 (direct lookup)
+  if (n < 1000n) {
+    const segments = options.gender === 'feminine' ? SEGMENTS_FEM : SEGMENTS_MASC
+    return segments[Number(n)]
   }
 
-  convertWholePart (number) {
-    if (number === 0n) {
-      return this.zeroWord
-    }
-    const words = []
-    const chunks = this.splitByX(number.toString(), 3)
-    let index = chunks.length
-    for (const x of chunks) {
-      index = index - 1
-      if (x === 0n) {
-        continue
-      }
-      const [n1, n2, n3] = this.getDigits(x)
-      if (n3 > 0n) {
-        if (n3 === 1n && n2 === 0n && n1 > 0n) {
-          words.push(this.hundreds[2])
-        } else if (n3 > 1n) {
-          words.push(this.ones[n3], this.hundreds[1])
+  // For numbers >= 1000, feminine only applies to final segment if < 1000
+  // But we use masculine for all segments when n >= 1000
+  return buildLargeNumberWords(n, options)
+}
+
+/**
+ * Builds words for numbers >= 1000.
+ *
+ * @param {bigint} n - Number >= 1000
+ * @param {Object} options - Conversion options
+ * @returns {string} Latvian words
+ */
+function buildLargeNumberWords (n, options) {
+  const numStr = n.toString()
+  const len = numStr.length
+
+  // Build segments of 3 digits from right to left
+  const segments = []
+  const segmentSize = 3
+
+  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) {
+      const segmentWord = SEGMENTS_MASC[segment]
+
+      if (scaleIndex === 0) {
+        // Units segment - use masculine (feminine doesn't apply when n >= 1000)
+        parts.push(segmentWord)
+      } else {
+        // Segment with scale word
+        const scaleForms = SCALE_FORMS[scaleIndex - 1]
+        const scaleWord = pluralize(segment, scaleForms)
+
+        // Latvian omits "one" before scale words
+        if (segment === 1) {
+          parts.push(scaleWord)
         } else {
-          words.push(this.hundreds[0])
+          parts.push(segmentWord + ' ' + scaleWord)
         }
       }
-      if (n2 > 1n) {
-        words.push(this.twenties[n2])
-      }
-      if (n2 === 1n) {
-        words.push(this.tens[n1])
-      } else if (n1 > 0n && !(index > 0 && x === 1n)) {
-        words.push(this.ones[n1])
-      }
-      if (index > 0) {
-        words.push(this.pluralize(x, this.thousands[index]))
-      }
     }
-    return words.join(' ')
+
+    scaleIndex--
+  }
+
+  return parts.join(' ')
+}
+
+/**
+ * Converts decimal digits to Latvian words.
+ *
+ * @param {string} decimalPart - Decimal digits (without the point)
+ * @param {Object} options - Conversion options
+ * @returns {string} Latvian words for decimal part
+ */
+function decimalPartToWords (decimalPart, options) {
+  let result = ''
+
+  // Handle leading zeros
+  let i = 0
+  while (i < decimalPart.length && decimalPart[i] === '0') {
+    if (result) result += ' '
+    result += ZERO
+    i++
+  }
+
+  // Convert remainder as a single number
+  const remainder = decimalPart.slice(i)
+  if (remainder) {
+    if (result) result += ' '
+    result += integerToWords(BigInt(remainder), options)
   }
+
+  return result
 }
 
 /**
- * Converts a number to Latvian cardinal (written) form.
+ * Converts a numeric value to Latvian 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 Latvian words
+ * @throws {TypeError} If value is not a valid numeric type
+ * @throws {Error} If value is not a valid number format
  *
- * @param {number|string|bigint} value The number to convert.
- * @param {Object} [options={}] Configuration options.
- * @param {boolean} [options.feminine=false] Use feminine forms for numbers.
- * @returns {string} The number expressed in Latvian words.
- * @throws {TypeError} If value is NaN or invalid type.
- * @throws {Error} If value is an invalid number string.
+ * @example
+ * toWords(42)                          // 'četrdesmit divi'
+ * toWords(1, { gender: 'feminine' })   // 'viena'
+ * toWords(1000)                        // 'tūkstotis'
  */
-export default function convertToWords (value, options = {}) {
-  return new Latvian(options).convertToWords(value)
+function toWords (value, options) {
+  options = validateOptions(options)
+  const { isNegative, integerPart, decimalPart } = parseNumericValue(value)
+
+  let result = ''
+
+  if (isNegative) {
+    result = NEGATIVE + ' '
+  }
+
+  result += integerToWords(integerPart, options)
+
+  if (decimalPart) {
+    result += ' ' + DECIMAL_SEP + ' ' + decimalPartToWords(decimalPart, options)
+  }
+
+  return result
 }
+
+// ============================================================================
+// Public API
+// ============================================================================
+
+export { toWords }

package/lib/languages/mr.d.ts

@@ -0,0 +1,7 @@
+/**
+ * Converts a numeric value to Marathi words.
+ *
+ * @param {number | string | bigint} value - The numeric value to convert
+ * @returns {string} The number in Marathi words
+ */
+export function toWords(value: number | string | bigint): string;