4e2450e800
- Removed the backend client compatibility wrapper and associated methods to streamline backend integration. - Updated session management to utilize control plane gateways and runtime configuration providers. - Adjusted TTS service implementations to remove the EdgeTTS service and simplify service dependencies. - Enhanced documentation to reflect changes in backend integration and service architecture. - Updated configuration files to remove deprecated TTS provider options and clarify available settings.
65 lines
1.9 KiB
Python
65 lines
1.9 KiB
Python
"""ASR extension port contracts."""
|
|
|
|
from __future__ import annotations
|
|
|
|
from dataclasses import dataclass
|
|
from typing import AsyncIterator, Awaitable, Callable, Optional, Protocol
|
|
|
|
from services.base import ASRResult
|
|
|
|
TranscriptCallback = Callable[[str, bool], Awaitable[None]]
|
|
|
|
|
|
@dataclass(frozen=True)
|
|
class ASRServiceSpec:
|
|
"""Resolved runtime configuration for ASR service creation."""
|
|
|
|
provider: str
|
|
sample_rate: int
|
|
language: str = "auto"
|
|
api_key: Optional[str] = None
|
|
api_url: Optional[str] = None
|
|
model: Optional[str] = None
|
|
interim_interval_ms: int = 500
|
|
min_audio_for_interim_ms: int = 300
|
|
on_transcript: Optional[TranscriptCallback] = None
|
|
|
|
|
|
class ASRPort(Protocol):
|
|
"""Port for speech recognition providers."""
|
|
|
|
async def connect(self) -> None:
|
|
"""Establish connection to ASR provider."""
|
|
|
|
async def disconnect(self) -> None:
|
|
"""Release ASR resources."""
|
|
|
|
async def send_audio(self, audio: bytes) -> None:
|
|
"""Push one PCM audio chunk for recognition."""
|
|
|
|
async def receive_transcripts(self) -> AsyncIterator[ASRResult]:
|
|
"""Stream partial/final recognition results."""
|
|
|
|
|
|
class ASRInterimControl(Protocol):
|
|
"""Optional extension for explicit interim transcription control."""
|
|
|
|
async def start_interim_transcription(self) -> None:
|
|
"""Start interim transcription loop if supported."""
|
|
|
|
async def stop_interim_transcription(self) -> None:
|
|
"""Stop interim transcription loop if supported."""
|
|
|
|
|
|
class ASRBufferControl(Protocol):
|
|
"""Optional extension for explicit ASR buffer lifecycle control."""
|
|
|
|
def clear_buffer(self) -> None:
|
|
"""Clear provider-side ASR buffer."""
|
|
|
|
async def get_final_transcription(self) -> str:
|
|
"""Return final transcription for the current utterance."""
|
|
|
|
def get_and_clear_text(self) -> str:
|
|
"""Return buffered text and clear internal state."""
|