estel/dify
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

fix: Before publish the app, preview the voice of tts, it raise an er… (#21821)

Browse Source 
Co-authored-by: 刘江波 <jiangbo721@163.com>
This commit is contained in:
jiangbo721
2025-07-03 10:53:14 +08:00
committed by GitHub
parent a45aa1e505
commit 31eb8548ef
5 changed files with 43 additions and 84 deletions
Download Patch File Download Diff File Expand all files Collapse all files

53
api/services/audio_service.py
View File

@@ -1,13 +1,16 @@
import io
import logging
import uuid
from collections.abc import Generator
from typing import Optional
from flask import Response, stream_with_context
from werkzeug.datastructures import FileStorage
from constants import AUDIO_EXTENSIONS
from core.model_manager import ModelManager
from core.model_runtime.entities.model_entities import ModelType
from extensions.ext_database import db
from models.model import App, AppMode, AppModelConfig, Message, MessageStatus
from services.errors.audio import (
AudioTooLargeServiceError,
@@ -16,6 +19,7 @@ from services.errors.audio import (
ProviderNotSupportTextToSpeechServiceError,
UnsupportedAudioTypeServiceError,
)
from services.workflow_service import WorkflowService
FILE_SIZE = 30
FILE_SIZE_LIMIT = FILE_SIZE * 1024 * 1024
@@ -74,35 +78,36 @@ class AudioService:
voice: Optional[str] = None,
end_user: Optional[str] = None,
message_id: Optional[str] = None,
is_draft: bool = False,
):
from collections.abc import Generator
from flask import Response, stream_with_context
from app import app
from extensions.ext_database import db
def invoke_tts(text_content: str, app_model: App, voice: Optional[str] = None):
def invoke_tts(text_content: str, app_model: App, voice: Optional[str] = None, is_draft: bool = False):
with app.app_context():
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")
if voice is None:
if app_model.mode in {AppMode.ADVANCED_CHAT.value, AppMode.WORKFLOW.value}:
if is_draft:
workflow = WorkflowService().get_draft_workflow(app_model=app_model)
else:
workflow = app_model.workflow
if (
workflow is None
or "text_to_speech" not in workflow.features_dict
or not workflow.features_dict["text_to_speech"].get("enabled")
):
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 = workflow.features_dict["text_to_speech"].get("voice")
else:
if not is_draft:
if app_model.app_model_config is None:
raise ValueError("AppModelConfig not found")
text_to_speech_dict = app_model.app_model_config.text_to_speech_dict
voice = features_dict["text_to_speech"].get("voice") if voice is None else voice
else:
if app_model.app_model_config is None:
raise ValueError("AppModelConfig not found")
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")
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
voice = text_to_speech_dict.get("voice")
model_manager = ModelManager()
model_instance = model_manager.get_default_model_instance(
@@ -136,14 +141,14 @@ class AudioService:
return None
else:
response = invoke_tts(message.answer, app_model=app_model, voice=voice)
response = invoke_tts(text_content=message.answer, app_model=app_model, voice=voice, is_draft=is_draft)
if isinstance(response, Generator):
return Response(stream_with_context(response), content_type="audio/mpeg")
return response
else:
if text is None:
raise ValueError("Text is required")
response = invoke_tts(text, app_model, voice)
response = invoke_tts(text_content=text, app_model=app_model, voice=voice, is_draft=is_draft)
if isinstance(response, Generator):
return Response(stream_with_context(response), content_type="audio/mpeg")
return response