92e4d992b2
Features: - JWT authentication with user registration and login - Video upload to Amazon S3 with file validation (200MB limit) - Audio transcription using OpenAI Whisper API - Text translation using GPT-4 API - Voice cloning and audio synthesis using ElevenLabs API - Video processing with ffmpeg for audio replacement - Complete SQLite database with proper models and migrations - Background task processing for long-running operations - Health endpoint and comprehensive API documentation Tech stack: - FastAPI with SQLAlchemy ORM - SQLite database with Alembic migrations - Amazon S3 for file storage - OpenAI APIs for transcription and translation - ElevenLabs API for voice cloning - ffmpeg for video processing - JWT authentication with bcrypt password hashing
60 lines
1.8 KiB
Python
60 lines
1.8 KiB
Python
import os
|
|
import openai
|
|
from typing import Optional
|
|
import logging
|
|
import tempfile
|
|
from app.services.s3_service import download_file_from_s3
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
|
|
|
|
if not OPENAI_API_KEY:
|
|
logger.warning("OpenAI API key not configured")
|
|
|
|
openai.api_key = OPENAI_API_KEY
|
|
|
|
|
|
async def transcribe_video_audio(video_s3_url: str) -> Optional[str]:
|
|
if not OPENAI_API_KEY:
|
|
logger.error("OpenAI API key not configured")
|
|
return None
|
|
|
|
try:
|
|
# Extract filename from S3 URL
|
|
file_name = video_s3_url.split('/')[-1]
|
|
|
|
# Download video from S3
|
|
video_content = await download_file_from_s3(file_name)
|
|
if not video_content:
|
|
logger.error("Failed to download video from S3")
|
|
return None
|
|
|
|
# Create temporary file for the video
|
|
with tempfile.NamedTemporaryFile(suffix='.mp4', delete=False) as temp_file:
|
|
temp_file.write(video_content)
|
|
temp_file_path = temp_file.name
|
|
|
|
try:
|
|
# Transcribe using OpenAI Whisper
|
|
with open(temp_file_path, 'rb') as audio_file:
|
|
transcript = openai.Audio.transcribe(
|
|
model="whisper-1",
|
|
file=audio_file,
|
|
response_format="text"
|
|
)
|
|
|
|
# Clean up temporary file
|
|
os.unlink(temp_file_path)
|
|
|
|
return transcript
|
|
|
|
except Exception as e:
|
|
# Clean up temporary file on error
|
|
if os.path.exists(temp_file_path):
|
|
os.unlink(temp_file_path)
|
|
raise e
|
|
|
|
except Exception as e:
|
|
logger.error(f"Error transcribing video: {e}")
|
|
return None |