RabbitMq consumption information, data processing and encoding settings

RabbitMq consumption information, data processing and encoding settings

Putting the message on the MQ queue must process the message, otherwise it would be meaningless to use MQ.

Ordinary consumption

Simply consuming messages only needs to write an implementation class that implements the onMessage method of MessageListener , such as

public class RabbitmqService implements MessageListener {

    private Logger logger = LoggerFactory.getLogger(RabbitmqService.class);
    
    @Autowired
    private UserService userService;

    @Autowired
    private MsgInfoService msgInfoService;

    public void onMessage(Message message) {
      
        try {
        	//将放入队列的消息已UTF-8的格式转为JSON字符串（前提是放入队列是也是JSON字符串，中文存在乱码）
            String  msg  = new String(message.getBody(),"utf-8");
            logger.info("消息：{}",JSON.toJSONString(msg));
            //将JSON字符串转换为自己想要处理的对象
            User user= JSON.parseObject(msg, User .class);
            //处理逻辑
            User userDto = userService.save(user);
            logger.info("调用结果返回：{}",JSON.toJSONString(userDto ));
        } catch (Exception e) {
            e.printStackTrace();
        }
    }

}

Manually confirm consumption

Manual confirmation means that you can choose your own processing method when there is an exception in message consumption, reenter the message into the queue, and persist the message. To manually confirm the message, you only need to implement the onMessage method of ChannelAwareMessageListener

public class AckConsumer implements ChannelAwareMessageListener {
    private Logger logger = LoggerFactory.getLogger(AckConsumer.class);
    @Autowired
    private UserService userService;

    @Autowired
    private MsgInfoService msgInfoService;

    @Override
    public void onMessage(Message message, Channel channel) throws Exception {
        logger.info("接受到消息：{}",message);
        channel.basicQos(1);//每次处理一条消息
        String state = null;
        try{
            String  msg = new String(message.getBody(),"utf-8");
            logger.info("处理的数据：{}",JSON.toJSONString(msg));
            User user= JSON.parseObject(msg, User.class);
            User userDto = userService.save(user);
            String returnCode = userDto .getReturnCode();
            logger.info("调用结果返回：{}",JSON.toJSONString(userDto ));
            if(returnCode.equals("1000")){
                state = "10A";
                //第一个参数Delivery Tag 用来标识信道中投递的消息,RabbitMQ 推送消息给 Consumer 时,
                // 会附带一个 Delivery Tag，以便 Consumer可以在消息确认时告诉 RabbitMQ 到底是哪条消息被确认了.
                //第二个参数 multiple 取值为 false 时，表示通知 RabbitMQ 当前消息被确认；
                // 如果为 true，则额外将比第一个参数指定的 delivery tag 小的消息一并确认。
                channel.basicAck(message.getMessageProperties().getDeliveryTag(),false);//确认消息已消费
            }else{
                state = "10I";
                //第一个参数，消息的标识
                //第二个参数是否批量处理消息,true:将一次性拒绝所有小于deliveryTag的消息。
                //第三个参数，消息是否重入队列，false将消息存队列删除。true:会重复消费该消息直到消息被确认消费
                channel.basicNack(message.getMessageProperties().getDeliveryTag(),false,false);//拒绝消费消息
            }
        }catch (Exception e){
            state = "10I";
            e.printStackTrace();
            channel.basicNack(message.getMessageProperties().getDeliveryTag(),false,false);
        }finally {
            //消息持久化
            msgInfoService.saveMsg(message, state);
        }
    }
}