d3ro-voice/apps/desktop/tests/main/services/VoiceModeService.test.ts
Yun Chan 983c60cda2 feat(bootstrap): Whisper large-v3-turbo 기본 전환 + 온보딩 2단계 다운로드 진행률
- 기본 STT 모델 base → large-v3-turbo (6배 빠름, 1.6GB)
- 사이드카: /download, /download/status, /download/cancel + --models-dir
- LocalSTTService: downloadModel/cancelDownload + download-progress 이벤트
- IPC: 설계서 02의 stt:downloadModel/cancelDownload/downloadProgress 구현
- OnboardingModal: LLM(gemma4:e4b) → STT(turbo) 2단계 순차 다운로드 UI
- SettingsModal turbo 선택지 + settings.model.largeTurbo 12 locale
- 테스트: 모노레포 잔재 import 수정 (src/shared → @d3ro/core), 41/41 통과
2026-07-21 11:59:49 +09:00

202 lines
5.9 KiB
TypeScript

// tests/main/services/VoiceModeService.test.ts
// 상태 머신 전이 + 이중 조건 플러시 + accidentalPress 테스트
import { describe, it, expect, beforeEach, vi } from 'vitest'
import { RecognitionState, AudioState } from '@d3ro/core/types'
import { TIMING } from '@d3ro/core/constants'
// 모든 하위 서비스 모킹
vi.mock('../../../src/main/services/LoggerService', () => ({
getLogger: () => ({
info: vi.fn(),
warn: vi.fn(),
error: vi.fn(),
debug: vi.fn()
})
}))
const mockSTT = {
initialize: vi.fn(() => Promise.resolve()),
transcribe: vi.fn(() =>
Promise.resolve({ text: '테스트 전사', segments: [], language: 'ko', duration: 2, processingTime: 500 })
),
getStatus: vi.fn(() => ({ state: 'ready', modelId: 'base', uptime: 0 })),
on: vi.fn(),
off: vi.fn()
}
vi.mock('../../../src/main/services/LocalSTTService', () => ({
getLocalSTTService: () => mockSTT
}))
const mockAudio = {
start: vi.fn(() => Promise.resolve()),
stop: vi.fn(() => Promise.resolve()),
on: vi.fn(),
off: vi.fn()
}
vi.mock('../../../src/main/services/AudioCaptureService', () => ({
getAudioCaptureService: () => mockAudio
}))
const mockHotkey = {
on: vi.fn(),
off: vi.fn()
}
vi.mock('../../../src/main/services/HotkeyService', () => ({
getHotkeyService: () => mockHotkey
}))
vi.mock('../../../src/main/services/ConfigService', () => ({
configGet: vi.fn((key: string) => {
const defaults: Record<string, unknown> = {
sttModelId: 'base',
defaultLLMAction: 'refine',
ollamaServerUrl: 'http://localhost:11434',
llmModelId: 'gemma4:e4b'
}
return defaults[key]
})
}))
const mockTextInsert = {
insertText: vi.fn(() => Promise.resolve({ success: true, method: 'clipboard', textLength: 10, durationMs: 50 }))
}
vi.mock('../../../src/main/services/TextInsertService', () => ({
getTextInsertService: () => mockTextInsert
}))
const mockLLM = {
isAvailable: vi.fn(() => false),
processText: vi.fn(() => Promise.resolve('다듬어진 텍스트')),
on: vi.fn(),
off: vi.fn()
}
vi.mock('../../../src/main/services/LocalLLMService', () => ({
getLocalLLMService: () => mockLLM
}))
let getVoiceModeService: () => ReturnType<typeof import('../../../src/main/services/VoiceModeService')['getVoiceModeService']>
beforeEach(async () => {
vi.resetModules()
vi.clearAllMocks()
const mod = await import('../../../src/main/services/VoiceModeService')
getVoiceModeService = mod.getVoiceModeService
})
describe('VoiceModeService', () => {
describe('상태 머신', () => {
it('초기 상태는 IDLE이다', () => {
const svc = getVoiceModeService()
const state = svc.getState()
expect(state.recognitionState).toBe(RecognitionState.IDLE)
expect(state.audioState).toBe(AudioState.IDLE)
expect(state.sessionId).toBeNull()
})
it('startSession 호출 시 PREPARING으로 전이한다', async () => {
const svc = getVoiceModeService()
const stateChanges: RecognitionState[] = []
svc.on('recognition-state-changed', (payload: { current: RecognitionState }) => {
stateChanges.push(payload.current)
})
await svc.startSession('dictation')
// PREPARING → CONNECTING → READY 순서
expect(stateChanges[0]).toBe(RecognitionState.PREPARING)
expect(stateChanges).toContain(RecognitionState.CONNECTING)
})
it('isActive는 세션이 활성일 때 true이다', async () => {
const svc = getVoiceModeService()
expect(svc.isActive).toBe(false)
// startSession은 완전 비동기이므로 await 후 세션 활성 확인
await svc.startSession('dictation')
expect(svc.isActive).toBe(true)
})
})
describe('accidentalPress', () => {
it('700ms 미만 세션은 자동 취소된다', async () => {
const svc = getVoiceModeService()
let cancelReason: string | null = null
svc.on('session-cancelled', (payload: { reason: string }) => {
cancelReason = payload.reason
})
// 세션 시작 즉시 종료 (700ms 미만)
await svc.startSession('dictation')
await svc.stopSession()
expect(cancelReason).toBe('too-short')
})
})
describe('cancelSession', () => {
it('user 취소로 세션을 종료한다', async () => {
const svc = getVoiceModeService()
let cancelReason: string | null = null
svc.on('session-cancelled', (payload: { reason: string }) => {
cancelReason = payload.reason
})
await svc.startSession('dictation')
svc.cancelSession()
expect(cancelReason).toBe('user')
})
})
describe('getState', () => {
it('현재 상태를 VoiceState 형태로 반환한다', () => {
const svc = getVoiceModeService()
const state = svc.getState()
expect(state).toHaveProperty('recognitionState')
expect(state).toHaveProperty('audioState')
expect(state).toHaveProperty('mode')
expect(state).toHaveProperty('sessionId')
expect(state).toHaveProperty('recordingStartedAt')
})
})
describe('터미널 상태', () => {
it('cancelSession 후 _resetToIdle의 200ms 딜레이 후 IDLE로 전이한다', async () => {
vi.useFakeTimers()
const svc = getVoiceModeService()
await svc.startSession('dictation')
svc.cancelSession()
// 200ms 딜레이로 IDLE 전이 예약됨
vi.advanceTimersByTime(250)
const state = svc.getState()
expect(state.recognitionState).toBe(RecognitionState.IDLE)
vi.useRealTimers()
})
})
})
describe('TIMING 상수', () => {
it('핵심 타이밍 값이 Speakly 패턴과 일치한다', () => {
expect(TIMING.MIN_AUDIO_DURATION).toBe(700)
expect(TIMING.DOUBLE_PRESS_DURATION).toBe(300)
expect(TIMING.POST_RECORDING_WAIT).toBe(4000)
expect(TIMING.POST_RECORDING_WAIT_BUFFERED).toBe(6000)
expect(TIMING.ABSOLUTE_MAX_WAIT).toBe(120000)
expect(TIMING.AUDIO_LEVEL_INTERVAL).toBe(100)
})
})