Phase 12~13 전체 구현: Pro+ 피처 6종 + 음성 대화 + RAG + OS 자동화

Phase 12:
- FileTranscriptionService: ffmpeg PCM 변환 + 30초 청크 순차 STT
- MeetingSummaryService: 자막 세션 → LLM 자동 요약 + DB summaryText
- DictationTemplateService: 필드별 음성 입력 상태 머신 + 프리셋 3개

Phase 13.1:
- VoiceConversationService: STT→Ollama /api/chat→TTS 대화 루프 (10턴)
- TTSPlaybackService: Windows SAPI 문장 단위 큐 재생
- LocalLLMService.chatStream: Ollama /api/chat 스트리밍

Phase 13.2:
- RAGService: Ollama 임베딩 + SQLite 벡터 + 코사인 유사도 검색
- KnowledgeBasePage: 문서 관리 + 질문/답변 UI
- PDF 파서: zlib FlateDecode 해제 + BT/ET 텍스트 추출

Phase 13.3:
- VoiceActionService: LLM JSON 액션 플랜 생성 + 실행
- 프리셋 6개 (크롬/메모장/탐색기/볼륨), 위험 명령 차단

공통: IPC ~70채널, 에러코드 780-878, i18n 100+키
버그픽스: 라이선스 로컬 키 우선, i18n featureLabel, DOM 중첩
This commit is contained in:
Yun Chan 2026-04-05 23:52:14 +09:00
parent a31f96bbb8
commit eb83682269
38 changed files with 5678 additions and 19 deletions

