to-romaji

A very basic kana to modern Hepburn-like rōmaji converter.

e.g. converts "にんじゃりばんばん" to "ninjaribanban".

Notes

• Handles repeated っ (e.g. "すっっっごくいい" → "suggggokuii", "あつっ" → "atsu'").

• Separates a final ん from a following vowel with a hyphen (e.g. かんいかきとめ → "kan-ikakitome"). Maybe we should use an apostrophe instead?

• Does not produce macrons for long vowels (長音) since it's too hard to distinguish between cases like みずうみ which should be "mizuumi" and すうがく which should be "sūgaku".

• Does not distinguish between は becoming "ha" and "wa". We tried to do this by using @birchill/tiny-segmenter to find 'は' particles but it turned up false positives on input like "おふろにはいる" (which it segments as "おふろ | に | は | いる").

• Does not convert half-width katakana. toNormalized from @birchill/normal-jp can be used as a pre-processing step to handle half-width katakana.

Installation

npm install @birchill/to-romaji

Usage

import { toRomaji } from '@birchill/to-romaji';

const romaji = toRomaji('にんじゃりばんばん');
// romaji = 'ninjaribanban'