- 语音:序数解析(第一个/第二个等)、解析失败计数与 API detail.retry_remaining; 百度 ASR 固定 dev_pid 为普通话;SurgeryPipelineError 支持 extra 并入 HTTP detail。 - Demo:demo 路由与假 RTSP、客户端 index 与 README;BackendResolver 与配置调整。 - 可观测:消耗 TSV 日志、语音文件日志、终端 Markdown 辅助;相关测试与依赖更新。 - 注意:.env 仍被 gitignore,本地密钥不会进入本提交。 Made-with: Cursor
72 lines
2.6 KiB
Python
72 lines
2.6 KiB
Python
from __future__ import annotations
|
||
|
||
from threading import Lock
|
||
from typing import Any
|
||
|
||
from aip import AipSpeech
|
||
|
||
from app.config import settings
|
||
|
||
|
||
class BaiduSpeechNotConfiguredError(RuntimeError):
|
||
"""未配置 BAIDU_SPEECH_APP_ID / API_KEY / SECRET_KEY 时调用接口会抛出。"""
|
||
|
||
|
||
class BaiduSpeechService:
|
||
"""百度短语音识别(asr)与在线语音合成(synthesis),基于 `baidu-aip` 的 `AipSpeech`。"""
|
||
|
||
def __init__(self) -> None:
|
||
self._client: AipSpeech | None = None
|
||
self._lock = Lock()
|
||
|
||
@property
|
||
def configured(self) -> bool:
|
||
return settings.baidu_speech_configured
|
||
|
||
def _client_or_raise(self) -> AipSpeech:
|
||
if not self.configured:
|
||
raise BaiduSpeechNotConfiguredError(
|
||
"百度语音未配置:请设置 BAIDU_SPEECH_APP_ID、BAIDU_SPEECH_API_KEY、"
|
||
"BAIDU_SPEECH_SECRET_KEY"
|
||
)
|
||
with self._lock:
|
||
if self._client is None:
|
||
client = AipSpeech(
|
||
settings.baidu_speech_app_id,
|
||
settings.baidu_speech_api_key,
|
||
settings.baidu_speech_secret_key,
|
||
)
|
||
if settings.baidu_speech_connection_timeout_ms is not None:
|
||
client.setConnectionTimeoutInMillis(
|
||
settings.baidu_speech_connection_timeout_ms
|
||
)
|
||
if settings.baidu_speech_socket_timeout_ms is not None:
|
||
client.setSocketTimeoutInMillis(settings.baidu_speech_socket_timeout_ms)
|
||
self._client = client
|
||
return self._client
|
||
|
||
def asr(
|
||
self,
|
||
speech: bytes | None = None,
|
||
format: str = "pcm",
|
||
rate: int = 16000,
|
||
options: dict[str, Any] | None = None,
|
||
) -> dict[str, Any]:
|
||
"""短语音识别。返回百度 JSON(含 `err_no`、`result` 等)。
|
||
|
||
固定使用普通话模型(`dev_pid` 来自配置),避免未传参时误用服务端默认导致偏英语等结果。
|
||
"""
|
||
merged: dict[str, Any] = dict(options or {})
|
||
merged["dev_pid"] = int(settings.baidu_speech_asr_dev_pid)
|
||
return self._client_or_raise().asr(speech, format, rate, merged)
|
||
|
||
def synthesis(
|
||
self,
|
||
text: str,
|
||
lang: str = "zh",
|
||
ctp: int = 1,
|
||
options: dict[str, Any] | None = None,
|
||
) -> bytes | dict[str, Any]:
|
||
"""在线语音合成。成功为音频二进制;失败为错误信息 dict。"""
|
||
return self._client_or_raise().synthesis(text, lang, ctp, options)
|