feat(llm): speech-to-text for patient voice input via OpenRouter (+ offline mock)
CI / ruff · pytest (pull_request) Successful in 2m57s
CI / ruff · pytest (pull_request) Successful in 2m57s
POST /api/v1/me/{token}/stt takes a 16 kHz mono WAV and only transcribes:
the patient checks the text and sends it as an ordinary message, so safety,
scrub() and the LLM gateway still apply. Audio is neither stored nor logged.
OpenRouter STT is synthetic-data only (ADR-014); SaluteSpeech for the pilot is #121.
Refs #116
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
+54
-2
@@ -17,17 +17,19 @@ issued by the server instead of handing out the pseudonymous token itself.
|
||||
|
||||
import logging
|
||||
import secrets
|
||||
import time as monotonic_time
|
||||
from collections import defaultdict
|
||||
from datetime import date, datetime, time, timedelta
|
||||
from typing import Any, Literal
|
||||
from zoneinfo import ZoneInfo
|
||||
|
||||
from fastapi import APIRouter, HTTPException
|
||||
from fastapi import APIRouter, HTTPException, Request
|
||||
from pydantic import BaseModel, ConfigDict, Field
|
||||
from sqlalchemy import select
|
||||
|
||||
from app.api import fixtures
|
||||
from app.api.schemas import DemoPersona
|
||||
from app.config import get_settings
|
||||
from app.core import clock
|
||||
from app.core.instrument_catalog import Catalog, catalog
|
||||
from app.core.intake_questions import DIARY_KINDS
|
||||
@@ -42,6 +44,7 @@ from app.db.identity import TOKEN_LEN
|
||||
from app.db.intake import QuestionnaireRecord
|
||||
from app.db.session import get_clinical_session, get_identity_session
|
||||
from app.integrations.emias_mock import CODE_MAX_LEN, EmiasMock
|
||||
from app.llm.stt import InvalidAudio, SttError, SttNoSpeech, get_stt, wav_seconds
|
||||
from app.services import anonymous_care
|
||||
|
||||
log = logging.getLogger(__name__)
|
||||
@@ -219,7 +222,56 @@ async def auth_anonymous(body: AnonymousIn | None = None) -> AuthOut:
|
||||
@router.get("/patient/capabilities")
|
||||
async def patient_capabilities() -> dict[str, bool]:
|
||||
"""Expose availability without leaking clinical policy or server configuration."""
|
||||
return {"primary_screening": intake_enabled()}
|
||||
return {"primary_screening": intake_enabled(), "voice_input": get_stt(get_settings()) is not None}
|
||||
|
||||
|
||||
class SttOut(BaseModel):
|
||||
text: str
|
||||
provider: str
|
||||
|
||||
|
||||
# 16 kHz mono 16-bit PCM = 32 000 bytes/s, + header slack.
|
||||
_WAV_BYTES_PER_SECOND = 32_000
|
||||
|
||||
|
||||
@router.post("/me/{token}/stt", response_model=SttOut)
|
||||
async def transcribe_voice(token: str, request: Request) -> SttOut:
|
||||
"""Voice → text for the patient to check and edit (#116). Body: raw `audio/wav`, 16 kHz mono.
|
||||
|
||||
Only transcribes: nothing is written to the dialog and the agent is not called — the patient sends the
|
||||
checked text as an ordinary message, where safety and `scrub()` apply. The audio is neither stored nor logged.
|
||||
"""
|
||||
settings = get_settings()
|
||||
stt = get_stt(settings)
|
||||
if stt is None:
|
||||
raise HTTPException(404, "voice input is off")
|
||||
async with get_clinical_session() as s:
|
||||
if await repo.get_profile(s, token) is None:
|
||||
raise HTTPException(404, "patient not found")
|
||||
max_bytes = (settings.stt_max_seconds + 2) * _WAV_BYTES_PER_SECOND + 4096
|
||||
declared = request.headers.get("content-length")
|
||||
if declared and declared.isdigit() and int(declared) > max_bytes:
|
||||
raise HTTPException(413, "recording too long")
|
||||
audio = await request.body()
|
||||
if len(audio) > max_bytes:
|
||||
raise HTTPException(413, "recording too long")
|
||||
try:
|
||||
seconds = wav_seconds(audio)
|
||||
except InvalidAudio as exc:
|
||||
raise HTTPException(415, "expected audio/wav") from exc
|
||||
if seconds > settings.stt_max_seconds + 1:
|
||||
raise HTTPException(413, "recording too long")
|
||||
started = monotonic_time.monotonic() # latency only, not domain time (that is app.core.clock)
|
||||
try:
|
||||
text = await stt.transcribe(audio, "wav")
|
||||
except SttNoSpeech as exc:
|
||||
raise HTTPException(422, "no speech recognised") from exc
|
||||
except SttError as exc:
|
||||
log.warning("stt failed tok=%s provider=%s: %s", token[:6], stt.name, exc)
|
||||
raise HTTPException(503, "speech recognition unavailable") from exc
|
||||
ms = int((monotonic_time.monotonic() - started) * 1000)
|
||||
log.info("stt tok=%s provider=%s secs=%.1f ms=%d chars=%d", token[:6], stt.name, seconds, ms, len(text))
|
||||
return SttOut(text=text, provider=stt.name)
|
||||
|
||||
|
||||
@router.post("/me/{token}/doctor/connect")
|
||||
|
||||
Reference in New Issue
Block a user