Korean tokenizer plugin: garu-orama-tokenizer (1.9MB WASM analyzer)

[ongjin]

Hi Orama team! 👋

I built a Korean morphological tokenizer plugin for Orama: garu-orama-tokenizer.

Korean search with the default tokenizer has a real pain point — particles like 은/는/이/가/을/를 glue onto the noun, so searching for 학교 misses 학교에, 학교를, 학교가. And verb inflections like 먹었다 vs 먹는다 never match each other. This plugin runs real morphological analysis under the hood, powered by garu-ko (a 1.9MB WASM analyzer that runs in browsers, Node, and edge).

Usage follows the standard Orama pattern (same as @orama/tokenizers/mandarin):

import { create, insert, search } from '@orama/orama'
import { createTokenizer } from 'garu-orama-tokenizer'

const db = await create({
  schema: { title: 'string', body: 'string' },
  components: { tokenizer: await createTokenizer() }
})

Would you consider linking this from the Tokenizers docs page or your community integrations list? Happy to send a docs PR if that's easier.

Links:

• npm: https://www.npmjs.com/package/garu-orama-tokenizer
• GitHub: https://github.com/ongjin/garu/tree/main/integrations/orama-tokenizer
• Garu (the analyzer): https://github.com/ongjin/garu