Files
hold-slayer/services/tts.py
Robert Helewka 4048ce1db6 Stage 4: honest health, explicit error policy, event-bus integrity
Engine mode is now explicit: USE_MOCK_SIP=true is the only way to get
the mock engine; an unconfigured trunk fails startup with guidance
instead of silently degrading. Root-caused why the engine always ran
mock: nested pydantic-settings never read .env (no env_file on the
sub-settings classes) — all 8 now declare it.

/health stops lying: reports engine mode (sippy|mock), a live DB
SELECT 1, trunk registration state with reason, and TTS/STT
availability from their last real request; "healthy" now requires
ready + db + sippy + registered trunk.

Error policy: leaf services (tts/transcription/llm_client) raise and
track availability; call-loop callers catch, publish EventType.ERROR
naming the failed service, and apply an explicit fallback. Persistence
writes get one bounded 3x exponential retry, then an ERROR log — no
more silent data loss.

Event bus: a full subscriber queue drops its oldest event (counted)
instead of silently evicting the subscription; subscribe(replay_last=N)
delivers the advertised history replay, used by /ws/events (25).

Receptionist correctness: a matched TAKE_MESSAGE rule beats the LLM;
voicemail polls for early hangup and stops/transcribes/hangs up in
finally; RecordingSession finally keeps its leg_ids so taps detach.

Dead code removed: models/contact.py + Contact table, dtmf_buffer,
transcribe_stream stub, SMS stub in notification.py.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-07-10 07:01:45 -04:00

91 lines
2.9 KiB
Python

"""
TTS Service — Rhema (OpenAI-compatible) text-to-speech client.
Synthesizes speech for the SPEAK call-flow step and the AI Receptionist.
Rhema exposes POST /v1/audio/speech (OpenAI-compatible) with Kokoro voices.
"""
import logging
from pathlib import Path
from typing import Optional
import httpx
from config import TTSSettings
logger = logging.getLogger(__name__)
class TTSService:
"""Client for Rhema TTS service."""
def __init__(self, settings: TTSSettings):
self.settings = settings
self._client: Optional[httpx.AsyncClient] = None
# Last-known reachability, surfaced by /health (None = no requests yet)
self.available: Optional[bool] = None
async def _get_client(self) -> httpx.AsyncClient:
if self._client is None or self._client.is_closed:
headers = {}
if self.settings.api_key.get_secret_value():
headers["Authorization"] = f"Bearer {self.settings.api_key.get_secret_value()}"
self._client = httpx.AsyncClient(
base_url=self.settings.base_url,
timeout=httpx.Timeout(self.settings.timeout, connect=5.0),
headers=headers,
)
return self._client
async def synthesize(
self,
text: str,
voice: Optional[str] = None,
response_format: str = "wav",
) -> bytes:
"""Synthesize speech and return audio bytes."""
if not text or not text.strip():
return b""
client = await self._get_client()
body = {
"model": self.settings.model,
"input": text,
"voice": voice or self.settings.voice,
"response_format": response_format,
"sample_rate": self.settings.sample_rate,
}
# Raises on failure — callers decide the per-call fallback and
# publish a service-error event; swallowing here made a down
# Rhema look like "the AI went quiet".
try:
response = await client.post("/v1/audio/speech", json=body)
response.raise_for_status()
except Exception:
self.available = False
raise
self.available = True
return response.content
async def synthesize_to_file(
self,
text: str,
filepath: str | Path,
voice: Optional[str] = None,
) -> bool:
"""Synthesize to a WAV file. Returns True on success."""
audio = await self.synthesize(text, voice=voice, response_format="wav")
if not audio:
return False
path = Path(filepath)
path.parent.mkdir(parents=True, exist_ok=True)
path.write_bytes(audio)
logger.debug(f"TTS wrote {len(audio)} bytes to {path}")
return True
async def close(self) -> None:
if self._client and not self._client.is_closed:
await self._client.aclose()
self._client = None