dify/api/services/audio_service.py

120 lines
4.4 KiB
Python
Raw Normal View History

2023-07-07 17:50:42 +08:00
import io
from typing import Optional
from werkzeug.datastructures import FileStorage
from core.model_manager import ModelManager
from core.model_runtime.entities.model_entities import ModelType
from models.model import App, AppMode, AppModelConfig
from services.errors.audio import (
AudioTooLargeServiceError,
NoAudioUploadedServiceError,
ProviderNotSupportSpeechToTextServiceError,
ProviderNotSupportTextToSpeechServiceError,
UnsupportedAudioTypeServiceError,
)
2023-07-07 17:50:42 +08:00
FILE_SIZE = 30
2023-07-12 17:18:56 +08:00
FILE_SIZE_LIMIT = FILE_SIZE * 1024 * 1024
ALLOWED_EXTENSIONS = ['mp3', 'mp4', 'mpeg', 'mpga', 'm4a', 'wav', 'webm', 'amr']
2023-07-07 17:50:42 +08:00
2023-07-07 17:50:42 +08:00
class AudioService:
@classmethod
def transcript_asr(cls, app_model: App, file: FileStorage, end_user: Optional[str] = None):
if app_model.mode in [AppMode.ADVANCED_CHAT.value, AppMode.WORKFLOW.value]:
workflow = app_model.workflow
if workflow is None:
raise ValueError("Speech to text is not enabled")
features_dict = workflow.features_dict
if 'speech_to_text' not in features_dict or not features_dict['speech_to_text'].get('enabled'):
raise ValueError("Speech to text is not enabled")
else:
app_model_config: AppModelConfig = app_model.app_model_config
if not app_model_config.speech_to_text_dict['enabled']:
raise ValueError("Speech to text is not enabled")
2023-07-07 17:50:42 +08:00
if file is None:
raise NoAudioUploadedServiceError()
2023-07-07 17:50:42 +08:00
extension = file.mimetype
if extension not in [f'audio/{ext}' for ext in ALLOWED_EXTENSIONS]:
raise UnsupportedAudioTypeServiceError()
file_content = file.read()
file_size = len(file_content)
if file_size > FILE_SIZE_LIMIT:
2023-07-12 17:18:56 +08:00
message = f"Audio size larger than {FILE_SIZE} mb"
2023-07-07 17:50:42 +08:00
raise AudioTooLargeServiceError(message)
model_manager = ModelManager()
model_instance = model_manager.get_default_model_instance(
tenant_id=app_model.tenant_id,
model_type=ModelType.SPEECH2TEXT
)
if model_instance is None:
raise ProviderNotSupportSpeechToTextServiceError()
2023-07-07 17:50:42 +08:00
buffer = io.BytesIO(file_content)
2023-07-12 17:18:56 +08:00
buffer.name = 'temp.mp3'
2023-07-07 17:50:42 +08:00
return {"text": model_instance.invoke_speech2text(file=buffer, user=end_user)}
@classmethod
def transcript_tts(cls, app_model: App, text: str, streaming: bool,
voice: Optional[str] = None, end_user: Optional[str] = None):
if app_model.mode in [AppMode.ADVANCED_CHAT.value, AppMode.WORKFLOW.value]:
workflow = app_model.workflow
if workflow is None:
raise ValueError("TTS is not enabled")
features_dict = workflow.features_dict
if 'text_to_speech' not in features_dict or not features_dict['text_to_speech'].get('enabled'):
raise ValueError("TTS is not enabled")
voice = features_dict['text_to_speech'].get('voice') if voice is None else voice
else:
text_to_speech_dict = app_model.app_model_config.text_to_speech_dict
if not text_to_speech_dict.get('enabled'):
raise ValueError("TTS is not enabled")
voice = text_to_speech_dict.get('voice') if voice is None else voice
model_manager = ModelManager()
model_instance = model_manager.get_default_model_instance(
tenant_id=app_model.tenant_id,
model_type=ModelType.TTS
)
if model_instance is None:
raise ProviderNotSupportTextToSpeechServiceError()
try:
return model_instance.invoke_tts(
content_text=text.strip(),
user=end_user,
streaming=streaming,
tenant_id=app_model.tenant_id,
voice=voice
)
except Exception as e:
raise e
@classmethod
def transcript_tts_voices(cls, tenant_id: str, language: str):
model_manager = ModelManager()
model_instance = model_manager.get_default_model_instance(
tenant_id=tenant_id,
model_type=ModelType.TTS
)
if model_instance is None:
raise ProviderNotSupportTextToSpeechServiceError()
try:
return model_instance.get_tts_voices(language)
except Exception as e:
raise e