Skip to Content
Introduction

rosetta-date

A translation engine for date-formatting languages.

rosetta-date translates date-format strings between formatting languages by routing every token through a canonical intermediate representation before rendering it into the target language.

import { convert } from 'rosetta-date'
import { ldml, moment } from 'rosetta-date/dialects'
 
convert('YYYY-MM-DD', { from: moment, to: ldml }) // 'yyyy-MM-dd'

Although YYYY-MM-DD and yyyy-MM-dd look different, they express the same date format using different formatting languages. rosetta-date understands the difference and rewrites the tokens — not the date.

Why it exists

Moving a format string from one date library to another is deceptively error-prone. YYYY means a calendar year in Moment.js but a week-numbering year in LDML; D is a day of month in one and a day of year in another. Hand-translating these tables is exactly the kind of work that looks trivial and quietly ships bugs. Migrating a Moment.js format to date-fns is one call:

import { convert } from 'rosetta-date'
import { dateFns, momentjs } from 'rosetta-date/libraries'
 
convert('dddd, MMMM Do YYYY', { from: momentjs, to: dateFns }) // 'EEEE, MMMM do yyyy'

rosetta-date encodes the full token grammar once, in one place. A token that converts cleanly round-trips faithfully; one with no clean counterpart is handled by a policy you control — preserved as a safe literal by default.

Highlights

  • 🔁 Bidirectional — every mapping works in both directions.
  • 🧭 Canonical semantic model — tokens convert through a shared meaning, never dialect-to-dialect.
  • 🧩 Extensible by design — a new dialect or library connects to the hub with two mappings, not a fleet of pairwise converters.
  • 🪶 Zero runtime dependencies.
  • 🌳 Tree-shakeable — dialects and libraries ship from their own entrypoints.
  • 🛡️ Escape-aware tokenizer — literals round-trip intact across dialects.

Where to go next

Last updated on
Installation