UnetAgent路由算法实现遇错：startnode代理加载失败求助

我正在实现一种路由算法，核心逻辑：

• 节点A向不存在的节点发送RouteDiscoveryReq，触发邻居发现
• 邻居节点a1、a2收到RouteDiscoveryNtf后，等待随机时长（避免冲突）再向节点A回复随机数

给a1、a2部署startnode.groovy代理时，加载报错：

BUG! exception in phase 'semantic analysis' ...The lookup for startnode caused a failed compilation

相关代码片段 #

1. multiHop.groovy（仿真脚本）

部署6个节点，其中a1、a2为节点A的邻居并运行startnode代理，以下是关键代码：

def n2 = node 'A', address: 2, location: [ 0.km, 2.km, 0.m], web: 8087, stack: "$home/etc/setup"
n2.startup={
    def rdp = agentForService Services.ROUTE_MAINTENANCE
    subscribe topic(rdp);
    def datagram = agentForService(org.arl.unet.Services.DATAGRAM)
    subscribe topic(datagram);
    n = []                        
    rdp << new RouteDiscoveryReq(to: 100, count: 1)
    while (ntf = receive(RouteDiscoveryNtf, 10000)) {
      n << ntf.nextHop            // 添加邻居到列表
    }
    n = n.unique() // 去重
    println("Neighbors are ${n}")
}

def a1 = node 'a1', address: 3, location: [ 0.km, 3.km, -1850.m], web: 8088, stack: "$home/etc/setup"
a1.startup={
    container.add 'sn', new startnode();
}

def a2 = node 'a2', address: 4, location: [ 1.5.km, 3.km, -1850.m], web: 8089, stack: "$home/etc/setup"
a2.startup={
    container.add 'sn', new startnode();
}

2. startnode.groovy（代理脚本）

实现等待随机时长后回复随机数的逻辑：

import org.arl.fjage.*
import org.arl.fjage.param.Parameter
// import static org.arl.unet.utils.MathUtils.*
import org.arl.unet.*

class startnode extends UnetAgent {
    
  int delay = Math.abs(new Random().nextInt()) % 1000 + 1                           
  int rand_num = Math.abs(new Random().nextInt()) % 10 + 1
  
  void startup() {
    subscribeForServices(Services.DATAGRAM)
    subscribeForServices(Services.ROUTE_MAINTENANCE)
  }
  
  void processMessage(Message msg) {
     
    if (msg instanceof RouteDiscoveryNtf) {
      add new WakerBehavior(delay, {
        send new DatagramReq(
          recipient: msg.sender,
          to: msg.from,
          data: rand_num
        )
      })
    }
  }
}

错误原因及修复方案 #

1. 类名不符合Groovy规范：Groovy要求类名采用大驼峰命名，将startnode改为StartNode，同时在multiHop.groovy中实例化时改为new StartNode()
2. 缺少Random类导入：startnode.groovy中使用了Random但未导入，需添加import java.util.Random
3. 变量作用域问题：节点A的启动脚本中n未声明，改为def n = []避免作用域混淆
4. 订阅服务简化：可以直接用subscribe Services.ROUTE_MAINTENANCE替代通过代理订阅topic的写法

修复后的StartNode.groovy：

import org.arl.fjage.*
import org.arl.fjage.param.Parameter
import org.arl.unet.*
import java.util.Random

class StartNode extends UnetAgent {
    
  int delay = Math.abs(new Random().nextInt()) % 1000 + 1                           
  int rand_num = Math.abs(new Random().nextInt()) % 10 + 1
  
  void startup() {
    subscribe Services.DATAGRAM
    subscribe Services.ROUTE_MAINTENANCE
  }
  
  void processMessage(Message msg) {
     
    if (msg instanceof RouteDiscoveryNtf) {
      add new WakerBehavior(delay, {
        send new DatagramReq(
          recipient: msg.sender,
          to: msg.from,
          data: rand_num
        )
      })
    }
  }
}

修复后的multiHop.groovy中a1、a2启动代码：

def a1 = node 'a1', address: 3, location: [ 0.km, 3.km, -1850.m], web: 8088, stack: "$home/etc/setup"
a1.startup={
    container.add 'sn', new StartNode();
}

def a2 = node 'a2', address: 4, location: [ 1.5.km, 3.km, -1850.m], web: 8089, stack: "$home/etc/setup"
a2.startup={
    container.add 'sn', new StartNode();
}

内容的提问来源于stack exchange，提问作者Radhika Suradkar