网络编程：使用UDP实现数据帧的接收

1、需求

2、逻辑处理

3、代码实现

4、总结

1、需求

使用java代码实现数据帧的接收需求，完成数据到数据库的存储。

2、逻辑处理

由于udp传输不保证数据的有序性、可靠性，所以在做业务开发的时候，要程序员自己考虑需求完善udp的缺陷。因此我们定义几个字段保证数据的有序，对于可靠性未进行考虑。

3、代码实现

监听端口55000，等待数据的发送：

@Service
public class UDPReceiverSuper {private static final int BUFFER_SIZE = 1044;private static final int HEAD_SIZE = 20;private static final int DATA_SIZE = 1024;private static final int MAX_BUFFER_SIZE = 1 * 1024 * 1024; // 缓冲器大小设置为1MBprivate static final double MAX_BUFFER_THRESHOLD = 0.8; // 缓冲区阈值private static final int MAX_BUFFER_INDEX = (int) (MAX_BUFFER_SIZE * MAX_BUFFER_THRESHOLD / DATA_SIZE); //缓冲区元素数量阈值//timestampToBufferMap存储的是：时间戳，TreeMap，TreeMap里面存储的是：当前包序号，接受数据的对象private Map<Long, ConcurrentHashMap<Long, DatagramPacket>> timestampToBufferMap = new HashMap();private long timeStamp;private boolean isClosed = false;// 使用阻塞队列作为缓冲区private long errorPackageSum = 0;private int frameNum;        //用于帧计数Thread udpReceiverThread;@Value("${GK.GKOriginalDataFilePath}")private String GKOriginalDataFilePath; // 管控原始数据文件存储路径@Value("${HP.storagePath}")private String storagePath;    //高性能数据接收路径@Autowiredprivate INetworkConfigService networkConfigService;@Autowiredprivate DealGkDataServiceSuperWithNewThread dealGkDataServiceSuperWithNewThread;@Autowiredprivate DealGkDataServiceSuperWithThreadPoolAndBuffer dealGkDataServiceSuperWithThreadPoolAndBuffer;@Autowiredprivate DealGkDataServiceSuperWithThreadPool dealGkDataServiceSuperWithThreadPool;@Autowiredprivate SaveGKOriginalDataService saveGKOriginalDataService;@Autowiredprivate SaveGKOriginalDataServiceWithBuffer saveGKOriginalDataServiceWithBuffer;public UDPReceiverSuper() {}public void start() {//创建父文件夹Path path = Paths.get(storagePath);if (Files.notExists(path)) {try {Files.createDirectories(path);System.out.println("Directories created successfully: " + storagePath);} catch (IOException e) {System.err.println("Failed to create directories: " + e.getMessage());}} else {System.out.println("Directories already exist: " + storagePath);}// 启动接收数据的线程if (udpReceiverThread == null) {udpReceiverThread = new Thread(new Receiver());udpReceiverThread.start();}}//数据帧头定义private class PackageHeader {public long id = 0;public long timestamp = 0;public long totalPackageNum = 0;public long currentPackageNum = 0;public long dataLength = 0;}// 接收数据的线程private class Receiver implements Runnable {@Overridepublic void run() {NetworkConfig networkConfig = networkConfigService.selectNetworkConfigById(1L);String port = networkConfig.getPort();String ip = networkConfig.getIp();System.out.println("实际未绑定ip");System.out.println("ip: " + ip + "  port: " + port);try {DatagramSocket ds = new DatagramSocket(Integer.parseInt(port));if (ds != null) {isClosed = false;}System.out.println("udpReceiver_ds: " + ds + "   等待接收数据......");while (true) {if (isClosed) {break;}byte[] receiveData = new byte[BUFFER_SIZE];   //接收数据缓存区，大小为1044DatagramPacket receivePacket = new DatagramPacket(receiveData, receiveData.length);ds.receive(receivePacket);     //接收数据byte[] data1 = receivePacket.getData();frameNum++;
//                    System.out.println("当前帧数为： " + frameNum);   //todo 用于打印输出当前接收到的帧数ByteBuffer byteBuffer1 = ByteBuffer.allocate(data1.length);byteBuffer1.put(data1);byteBuffer1.flip();   //flip操作是将：写模式切换到读模式，将‘limit’设置为当前的‘position’，将‘position’重置为0
//                    ByteBuffer byteBuffer1 = ByteBuffer.allocate(receiveData.length);
//                    byteBuffer1.put(receiveData);
//                    byteBuffer1.flip();   //flip操作是将：写模式切换到读模式，将‘limit’设置为当前的‘position’，将‘position’重置为0/*两种情况：1、接收管控  2、接收高性能*/byteBuffer1.order(ByteOrder.LITTLE_ENDIAN);  //转化为小端int headerType = byteBuffer1.getInt();       //得到设备标识符if (headerType == 1) {/*解决方法一： 这个是采用多线程的方式进行写入数据到txt文件*/saveGKOriginalDataService.saveGKOriginalData(receivePacket, GKOriginalDataFilePath);/*解决方法二：直接处理管控的函数*/
//                        dealGkDataServiceSuperWithNewThread.dealGKApi(byteBuffer1);/*解决方法三：在UDPReceiverSuperSuper类里面，并且要在NetworkConfigController中进行函数*//*解决方法四：使用线程池的方式解决，每接收一帧，就开始处理*/
//                        dealGkDataServiceSuperWithThreadPoolAndBuffer.dealGKApi(byteBuffer1);/*解决方法五：直接开启线程进行处理数据,这个方法是对的*/dealGkDataServiceSuperWithThreadPool.dealGKApi(byteBuffer1);/*解决方法六：将接收到的数据存储到缓冲区中，然后使用多线程从缓冲区中取出，方法实现写在method3中*/}