View file

@ -432,7 +432,7 @@ export interface HistoryEntry {
focusedApp: string | null
focusedAppName: string | null
focusedAppWindowTitle: string | null
mode: 'dictation' | 'translate' | 'command'
mode: 'dictation' | 'translate' | 'command' | 'caption' | 'file-transcription'
status: 'completed' | 'cancelled' | 'error'
errorCode: string | null
audioLocalPath: string | null
@ -447,6 +447,8 @@ export interface HistoryEntry {
createdAt: number
updatedAt: number
appVersion: string
/** Phase 12.2: 회의록 요약 마크다운 */
summaryText: string | null
}
export interface HistoryQueryParams {
@ -947,3 +949,316 @@ export interface TierComparison {
pro: boolean | string
proPlus: boolean | string
}
// ============================================================
// Phase 12: File Transcription (12.1)
// ============================================================
export type FileTranscriptionState = 'idle' | 'converting' | 'transcribing' | 'completed' | 'error'
export interface FileTranscriptionStartParams {
filePath: string
language?: string
}
export interface FileTranscriptionProgress {
jobId: string
currentChunk: number
totalChunks: number
percent: number
currentText: string
}
export interface FileTranscriptionSegment {
text: string
start: number
end: number
confidence: number
}
export interface FileTranscriptionResult {
jobId: string
filePath: string
fileName: string
fullText: string
segments: FileTranscriptionSegment[]
totalDurationSec: number
processingTimeMs: number
}
export interface FileTranscriptionStateInfo {
state: FileTranscriptionState
jobId: string | null
progress: FileTranscriptionProgress | null
}
// ============================================================
// Phase 12: Meeting Summary (12.2)
// ============================================================
export interface MeetingSummaryResult {
historyId: string
summary: string
decisions: string[]
actionItems: string[]
rawMarkdown: string
generatedAt: number
}
export interface MeetingSummarizeParams {
historyId: string
}
export interface MeetingSummaryGetParams {
historyId: string
}
export interface MeetingSummaryExportParams {
historyId: string
}
export interface MeetingSummaryProgress {
historyId: string
status: 'generating' | 'done' | 'error'
}
// ============================================================
// Phase 12: Dictation Templates (12.3)
// ============================================================
export interface TemplateField {
id: string
name: string
label: string
promptText: string
required: boolean
maxDurationSec: number
}
export interface DictationTemplate {
id: string
name: string
description: string
fields: TemplateField[]
outputFormat: string
isBuiltin: boolean
createdAt: number
updatedAt: number
}
export interface CreateTemplateParams {
name: string
description: string
fields: TemplateField[]
outputFormat: string
}
export interface UpdateTemplateParams {
id: string
name?: string
description?: string
fields?: TemplateField[]
outputFormat?: string
}
export interface DeleteTemplateParams {
id: string
}
export type TemplateSessionState =
| 'idle'
| 'field-prompting'
| 'field-recording'
| 'completing'
export interface TemplateSessionInfo {
templateId: string
templateName: string
state: TemplateSessionState
currentFieldIndex: number
totalFields: number
currentField: TemplateField | null
fieldValues: Record<string, string>
}
export interface StartTemplateSessionParams {
templateId: string
}
export interface SetFieldValueParams {
fieldId: string
value: string
}
export interface TemplateFieldCompletedEvent {
fieldId: string
fieldName: string
value: string
nextField: TemplateField | null
}
export interface TemplateSessionCompletedEvent {
templateId: string
outputText: string
fieldValues: Record<string, string>
}
// ============================================================
// Phase 13: Voice Conversation (13.1)
// ============================================================
export type ConversationState =
| 'idle'
| 'listening'
| 'thinking'
| 'speaking'
export type ConversationRole = 'user' | 'assistant' | 'system'
export interface ConversationMessage {
id: string
role: ConversationRole
content: string
timestamp: number
}
export interface ConversationSessionInfo {
state: ConversationState
messages: ConversationMessage[]
isActive: boolean
}
export interface ConversationSendParams {
text: string
}
export interface ConversationAssistantDelta {
messageId: string
delta: string
accumulated: string
}
export interface ConversationAssistantMessage {
messageId: string
content: string
}
export interface ConversationError {
message: string
phase: 'stt' | 'llm' | 'tts'
}
// ============================================================
// Phase 13: Local RAG (13.2)
// ============================================================
export interface RAGDocument {
id: string
fileName: string
filePath: string
fileType: 'txt' | 'md' | 'pdf' | 'docx'
chunkCount: number
indexed: boolean
indexedAt: number | null
addedAt: number
}
export interface RAGChunk {
id: string
documentId: string
content: string
embedding: number[]
chunkIndex: number
}
export interface RAGQueryParams {
query: string
topK?: number
}
export interface RAGQueryResult {
query: string
results: Array<{
documentId: string
fileName: string
content: string
similarity: number
}>
/** LLM 답변 (컨텍스트 주입 후) */
answer: string
}
export interface RAGAddDocumentParams {
filePath: string
}
export interface RAGRemoveDocumentParams {
documentId: string
}
export interface RAGIndexProgress {
documentId: string
fileName: string
currentChunk: number
totalChunks: number
percent: number
}
export type RAGState = 'idle' | 'indexing' | 'querying'
export interface RAGStateInfo {
state: RAGState
documentCount: number
totalChunks: number
}
// ============================================================
// Phase 13: Voice Action / OS Automation (13.3)
// ============================================================
export type VoiceActionType =
| 'open_app'
| 'open_url'
| 'open_file'
| 'keyboard_shortcut'
| 'type_text'
| 'system_command'
export interface VoiceActionPlan {
action: VoiceActionType
target: string
description: string
safe: boolean
}
export interface VoiceActionExecuteParams {
text: string
}
export interface VoiceActionPreset {
keywords: string[]
action: VoiceActionPlan
}
export interface VoiceActionHistoryEntry {
id: string
userText: string
plan: VoiceActionPlan
executed: boolean
timestamp: number
}
export interface VoiceActionPlannedEvent {
plan: VoiceActionPlan
userText: string
}
export interface VoiceActionExecutedEvent {
plan: VoiceActionPlan
success: boolean
}
export interface VoiceActionErrorEvent {
message: string
userText: string
}