mirror of
https://github.com/hanzoai/i18n.git
synced 2026-10-30 14:24:32 +00:00
zeekay
066c5220dc
--mode madlad runs MADLAD-400 through CTranslate2 on this machine (tools/translate/madlad.py). It takes no instructions, so protected names cross it as numbered placeholders and are restored after. A label, a protected name with at most one other word and no slot, translates to itself: when an answer transliterates its name it is written as the source and recorded under a provenance run of model "source". Variant notes for the new locales go with the prompt.
40 lines
1.9 KiB
TypeScript
40 lines
1.9 KiB
TypeScript
import assert from 'node:assert/strict'
|
|
import { test } from 'node:test'
|
|
|
|
import { madlad, type Worker } from '../src/madlad.ts'
|
|
|
|
const terms = { keep: ['Claude Code', 'Claude', 'Hanzo', 'Cursor'] }
|
|
|
|
/** A worker that records what it was handed and answers with `reply`. */
|
|
function recording(reply: (texts: string[]) => string[]): { worker: Worker; asked: [string, string[]][] } {
|
|
const asked: [string, string[]][] = []
|
|
return { asked, worker: async (target, texts) => (asked.push([target, texts]), reply(texts)) }
|
|
}
|
|
|
|
test('protected names cross MADLAD as numbered placeholders and come back as written', async () => {
|
|
const { worker, asked } = recording((texts) => texts.map((t) => `T(${t})`))
|
|
const answer = await madlad(worker, terms)({ a: 'Accelerate Cursor with Hanzo', b: 'Claude Code, Claude', c: ' Start ' }, 'my')
|
|
assert.deepEqual(asked, [['my', ['Accelerate {0} with {1}', '{0}, {1}', 'Start']]])
|
|
assert.deepEqual(answer, { a: 'T(Accelerate Cursor with Hanzo)', b: 'T(Claude Code, Claude)', c: 'T(Start)' })
|
|
})
|
|
|
|
test('a numbered slot the source holds is not mistaken for a name', async () => {
|
|
const { worker, asked } = recording((texts) => texts)
|
|
const answer = await madlad(worker, terms)({ a: '{0} files in Hanzo' }, 'km')
|
|
assert.deepEqual(asked[0][1], ['{0} files in {1}'])
|
|
assert.deepEqual(answer, { a: '{0} files in Hanzo' })
|
|
})
|
|
|
|
test('the target is the language MADLAD names', async () => {
|
|
const { worker, asked } = recording((texts) => texts)
|
|
for (const locale of ['my', 'fa-AF', 'zh-HK', 'pa-PK', 'es-MX']) {
|
|
await madlad(worker)({ a: 'A' }, locale)
|
|
}
|
|
assert.deepEqual(asked.map(([target]) => target), ['my', 'prs', 'zh_Hant', 'pa', 'es'])
|
|
})
|
|
|
|
test('an answer that does not line up with the batch is an error', async () => {
|
|
const { worker } = recording(() => ['one'])
|
|
await assert.rejects(madlad(worker)({ a: 'A', b: 'B' }, 'my'), /1 translations for 2 strings/)
|
|
})
|