UML MCP

#!/bin/bash # UML MCP 服务启动脚本 # 用途：简化本地开发和部署流程 set -e # 颜色定义 RED='\033[0;31m' GREEN='\033[0;32m' YELLOW='\033[1;33m' BLUE='\033[0;34m' NC='\033[0m' # No Color # 日志函数 log_info() { echo -e "${BLUE}[INFO]${NC} $1" } log_success() { echo -e "${GREEN}[SUCCESS]${NC} $1" } log_warning() { echo -e "${YELLOW}[WARNING]${NC} $1" } log_error() { echo -e "${RED}[ERROR]${NC} $1" } # 检查命令是否存在 command_exists() { command -v "$1" >/dev/null 2>&1 } # 检查Java环境 check_java() { if ! command_exists java; then log_error "Java未安装，请先安装Java 8+" exit 1 fi java_version=$(java -version 2>&1 | head -n1 | cut -d'"' -f2) log_info "检测到Java版本: $java_version" } # 检查Python环境 check_python() { if ! command_exists python3; then log_error "Python3未安装，请先安装Python 3.9+" exit 1 fi python_version=$(python3 --version) log_info "检测到Python版本: $python_version" } # 检查uv包管理器 check_uv() { if ! command_exists uv; then log_error "uv包管理器未安装，请先安装uv" log_info "安装命令: curl -LsSf https://astral.sh/uv/install.sh | sh" exit 1 fi uv_version=$(uv --version) log_info "检测到uv版本: $uv_version" } # 下载PlantUML JAR download_plantuml() { local jar_path="./plantuml.jar" if [ ! -f "$jar_path" ]; then log_info "下载PlantUML JAR文件..." # 优先使用 curl，如果不存在则尝试 wget if command_exists curl; then curl -L -o "$jar_path" "https://github.com/plantuml/plantuml/releases/download/v1.2024.0/plantuml-1.2024.0.jar" elif command_exists wget; then wget -O "$jar_path" "https://github.com/plantuml/plantuml/releases/download/v1.2024.0/plantuml-1.2024.0.jar" else log_error "未找到 curl 或 wget 命令，请手动下载 PlantUML JAR 文件" log_info "下载地址: https://github.com/plantuml/plantuml/releases/download/v1.2024.0/plantuml-1.2024.0.jar" log_info "保存为: $jar_path" exit 1 fi log_success "PlantUML JAR下载完成" else log_info "PlantUML JAR文件已存在" fi # 测试PlantUML if java -jar "$jar_path" -version >/dev/null 2>&1; then log_success "PlantUML测试通过" else log_error "PlantUML测试失败" exit 1 fi } # 安装Python依赖 install_dependencies() { log_info "安装Python依赖..." # 优先使用uv，如果存在pyproject.toml if [ -f "pyproject.toml" ] && command_exists uv; then log_info "使用uv安装依赖..." uv sync log_success "uv依赖安装完成" elif [ -f "requirements.txt" ]; then log_info "使用pip安装依赖..." python3 -m pip install -r requirements.txt log_success "pip依赖安装完成" else log_error "未找到pyproject.toml或requirements.txt文件" exit 1 fi } # 创建必要目录 create_directories() { log_info "创建必要目录..." mkdir -p temp cache logs log_success "目录创建完成" } # 设置环境变量 setup_environment() { log_info "设置环境变量..." export PLANTUML_JAR_PATH="$(pwd)/plantuml.jar" export UML_TEMP_DIR="$(pwd)/temp" export UML_CACHE_DIR="$(pwd)/cache" export UML_LOG_LEVEL="INFO" export UML_MAX_FILE_SIZE="10485760" export UML_TIMEOUT="30" export UML_MAX_CONCURRENT="10" export PYTHONPATH="$(pwd)" log_success "环境变量设置完成" } # 启动服务 start_server() { log_info "启动UML MCP服务..." # 优先使用uv运行，如果存在pyproject.toml if [ -f "pyproject.toml" ] && command_exists uv; then log_info "使用uv运行服务..." uv run python server.py else log_info "使用python3运行服务..." python3 server.py fi } # Docker模式启动 start_docker() { log_info "使用Docker启动服务..." if ! command_exists docker; then log_error "Docker未安装，请先安装Docker" exit 1 fi if ! command_exists docker-compose; then log_error "Docker Compose未安装，请先安装Docker Compose" exit 1 fi log_info "构建并启动Docker容器..." docker-compose up --build } # 显示帮助信息 show_help() { echo "UML MCP 服务启动脚本" echo "" echo "用法: $0 [选项]" echo "" echo "选项:" echo " --local 本地模式启动（默认）" echo " --docker Docker模式启动" echo " --check 仅检查环境依赖" echo " --help 显示此帮助信息" echo "" echo "示例:" echo " $0 # 本地模式启动" echo " $0 --docker # Docker模式启动" echo " $0 --check # 检查环境" } # 主函数 main() { local mode="local" # 解析命令行参数 while [[ $# -gt 0 ]]; do case $1 in --local) mode="local" shift ;; --docker) mode="docker" shift ;; --check) mode="check" shift ;; --help) show_help exit 0 ;; *) log_error "未知参数: $1" show_help exit 1 ;; esac done log_info "UML MCP 服务启动脚本" log_info "模式: $mode" echo "" case $mode in "local") check_java check_python # 如果存在pyproject.toml，则检查uv if [ -f "pyproject.toml" ]; then check_uv fi download_plantuml install_dependencies create_directories setup_environment start_server ;; "docker") start_docker ;; "check") check_java check_python log_success "环境检查完成" ;; esac } # 执行主函数 main "$@"