业务处理逻辑：

package com.ruoyi.system.service.customService.dealGKService_ThreadPool;import com.ruoyi.system.domain.*;
import com.ruoyi.system.mapper.*;
import com.ruoyi.system.utlis.CSVFileUtil;
import com.ruoyi.system.utlis.ConvertUtil;
import org.springframework.beans.factory.annotation.Autowired;
import org.springframework.beans.factory.annotation.Value;
import org.springframework.stereotype.Service;import java.io.IOException;
import java.nio.ByteBuffer;
import java.nio.ByteOrder;
import java.nio.charset.StandardCharsets;
import java.nio.file.Files;
import java.nio.file.Path;
import java.nio.file.Paths;
import java.sql.SQLIntegrityConstraintViolationException;
import java.util.HashMap;
import java.util.Map;
import java.util.concurrent.*;
import java.util.concurrent.atomic.LongAdder;/*** @Author 不要有情绪的  ljy* @Date 2024/3/22 15:54* @Description:*/
@Service
public class DealGkDataServiceSuperWithThreadPool {@Value("${GK.statusLogPath}")private String stateLogFilePath;  //状态日志存储路径@Autowiredprivate OperateLogInfoMapper operateLogInfoMapper;@Autowiredprivate InstructLogInfoMapper instructLogInfoMapper;@Autowiredprivate Instruct1553bLogInfoMapper instruct1553bLogInfoMapper;@Autowiredprivate InstructRs422LogInfoMapper instructRs422LogInfoMapper;@Autowiredprivate StateLogInfoMapper stateLogInfoMapper;@Autowiredprivate ErrorLogInfoMapper errorLogInfoMapper;int frontTimeFlag = -1;private int currentReceivedFrameNum = 0;  //用于计算管控接收帧数private Map<Integer, BlockingQueue<byte[]>> currTimeFlagToQueue = new HashMap<>();int threadNum = 1;private ExecutorService threadPool;private Counter counter = new Counter();public DealGkDataServiceSuperWithThreadPool() {int corePoolSize = Runtime.getRuntime().availableProcessors() * 2;System.out.println("核心线程数：   " + corePoolSize);int maximumPoolSize = corePoolSize * 2;long keepAliveTime = 60L;threadPool = new ThreadPoolExecutor(corePoolSize,maximumPoolSize,keepAliveTime,TimeUnit.SECONDS,new LinkedBlockingQueue<>());}public void dealGKApi(ByteBuffer byteBuffer) {currentReceivedFrameNum++;if (currentReceivedFrameNum % 1000 == 0) {System.out.println("管控当前接收的数据帧数(每隔1000打印一次)： " + currentReceivedFrameNum);}int currTimeFlag = byteBuffer.getInt();       //当前时间标识，用于区分是否丢包int packagesTotalNum = byteBuffer.getInt();   //表示当前发送的包总数int currPackageNum = byteBuffer.getInt();     //表示当前包序号int messageLength = byteBuffer.getInt();     //消息长度int remainingBytes = byteBuffer.remaining();byte[] remainingData = new byte[messageLength];   //用于获取日志长度 1024if (remainingBytes > 0) {byteBuffer.get(remainingData); // 获取剩余的字节threadPool.submit(new GKRunnable(remainingData));}}class GKRunnable implements Runnable {private byte[] bytes;public GKRunnable(byte[] remainingData) {this.bytes = remainingData;}@Overridepublic void run() {System.out.println("新启动一个线程用于处理管控日志，当前线程名:  " + Thread.currentThread().getName());ByteBuffer byteBuffer = ByteBuffer.allocate(bytes.length);   //将不用byte数组放到ByteBuffer中byteBuffer.put(bytes);dealWithAssembledFrame(byteBuffer);}private void dealWithAssembledFrame(ByteBuffer byteBuffer) {byteBuffer.flip();byteBuffer.order(ByteOrder.LITTLE_ENDIAN);  //转化为小端dealLogPart(byteBuffer);}/*todo 要写的内容为：出现错误就要将异常抛出，并且找到帧头的位置*/private void dealLogPart(ByteBuffer byteBuffer) {
//            /*找到帧头的位置*/
