import os import uuid from django.http import FileResponse from rest_framework import status from rest_framework.response import Response from rest_framework.generics import CreateAPIView from TTS.api import TTS from rest_framework.authentication import TokenAuthentication from rest_framework.permissions import IsAuthenticated from .serializers import TextToSpeechSerializer from rest_framework.parsers import MultiPartParser from drf_yasg import openapi from drf_yasg.utils import swagger_auto_schema class TextToSpeechCreateView(CreateAPIView): serializer_class = TextToSpeechSerializer authentication_classes = [TokenAuthentication] # Apply token authentication permission_classes = [IsAuthenticated] # Require authentication for this view parser_classes = [MultiPartParser] @swagger_auto_schema( operation_id='Create a document', operation_description='Create a document by providing file and s3_key', manual_parameters=[ openapi.Parameter('file', openapi.IN_FORM, type=openapi.TYPE_FILE, description='Document to be uploaded'), openapi.Parameter('s3_key', openapi.IN_FORM, type=openapi.TYPE_STRING, description='S3 Key of the Document ' '(folders along with name)') ], responses={ status.HTTP_200_OK: openapi.Response( 'Success', schema=openapi.Schema(type=openapi.TYPE_OBJECT, properties={ 'doc_id': openapi.Schema(type=openapi.TYPE_STRING, description='Document ID'), 'mime_type': openapi.Schema(type=openapi.TYPE_STRING, description='Mime Type of the Document'), 'version_id': openapi.Schema(type=openapi.TYPE_STRING, description='S3 version ID of the document') }) ) } ) def create(self, request, *args, **kwargs): serializer = self.get_serializer(data=request.data) if serializer.is_valid(): text = serializer.validated_data.get("text") speaker_wav = serializer.validated_data.get("speaker_wav") language = serializer.validated_data.get("language") output_filename = f"output_{uuid.uuid4()}.wav" try: # Save the uploaded speaker file to a temporary location speaker_file_path = os.path.join("/tmp", speaker_wav.name) with open(speaker_file_path, "wb") as destination: for chunk in speaker_wav.chunks(): destination.write(chunk) # Generate speech using tts.tts_to_file tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2", gpu=False) tts.tts_to_file(text=text, file_path=output_filename, speaker_wav=speaker_file_path, language=language) # Define a function to delete the output file def file_iterator(file_name): with open(file_name, 'rb') as f: yield from f # Delete the file after sending it try: os.remove(file_name) except Exception as e: # You might want to log this error pass # Use the file_iterator to create a FileResponse response = FileResponse(file_iterator(output_filename), as_attachment=True, content_type='audio/wav') return response except Exception as e: return Response({"error": str(e)}, status=status.HTTP_500_INTERNAL_SERVER_ERROR) return Response(serializer.errors, status=status.HTTP_400_BAD_REQUEST)