@poppinss/intl-formatter

Memoized API for Intl (To be used within Node.js)

The intl-formatter package ships with the memoized version of the Intl API. Creating new instances of the new Intl.<AnyFormatter> is painfully slow (see benchmarks), and this package caches those instances for re-use.

• The API is 100% identical to the official spec, instead of writing new Intl.DateTimeFormat(), you write formatters.date(), and the rest is all the same.
• All arguments are deeply compared during memoization.

Usage

Install the package from the npm registry as follows:

npm i @poppinss/intl-formatter

# Yarn friends
yarn add @poppinss/intl-formatter

And use it as follows:

import formatters from '@poppinss/intl-formatter'

const amount = formatters
  .number('en-in', { style: 'currency', currency: 'INR' })
  .format(10)

console.log(amount)

Available formatters

• formatters.number same as Intl.NumberFormat
• formatters.date same as Intl.DateTimeFormat
• formatters.relative same as Intl.RelativeTimeFormat
• formatters.plural same as Intl.PluralRules
• formatters.list same as Intl.ListFormat
• formatters.displayNames same as Intl.DisplayNames

Why not use FormatJS?

FormatJS is a great and popular library for Internationalization. However, it has a large set of polyfills for different platforms (especially for browsers) that do not have complete support for Intl.

This package relies on the native Intl APIs available in Node.js runtime and caches the instances for re-use and performance.

Benchmarks

DateTimeFormat

DateTimeFormat@memoize x 1,115,507 ops/sec ±0.22% (94 runs sampled)
DateTimeFormat x 20,042 ops/sec ±20.76% (75 runs sampled)

Fastest is DateTimeFormat@memoize

NumberFormat

NumberFormat@memoize x 2,874,842 ops/sec ±0.28% (97 runs sampled)
NumberFormat x 74,720 ops/sec ±1.14% (97 runs sampled)

Fastest is NumberFormat@memoize

PluralRules

PluralRules@memoize x 2,381,739 ops/sec ±0.63% (97 runs sampled)
PluralRules x 62,113 ops/sec ±2.88% (91 runs sampled)

Fastest is PluralRules@memoize

RelativeTimeFormat

RelativeTimeFormat@memoize x 2,426,178 ops/sec ±1.11% (92 runs sampled)
RelativeTimeFormat x 89,485 ops/sec ±3.48% (84 runs sampled)

Fastest is RelativeTimeFormat@memoize

ListFormat

ListFormat@memoize x 1,501,976 ops/sec ±1.94% (97 runs sampled)
ListFormat x 244,943 ops/sec ±0.93% (97 runs sampled)

Fastest is ListFormat@memoize

DisplayNames

DisplayNames@memoize x 299,573 ops/sec ±0.56% (93 runs sampled)
DisplayNames x 106,279 ops/sec ±1.32% (94 runs sampled)

Fastest is DisplayNames@memoize