app.py

from fastapi import FastAPI, File, UploadFile
from magic_pdf.user_api import MagicPDF
import tempfile
import os
import json
import shutil

app = FastAPI(title="MinerU PDF Processing API",
             description="API for processing PDF documents using MinerU",
             version="1.0.0")

# 初始化 MagicPDF
config_file = os.path.join(os.path.expanduser('~'), 'magic-pdf.json')
magic_pdf = MagicPDF(config_file)

@app.post("/process_pdf")
async def process_pdf(file: UploadFile = File(...)):
    """
    处理上传的PDF文件并返回分析结果
    """
    if not file.filename.lower().endswith('.pdf'):
        return {"error": "Only PDF files are supported"}
    
    try:
        # 创建临时目录
        with tempfile.TemporaryDirectory() as temp_dir:
            # 保存上传的文件
            temp_pdf = os.path.join(temp_dir, file.filename)
            with open(temp_pdf, "wb") as buffer:
                shutil.copyfileobj(file.file, buffer)
            
            # 处理 PDF
            result = magic_pdf.process(temp_pdf)
            
            return result
    except Exception as e:
        return {"error": str(e)}

@app.get("/health")
async def health_check():
    """
    健康检查接口
    """
    return {"status": "healthy", "version": "1.0.0"}