talisman/phonetics/eudex.js

Straightforward fuzzy matching, information retrieval and NLP building blocks for JavaScript.

'use strict';

Object.defineProperty(exports, "__esModule", {
  value: true
});
exports.default = eudex;

var _long = require('long');

var _long2 = _interopRequireDefault(_long);

function _interopRequireDefault(obj) { return obj && obj.__esModule ? obj : { default: obj }; }

/**
 * Maps.
 */
var binary = function binary(str) {
  return typeof str === 'string' ? parseInt(str, 2) : str;
},
    charCode = function charCode(char) {
  return char.charCodeAt(0);
},
    charArray = function charArray(str) {
  return str.split('').map(charCode);
}; /* eslint no-confusing-arrow: 0 */
/**
 * Talisman phonetics/eudex
 * =========================
 *
 * JavaScript implementation of the Eudex phonetic hashing algorithm.
 *
 * [Reference]:
 * https://github.com/ticki/eudex
 *
 * [Author]:
 * @ticki (https://github.com/ticki)
 */


var PHONES = ['0', // a
// +--------- Confident
// |+-------- Labial
// ||+------- Liquid
// |||+------ Dental
// ||||+----- Plosive
// |||||+---- Fricative
// ||||||+--- Nasal
// |||||||+-- Discriminant
// ||||||||
'01001000', // b
'00001100', // c
'00011000', // d
'0', // e
'01000100', // f
'00001000', // g
'00000100', // h
'1', // i
'00000101', // j
'00001001', // k
'10100000', // l
'00000010', // m
'00010010', // n
'0', // o
'01001001', // p
'10101000', // q
'10100001', // r
'00010100', // s
'00011101', // t
'1', // u
'01000101', // v
'00000000', // w
'10000100', // x
'1', // y
'10010100'].map(binary);

var LETTERS = PHONES.length;

var PHONES_C1 = [PHONES[charCode('s') - charCode('a')] ^ 1, // ß
'0', // à
'0', // á
'0', // â
'0', // ã
'0', // ä [æ]
'1', // å [oː]
'0', // æ [æ]
PHONES[charCode('z') - charCode('a')] ^ 1, // ç [t͡ʃ]
'1', // è
'1', // é
'1', // ê
'1', // ë
'1', // ì
'1', // í
'1', // î
'1', // ï
'00010101', // ð [ð̠] (represented as a non-plosive T)
'00010111', // ñ [nj] (represented as a combination of n and j)
'0', // ò
'0', // ó
'0', // ô
'0', // õ
'1', // ö [ø]
'1', // ÷
'1', // ø [ø]
'1', // ù
'1', // ú
'1', // û
'1', // ü
'1', // ý
'00010101', // þ [ð̠] (represented as a non-plosive T)
'1'].map(binary);

var INJECTIVE_PHONES = [
// +--------- Vowel
// |+-------- Closer than ɜ
// ||+------- Close
// |||+------ Front
// ||||+----- Close-mid
// |||||+---- Central
// ||||||+--- Open-mid
// |||||||+-- Discriminant
// ||||||||   (*=vowel)
'10000100', // a*
'00100100', // b
'00000110', // c
'00001100', // d
'11011000', // e*
'00100010', // f
'00000100', // g
'00000010', // h
'11111000', // i*
'00000011', // j
'00000101', // k
'01010000', // l
'00000001', // m
'00001001', // n
'10010100', // o*
'00100101', // p
'01010100', // q
'01010001', // r
'00001010', // s
'00001110', // t
'11100000', // u*
'00100011', // v
'00000000', // w
'01000010', // x
'11100100', // y*
'01001010'].map(binary);

var INJECTIVE_PHONES_C1 = [INJECTIVE_PHONES[charCode('s') - charCode('a')] ^ 1, // ß
INJECTIVE_PHONES[charCode('a') - charCode('a')] ^ 1, // à
INJECTIVE_PHONES[charCode('a') - charCode('a')] ^ 1, // á
// +--------- Vowel
// |+-------- Closer than ɜ
// ||+------- Close
// |||+------ Front
// ||||+----- Close-mid
// |||||+---- Central
// ||||||+--- Open-mid
// |||||||+-- Discriminant
// ||||||||
'10000000', // â
'10000110', // ã
'10100110', // ä [æ]
'11000010', // å [oː]
'10100111', // æ [æ]
'01010100', // ç [t͡ʃ]
INJECTIVE_PHONES[charCode('e') - charCode('a')] ^ 1, // è
INJECTIVE_PHONES[charCode('e') - charCode('a')] ^ 1, // é
INJECTIVE_PHONES[charCode('e') - charCode('a')] ^ 1, // ê
'11000110', // ë [ə] or [œ]
INJECTIVE_PHONES[charCode('i') - charCode('a')] ^ 1, // ì
INJECTIVE_PHONES[charCode('i') - charCode('a')] ^ 1, // í
INJECTIVE_PHONES[charCode('i') - charCode('a')] ^ 1, // î
INJECTIVE_PHONES[charCode('i') - charCode('a')] ^ 1, // ï
'00001011', // ð [ð̠] (represented as a non-plosive T)
'00001011', // ñ [nj] (represented as a combination of n and j)
INJECTIVE_PHONES[charCode('o') - charCode('a')] ^ 1, // ò
INJECTIVE_PHONES[charCode('o') - charCode('a')] ^ 1, // ó
INJECTIVE_PHONES[charCode('o') - charCode('a')] ^ 1, // ô
INJECTIVE_PHONES[charCode('o') - charCode('a')] ^ 1, // õ
'11011100', // ö [œ] or [ø]
'1', // ÷
'11011101', // ø [œ] or [ø]
INJECTIVE_PHONES[charCode('u') - charCode('a')] ^ 1, // ù
INJECTIVE_PHONES[charCode('u') - charCode('a')] ^ 1, // ú
INJECTIVE_PHONES[charCode('u') - charCode('a')] ^ 1, // û
INJECTIVE_PHONES[charCode('y') - charCode('a')] ^ 1, // ü
INJECTIVE_PHONES[charCode('y') - charCode('a')] ^ 1, // ý
'00001011', // þ [ð̠] (represented as a non-plosive T)
INJECTIVE_PHONES[charCode('y') - charCode('a')] ^ 1].map(binary);

/**
 * Function taking a single word and computing its Eudex hash.
 *
 * @param  {string} word - The word to process.
 * @return {number}      - The Eudex hash.
 */
var A = charCode('a'),
    Z = charCode('z');

function eudex(word) {
  var array = charArray(word);

  var entry = array.length > 0 ? (array[0] | 32) - A & 0xFF : 0;

  var firstByte = 0;

  if (entry < LETTERS) firstByte = INJECTIVE_PHONES[entry];else if (entry >= 0xDF && entry < 0xFF) firstByte = INJECTIVE_PHONES_C1[entry - 0xDF];

  firstByte = new _long2.default(firstByte);

  var res = _long2.default.UZERO,
      n = 0,
      b = 1;

  while (n < 8 && b < array.length) {
    entry = (array[b] | 32) - A & 0xFF;

    if (entry <= Z) {
      var x = void 0;

      if (entry < LETTERS) x = PHONES[entry];else if (entry >= 0xDF && entry < 0xFF) x = PHONES_C1[entry - 0xDF];else continue;

      if (!res.and(0xFE).equals(x & 0xFE)) {
        res = res.shiftLeft(8);
        res = res.or(x);
        n++;
      }
    }

    b++;
  }

  return res.or(firstByte.shiftLeft(56));
}
module.exports = exports['default'];