main.py

from fastapi import FastAPI, UploadFile, File
from fastapi.responses import JSONResponse
import subprocess
import json
import os
import tempfile
import shutil
from pydantic import BaseModel
import torch

app = FastAPI()

class LineDetectionResponse(BaseModel):
    lines: list

@app.post("/detect_lines", response_model=LineDetectionResponse)
async def detect_lines(file: UploadFile = File(...)):
    # Check if CUDA is available
    device = "cuda" if torch.cuda.is_available() else "cpu"
    
    # Create a temporary directory
    with tempfile.TemporaryDirectory() as temp_dir:
        # Save the uploaded file
        temp_file_path = os.path.join(temp_dir, file.filename)
        with open(temp_file_path, "wb") as buffer:
            shutil.copyfileobj(file.file, buffer)
        
        # Set up the output JSON path
        lines_json_path = os.path.join(temp_dir, "lines.json")
        
        # Run Kraken for line detection
        kraken_command = f"kraken -i {temp_file_path} {lines_json_path} segment -bl --device {device}"
        subprocess.run(kraken_command, shell=True, check=True)
        
        # Load the lines from the JSON file
        with open(lines_json_path, 'r') as f:
            lines_data = json.load(f)
    
    # Return the lines data
    return LineDetectionResponse(lines=lines_data['lines'])

# Optionally, you can add a root endpoint for basic information
@app.get("/")
async def root():
    return {"message": "Welcome to the Kraken Line Detection API"}

# To run the app with GPU support on Hugging Face Spaces, you need to use uvicorn with the following settings:
if __name__ == "__main__":
    import uvicorn
    uvicorn.run(app, host="0.0.0.0", port=7860, workers=1)