setup.py

#!/usr/bin/env python3
"""
pdf2othersAPI 安装脚本
"""
import os
import subprocess
import sys
from pathlib import Path

# 加载环境变量
try:
    from dotenv import load_dotenv
    load_dotenv()
except ImportError:
    # 如果没有python-dotenv，尝试手动读取.env文件
    env_file = Path('.env')
    if env_file.exists():
        for line in env_file.read_text().strip().split('\n'):
            if line and not line.startswith('#') and '=' in line:
                key, value = line.split('=', 1)
                os.environ[key.strip()] = value.strip()


def run_command(command, description=""):
    """执行命令并处理错误"""
    print(f"🔄 {description}")
    try:
        result = subprocess.run(command, shell=True, check=True, capture_output=True, text=True)
        print(f"✅ {description} - 完成")
        return result.stdout
    except subprocess.CalledProcessError as e:
        print(f"❌ {description} - 失败")
        print(f"错误: {e.stderr}")
        return None


def check_python_version():
    """检查Python版本"""
    if sys.version_info < (3, 8):
        print("❌ 需要Python 3.8或更高版本")
        sys.exit(1)
    print(f"✅ Python版本: {sys.version}")


def install_dependencies():
    """安装Python依赖"""
    print("\n📦 安装Python依赖...")
    
    # 检查pip
    pip_version = run_command("pip3 --version", "检查pip")
    if not pip_version:
        print("❌ pip3未找到，请先安装pip")
        return False
    
    # 安装依赖
    result = run_command("pip3 install -r requirements.txt", "安装Python包")
    return result is not None


def install_system_dependencies():
    """安装系统依赖"""
    print("\n🔧 检查系统依赖...")
    
    # 检查Pandoc
    pandoc_version = run_command("pandoc --version", "检查Pandoc")
    if not pandoc_version:
        print("📝 Pandoc未安装，尝试安装...")
        
        # 根据操作系统安装Pandoc
        import platform
        system = platform.system().lower()
        
        if system == "darwin":  # macOS
            run_command("brew install pandoc", "安装Pandoc (macOS)")
        elif system == "linux":
            run_command("sudo apt-get update && sudo apt-get install -y pandoc", "安装Pandoc (Linux)")
        else:
            print("❌ 请手动安装Pandoc: https://pandoc.org/installing.html")
            return False
    
    # 检查MinerU
    mineru_check = run_command("python3 -c 'import mineru'", "检查MinerU")
    if not mineru_check:
        print("📝 安装MinerU...")
        run_command("pip3 install mineru", "安装MinerU")
    
    return True


def setup_directories():
    """创建必要的目录"""
    print("\n📁 创建存储目录...")
    
    # 从环境变量读取目录配置，如果没有则使用默认值
    directories = [
        Path(os.getenv('PDFS_PATH', '/var/lib/pdf2others/storage')),
        Path(os.getenv('TMP_TRS_DIR', '/tmp/pdf2others')),
        Path(os.getenv('LOGS_DIR', 'logs'))
    ]
    
    for directory in directories:
        try:
            directory.mkdir(parents=True, exist_ok=True)
            print(f"✅ 创建目录: {directory}")
        except PermissionError:
            print(f"❌ 权限不足，无法创建目录: {directory}")
            print(f"   请手动创建或使用sudo")


def setup_environment():
    """设置环境配置"""
    print("\n⚙️  设置环境配置...")
    
    env_file = Path(".env")
    env_example = Path(".env.example")
    
    if not env_file.exists() and env_example.exists():
        # 复制示例配置文件
        env_content = env_example.read_text()
        env_file.write_text(env_content)
        print(f"✅ 创建配置文件: {env_file}")
        print("⚠️  请编辑 .env 文件，设置正确的配置值")
    else:
        print("ℹ️  .env 文件已存在，跳过创建")


def run_tests():
    """运行测试"""
    print("\n🧪 运行测试...")
    
    # 检查pytest
    pytest_check = run_command("python3 -c 'import pytest'", "检查pytest")
    if not pytest_check:
        run_command("pip3 install pytest", "安装pytest")
    
    # 运行测试
    test_result = run_command("python3 -m pytest tests/ -v", "运行测试")
    return test_result is not None


def main():
    """主安装流程"""
    print("🚀 pdf2othersAPI 安装脚本")
    print("=" * 50)
    
    # 检查Python版本
    check_python_version()
    
    # 安装系统依赖
    if not install_system_dependencies():
        print("❌ 系统依赖安装失败")
        sys.exit(1)
    
    # 安装Python依赖
    if not install_dependencies():
        print("❌ Python依赖安装失败")
        sys.exit(1)
    
    # 创建目录
    setup_directories()
    
    # 设置环境
    setup_environment()
    
    # 运行测试
    if run_tests():
        print("\n🎉 安装完成！")
        print("\n📖 下一步:")
        print("1. 编辑 .env 文件，设置数据库和vLLM配置")
        print("2. 启动vLLM服务")
        print("3. 运行应用: python3 main.py")
    else:
        print("\n⚠️  安装完成，但测试失败")
        print("请检查配置和依赖")


if __name__ == "__main__":
    main()