Phase 15.5 구현: 화자 구분 (Speaker Diarization)

- Sidecar: pyannote /diarize 엔드포인트 + requirements.txt 업데이트
- LLM 기반 화자 추정 (Phase 1 — 오디오 보존 없이 전사 텍스트 분석)
- CaptionSegment에 speaker 필드 추가
- EditableSegment: 화자별 색상 바 + 화자 Chip 표시
- TranscriptTab: 화자 구분 버튼 + 진행률
- 설정: HuggingFace 토큰 입력 + Diarization 토글
- IPC: DIARIZE + DIARIZATION_PROGRESS 채널
- 에러코드: 895-897
- 12개 locale i18n
This commit is contained in:
Yun Chan 2026-04-08 12:14:00 +09:00
parent da25791c75
commit fc7628327a
25 changed files with 679 additions and 18 deletions

View file

@ -2,10 +2,26 @@
// Phase 14.5: 전사 세그먼트 인라인 편집 컴포넌트
import { useState, useRef, useCallback } from 'react'
import { Box, TextField, Tooltip } from '@mui/material'
import { Box, TextField, Tooltip, Chip } from '@mui/material'
import { d3roPalette, d3roFontMono } from '../../theme'
import { useI18n } from '../../i18n'
// Phase 15.5: 화자별 색상 매핑 (d3roPalette SSOT)
const SPEAKER_COLORS = [
d3roPalette.tag.purple,
d3roPalette.tag.green,
d3roPalette.tag.orange,
d3roPalette.tag.red,
'#3b82f6',
]
function getSpeakerColor(speaker: string): string {
// "화자 1" → 0, "화자 2" → 1, ...
const match = /(\d+)$/.exec(speaker)
const idx = match ? (parseInt(match[1], 10) - 1) : 0
return SPEAKER_COLORS[idx % SPEAKER_COLORS.length]
}
interface EditableSegmentProps {
segmentId: string
timestamp: number // ms, 상대 시간
@ -13,6 +29,7 @@ interface EditableSegmentProps {
edited: boolean // 수정 여부
onEdit: (segmentId: string, newText: string) => void
readOnly?: boolean
speaker?: string
}
function formatTimestamp(ms: number): string {
@ -29,6 +46,7 @@ export function EditableSegment({
edited,
onEdit,
readOnly = false,
speaker,
}: EditableSegmentProps): React.ReactElement {
const { t } = useI18n()
const [editing, setEditing] = useState(false)
@ -61,9 +79,20 @@ export function EditableSegment({
[handleCommit],
)
const speakerColor = speaker ? getSpeakerColor(speaker) : undefined
if (editing) {
return (
<Box sx={{ display: 'flex', alignItems: 'flex-start', gap: 1, mb: 0.5 }}>
<Box
sx={{
display: 'flex',
alignItems: 'flex-start',
gap: 1,
mb: 0.5,
borderLeft: speakerColor ? `4px solid ${speakerColor}` : undefined,
pl: speakerColor ? 0.75 : 0,
}}
>
<Box
component="span"
sx={{
@ -125,7 +154,16 @@ export function EditableSegment({
)
return (
<Box sx={{ display: 'flex', alignItems: 'flex-start', gap: 1, mb: 0.5 }}>
<Box
sx={{
display: 'flex',
alignItems: 'flex-start',
gap: 1,
mb: 0.5,
borderLeft: speakerColor ? `4px solid ${speakerColor}` : undefined,
pl: speakerColor ? 0.75 : 0,
}}
>
<Box
component="span"
sx={{
@ -139,6 +177,21 @@ export function EditableSegment({
>
{formatTimestamp(timestamp)}
</Box>
{speaker && (
<Chip
label={speaker}
size="small"
sx={{
fontSize: 10,
height: 18,
flexShrink: 0,
bgcolor: `${speakerColor}22`,
color: speakerColor,
border: `1px solid ${speakerColor}55`,
fontFamily: d3roFontMono,
}}
/>
)}
{edited ? (
<Tooltip title={t('meeting.modified')} placement="top">
{textNode}