eeac2eefb3
All checks were successful
Deploy / deploy (push) Successful in 5m44s
Шаг 1 миграции голосового стека из home-voice-assistant в сам tablet: - /api/voice/chat — Claude Haiku 4.5 с tool-loop (max 4 раунда), prompt caching на system + старой истории, история в /data/voice-history/. Эмитит command/response/error в voice-bus → орб моргает как раньше. - /api/voice/stt — Groq whisper-large-v3-turbo, multipart или raw audio. - lib/voice-text.ts — порт clean_for_speech (без pymorphy3, время в именительном падеже) и strip_fillers + RESET_PATTERNS. - lib/voice-executors.ts — tool executors через loopback fetch на существующие /api/voice/tools/* и /api/voice/timer. - Поддержка ANTHROPIC_PROXY/GROQ_PROXY (fallback на HTTPS_PROXY). После деплоя нужны GROQ_API_KEY и ANTHROPIC_API_KEY в tablet.env. Шаги 2 (push-to-talk в браузере) и 3 (wake-word) — отдельно.
72 lines
2.3 KiB
TypeScript
72 lines
2.3 KiB
TypeScript
export const dynamic = 'force-dynamic'
|
||
export const runtime = 'nodejs'
|
||
|
||
import { NextResponse } from 'next/server'
|
||
import Groq from 'groq-sdk'
|
||
import { HttpsProxyAgent } from 'https-proxy-agent'
|
||
import { toFile } from 'groq-sdk/uploads'
|
||
|
||
const STT_MODEL = process.env.GROQ_STT_MODEL || 'whisper-large-v3-turbo'
|
||
|
||
let _client: Groq | null = null
|
||
function client(): Groq {
|
||
if (_client) return _client
|
||
const apiKey = process.env.GROQ_API_KEY
|
||
if (!apiKey) throw new Error('GROQ_API_KEY not set')
|
||
const proxy = process.env.GROQ_PROXY || process.env.HTTPS_PROXY || ''
|
||
const httpAgent = proxy ? new HttpsProxyAgent(proxy) : undefined
|
||
_client = new Groq({ apiKey, httpAgent })
|
||
return _client
|
||
}
|
||
|
||
// Принимает либо multipart/form-data с полем "file",
|
||
// либо raw audio в теле (Content-Type: audio/* — например audio/webm).
|
||
// Возвращает {text: string}.
|
||
export async function POST(req: Request) {
|
||
let audio: { name: string; data: Buffer; mime: string }
|
||
|
||
const ct = req.headers.get('content-type') || ''
|
||
|
||
try {
|
||
if (ct.startsWith('multipart/form-data')) {
|
||
const fd = await req.formData()
|
||
const file = fd.get('file')
|
||
if (!(file instanceof Blob)) {
|
||
return NextResponse.json({ error: 'file field required' }, { status: 400 })
|
||
}
|
||
const ab = await file.arrayBuffer()
|
||
audio = {
|
||
name: (file as any).name || 'audio.webm',
|
||
data: Buffer.from(ab),
|
||
mime: file.type || 'audio/webm',
|
||
}
|
||
} else {
|
||
const ab = await req.arrayBuffer()
|
||
if (!ab.byteLength) {
|
||
return NextResponse.json({ error: 'empty body' }, { status: 400 })
|
||
}
|
||
audio = {
|
||
name: 'audio.webm',
|
||
data: Buffer.from(ab),
|
||
mime: ct || 'audio/webm',
|
||
}
|
||
}
|
||
} catch (e) {
|
||
return NextResponse.json({ error: 'failed_to_read_body' }, { status: 400 })
|
||
}
|
||
|
||
try {
|
||
const file = await toFile(audio.data, audio.name, { type: audio.mime })
|
||
const result = await client().audio.transcriptions.create({
|
||
file,
|
||
model: STT_MODEL,
|
||
language: 'ru',
|
||
})
|
||
const text = (result as any).text || ''
|
||
return NextResponse.json({ text })
|
||
} catch (e: any) {
|
||
console.error('[voice/stt] groq error:', e?.message || e)
|
||
return NextResponse.json({ error: 'stt_failed', detail: String(e?.message || e) }, { status: 502 })
|
||
}
|
||
}
|