feat: [#46] хендлер с рабочим stt

src/assistant/lib/api/v1/handlers/__init__.py

@@ -1,3 +1,7 @@
 from .health import basic_router
+from .voice_responce_handler import VoiceResponseHandler
 
-__all__ = ["basic_router"]
+__all__ = [
+    "VoiceResponseHandler",
+    "basic_router",
+]

src/assistant/lib/api/v1/handlers/voice_responce_handler.py

@@ -0,0 +1,28 @@
+import fastapi
+
+import lib.models.tts.voice as models_tts_voice
+import lib.stt.services as stt_services
+
+
+class VoiceResponseHandler:
+    def __init__(
+        self,
+        stt: stt_services.SpeechService,
+    ):
+        self.stt = stt
+        self.router = fastapi.APIRouter()
+        self.router.add_api_route(
+            "/",
+            self.voice_response,
+            methods=["POST"],
+            summary="Ответ голосового помощника",
+            description="Ответ голосового помощника",
+        )
+
+    async def voice_response(
+        self,
+        voice: bytes = fastapi.File(...),
+        voice_model: models_tts_voice.VoiceModelProvidersEnum = fastapi.Depends(),
+    ) -> dict[str, str]:
+        voice_text: str = await self.stt.recognize(voice)
+        return {"text": voice_text}

src/assistant/lib/api/v1/schemas/__init__.py

@@ -1,3 +1,6 @@
-from .base import HealthResponse
+from .base import HealthResponse, VoiceResponse
 
-__all__ = ["HealthResponse"]
+__all__ = [
+    "HealthResponse",
+    "VoiceResponse",
+]

src/assistant/lib/api/v1/schemas/base.py

@@ -3,3 +3,7 @@ import pydantic
 
 class HealthResponse(pydantic.BaseModel):
     status: str = pydantic.Field(default=..., examples=["healthy"], description="Схема доступности сервиса")
+
+
+class VoiceResponse(pydantic.BaseModel):
+    voice: bytes = pydantic.Field(default=..., description="Голосовой ответ")

src/assistant/lib/app/app.py

@@ -81,12 +81,12 @@ class Application:
         # Services
 
         logger.info("Initializing services")
-        stt_service: stt.SpeechService = stt.SpeechService(repository=stt_repository)  # type: ignore
-
+        stt_service: stt.SpeechService = stt.SpeechService(repository=stt_repository)
         # Handlers
 
         logger.info("Initializing handlers")
         liveness_probe_handler = api_v1_handlers.basic_router
+        voice_response_handler = api_v1_handlers.VoiceResponseHandler(stt=stt_service).router
 
         logger.info("Creating application")
 
@@ -100,6 +100,7 @@ class Application:
 
         # Routes
         fastapi_app.include_router(liveness_probe_handler, prefix="/api/v1/health", tags=["health"])
+        fastapi_app.include_router(voice_response_handler, prefix="/api/v1/voice", tags=["voice"])
 
         application = Application(
             settings=settings,

src/assistant/lib/stt/repositories.py

@@ -3,6 +3,7 @@ import tempfile
 
 import magic
 import openai
+import pydantic
 
 import lib.app.settings as app_settings
 import lib.stt as stt
@@ -26,13 +27,15 @@ class OpenaiSpeechRepository:
         file_extension = self.__get_file_extension_from_bytes(audio)
         if not file_extension:
             raise ValueError("File extension is not supported")
-
-        voice: stt.models.SttVoice = stt.models.SttVoice(
-            audio_size=len(audio) // 1024,  # audio size in MB,
-            audio_format=file_extension,
-            audio_data=audio,
-            voice_settings=self.settings.voice,
-        )
+        try:
+            voice: stt.models.SttVoice = stt.models.SttVoice(
+                audio_size=len(audio) // 1024,  # audio size in MB,
+                audio_format=file_extension,
+                audio_data=audio,
+                voice_settings=self.settings.voice,
+            )
+        except (pydantic.ValidationError, ValueError) as e:
+            raise ValueError(f"Voice validation error: {e}")
 
         try:
             with tempfile.NamedTemporaryFile(suffix=f".{file_extension}") as temp_file:

src/assistant/lib/tts/__init__.py

@@ -0,0 +1,5 @@
+from .services import TTSService
+
+__all__ = [
+    "TTSService",
+]