//            searchFrameHeaderAndSetPosition(byteBuffer);while (byteBuffer.position() != byteBuffer.capacity() && (byteBuffer.position() + 4) < byteBuffer.capacity()) {try {/*找到帧头的位置*/searchFrameHeaderAndSetPosition(byteBuffer);int startPosition = byteBuffer.position();//获取开始的长度//每个日志都包含的部分byte[] bytes2 = new byte[2];byteBuffer.get(bytes2);String frameHeaderInfo = ConvertUtil.byteArrayToHexString(bytes2);   //日志帧头字段short logLength = byteBuffer.getShort();                             //日志长度int logNumber = byteBuffer.getShort();                             //日志编号
//                    System.out.println(logNumber + logLength);byte logType = byteBuffer.get();//byte转化为字符串                     //日志类型String logTypeStr = String.format("%02X", logType);int time = byteBuffer.getInt();                //日志时间//根据日志类型，选择处理日志剩余方式if ("01".equals(logTypeStr) || "02".equals(logTypeStr) || "03".equals(logTypeStr)) {byte sendingAndReceivingBit = byteBuffer.get();byte sourceDeviceId = byteBuffer.get();byte targetDeviceId = byteBuffer.get();//得到日志内容长度    先将日志长度转化为十进制，然后减掉帧头信息，减掉日志长度，减掉日志编号，减掉日志类型、减掉时间，减掉校验码，减掉发送接收位，减掉源原设备ID，减掉目标设备IDint logContentLength = logLength - 2 - 2 - 2 - 1 - 4 - 1 - 1 - 1 - 1;String instructDataContent = null;if ("01".equals(logTypeStr)) {  //子地址+数据内容，子地址占1个字节byte[] bytes = new byte[1];byteBuffer.get(bytes);String subAddress = ConvertUtil.byteArrayToHexString(bytes);int dataContentLength = logContentLength - 1;byte[] bytes1 = new byte[dataContentLength];// 输出剩余字节的十六进制表示,即指令数据内容byteBuffer.get(bytes1);instructDataContent = ConvertUtil.byteArrayToHexString(bytes1);dealInstruct1553bLog(byteBuffer, startPosition, logLength, logNumber & 0xFFFF, time, sendingAndReceivingBit, sourceDeviceId, targetDeviceId, subAddress, instructDataContent);} else if ("02".equals(logTypeStr)) { //can指令类型（1个字节）  + ID（子地址）+数据内容，ID占4个字节byte[] bytes = new byte[1];byteBuffer.get(bytes);String canInstructType = ConvertUtil.byteArrayToHexString(bytes);byte[] bytes1 = new byte[4];byteBuffer.get(bytes1);String subAddress = ConvertUtil.byteArrayToHexString(bytes1);int dataContentLength = logContentLength - 1 - 4;byte[] ID = new byte[dataContentLength];// 输出剩余字节的十六进制表示,即指令数据内容byteBuffer.get(ID);instructDataContent = ConvertUtil.byteArrayToHexString(ID);dealInstructCANLog(byteBuffer, startPosition, logLength, logNumber & 0xFFFF, time, sendingAndReceivingBit, sourceDeviceId, targetDeviceId, canInstructType, subAddress, instructDataContent);} else if ("03".equals(logTypeStr)) { //数据内容byte[] bytes1 = new byte[logContentLength];byteBuffer.get(bytes1);instructDataContent = ConvertUtil.byteArrayToHexString(bytes1);dealInstructRs422Log(byteBuffer, startPosition, logLength, logNumber & 0xFFFF, time, sendingAndReceivingBit, sourceDeviceId, targetDeviceId, instructDataContent);}} else if ("04".equals(logTypeStr)) {//存储到excel表中
//                        dealStateLog(byteBuffer, startPosition, time);//存储到数据库中dealStateLog(byteBuffer, startPosition, logLength, logNumber & 0xFFFF, time);} else if ("05".equals(logTypeStr)) {dealOperateLog(byteBuffer, startPosition, logLength, logNumber & 0xFFFF, time);} else if ("06".equals(logTypeStr)) {dealErrorLog(byteBuffer, startPosition, logLength, logNumber & 0xFFFF, time);} else {System.out.println("as;dasd");}} catch (Exception e) {if (e.getCause().toString().contains("SQLIntegrityConstraintViolationException")) {counter.increment();}else {e.printStackTrace();//                        System.err.println(e.getMessage());}//在处理过程报错了，那么就认为日志内容错误，就要重新寻找帧头的位置searchFrameHeaderAndSetPosition(byteBuffer);}}System.out.println("线程名称：" + Thread.currentThread().getName() + "  