Table of contents

• @poppinss/intl-formatter
  • Usage
  • Available formatters
  • Why not use FormatJS?
  • Benchmarks

@poppinss/intl-formatter

Memoized API for Intl (To be used within Node)

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 just caches those instances for re-use.

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

Usage

Install the package from 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

Why not use FormatJS?

FormatJS is a great and a popular library for Internationalization. However, it comes with large polyfills for platforms (especially certain browser) that does not have complete support for Intl.

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

Benchmarks

DateTimeFormat

DateTimeFormat@memoize x 1,031,069 ops/sec ±0.22% (96 runs sampled)
DateTimeFormat x 16,338 ops/sec ±16.30% (82 runs sampled)

Fastest is DateTimeFormat@memoize

NumberFormat

NumberFormat@memoize x 2,740,775 ops/sec ±0.29% (94 runs sampled)
NumberFormat x 67,829 ops/sec ±1.75% (95 runs sampled)

Fastest is NumberFormat@memoize

PluralRules

PluralRules@memoize x 2,240,552 ops/sec ±0.22% (91 runs sampled)
PluralRules x 55,671 ops/sec ±5.13% (92 runs sampled)

Fastest is PluralRules@memoize

RelativeTimeFormat

RelativeTimeFormat@memoize x 2,344,764 ops/sec ±0.24% (96 runs sampled)
RelativeTimeFormat x 79,338 ops/sec ±4.08% (83 runs sampled)

Fastest is RelativeTimeFormat@memoize