线程id：" + Thread.currentThread().getId() + "  处理完成！");System.out.println("========"+counter.getCount());}private void dealStateLog(ByteBuffer byteBuffer, int startPosition, short logLength, int logNumber, int time) {byte[] bytes92 = new byte[92];  //将92个长度传递到byte数组中byteBuffer.get(bytes92);int endPosition = byteBuffer.position();byte checkCodeByte = calculateCheckCode(byteBuffer, startPosition, endPosition);byte checkCode = byteBuffer.get();  //用于校验日志的正确和完整性boolean logIsIntegrity = logIsIntegrity(checkCode, checkCodeByte);if (logIsIntegrity) {StateLogInfo stateLogInfo = new StateLogInfo();stateLogInfo.setLogLength(Short.toString(logLength));stateLogInfo.setLogNumber(Long.valueOf(logNumber));stateLogInfo.setTime(Long.valueOf(time));
//                Integer size = stateLogInfoMapper.searchDataIsDuplicate(stateLogInfo);
//                if (size > 0) { //判断数据库中是否已经存在
//                    return;
//                }//将参数设置到stateLogInfo实例中setParameter(bytes92, stateLogInfo);stateLogInfoMapper.insertStateLogInfo(stateLogInfo);} else {logIsNotIntegrity(byteBuffer, startPosition, endPosition);}}private void searchFrameHeaderAndSetPosition(ByteBuffer byteBuffer) {/*找到帧头的位置*/int frameHeaderPosition = findFrameHeaderPosition(byteBuffer, hexStringToByteArray("eafc"));if (frameHeaderPosition != -1) {byteBuffer.position(frameHeaderPosition);} else {System.out.println("未找到帧头为 eafc 的位置");return;   //说明从头查到尾都没有查到，就直接退出}}/*** 判断日志是否完整的函数，如果不完整，那么就要找下一个eafc帧头的位置** @param byteBuffer*/private void logIsNotIntegrity(ByteBuffer byteBuffer, int startPosition, int endPosition) {System.out.println("日志不完整！丢掉");//如果日志不完整，那么就要找到下一帧的头位置/*找到帧头的位置*/searchFrameHeaderAndSetPosition(byteBuffer);}/*** 判断是否为完整日志的函数isIntegrity（日志长度 + 帧头 - 校验码）** @param a* @param b* @return*/private boolean logIsIntegrity(byte a, byte b) {return a == b;}/*** 计算校验码的函数** @param byteBuffer* @param startPosition* @param endPosition* @return*/private byte calculateCheckCode(ByteBuffer byteBuffer, int startPosition, int endPosition) {int length = endPosition - startPosition;byteBuffer.position(startPosition);byte res = 0;for (int i = 0; i < length; i++) {byte b = byteBuffer.get();res += b;}return res;}/*** 通过遍历的方式得到帧头的位置** @param byteBuffer* @param frameHeader* @return*/public int findFrameHeaderPosition(ByteBuffer byteBuffer, byte[] frameHeader) {// 记录当前位置int startPosition = byteBuffer.position();// 遍历 ByteBuffer 从当前位置开始搜索帧头for (int i = startPosition; i < byteBuffer.limit() - frameHeader.length + 1; i++) {// 标记当前位置byteBuffer.position(i);boolean found = true;for (int j = 0; j < frameHeader.length; j++) {if (byteBuffer.get() != frameHeader[j]) {found = false;break;}}if (found) {// 恢复 ByteBuffer 的当前位置
//                byteBuffer.position(startPosition);return i; // 返回帧头 'e' 的位置}}// 恢复 ByteBuffer 的当前位置
//        byteBuffer.position(startPosition);// 如果没有找到，返回 -1 表示未找到return -1;}/*** 将十六进制字符串转换为字节数组的方法** @param hexString* @return*/public byte[] hexStringToByteArray(String hexString) {int len = hexString.length();byte[] byteArray = new byte[len / 2]; // 每两个十六进制字符表示一个字节for (int i = 0; i < len; i += 2) {byteArray[i / 2] = (byte) ((Character.digit(hexString.charAt(i), 16) << 4)+ Character.digit(hexString.charAt(i + 1), 16));}return byteArray;}}}class Counter {private LongAdder count = new LongAdder();public void increment() {count.increment();}public int getCount() {return count.intValue();}
}

解释：以上代码是采用多线程，将接收到的数据帧解析为日志，并将日志存储到数据库中，根据日志类型不同，存储到不同的数据库表中。