实现以上功能
方法1:定时任务批量执行
写一个定时任务,每隔 30分钟执行一次,列出所有超出时间范围得订单id的列表
@Async
@Scheduled(cron = "20 20 1 * * ?")
public void cancelOrder(){
log.info("【取消订单任务开始】");
QueryWrapper<Order> qw = new QueryWrapper<>();
qw.eq("status", Constants.OrderStatus.NOTPAID);
qw.eq("aftersale_status", 1);
List<Order> orderList = orderMapper.selectList(qw);
List<Long> idList = orderList.stream()
.filter(order -> LocalDateTimeUtil.between(order.getCreateTime(), LocalDateTime.now()).toMinutes() >= 15)
.map(Order::getId)
.collect(Collectors.toList());
CancelOrderRequest request = new CancelOrderRequest();
request.setIdList(idList);
h5OrderService.orderBatchCancel(request, null);
log.info("【取消订单任务结束】");
}
批量执行取消订单操作
@Transactional
public String orderBatchCancel(CancelOrderRequest request, Long userId) {
LocalDateTime optDate = LocalDateTime.now();
if (CollectionUtil.isEmpty(request.getIdList())) {
throw new RuntimeException("未指定需要取消的订单号");
}
QueryWrapper<Order> orderQw = new QueryWrapper<>();
orderQw.in("id", request.getIdList());
List<Order> orderList = orderMapper.selectList(orderQw);
if (orderList.size() < request.getIdList().size()) {
throw new RuntimeException("未查询到订单信息");
}
Order order = orderList.get(0);
//查orderItem
QueryWrapper<OrderItem> qw = new QueryWrapper<>();
qw.in("order_id", request.getIdList());
List<OrderItem> orderItems = orderItemMapper.selectList(qw);
if (CollectionUtil.isEmpty(orderItems)) {
throw new RuntimeException("未查询到订单信息");
}
long count = orderList.stream().filter(it -> !Constants.H5OrderStatus.UN_PAY.equals(it.getStatus())).count();
if (count > 0) {
throw new RuntimeException("订单状态已更新,请刷新页面");
}
List<OrderOperateHistory> addHistoryList = new ArrayList<>();
orderList.forEach(item -> {
item.setStatus(Constants.H5OrderStatus.CLOSED);
item.setUpdateTime(optDate);
item.setUpdateBy(userId);
OrderOperateHistory history = new OrderOperateHistory();
history.setOrderId(item.getId());
history.setOrderSn(item.getOrderSn());
history.setOperateMan(userId == null ? "后台管理员" : "" + item.getMemberId());
history.setOrderStatus(Constants.H5OrderStatus.CLOSED);
history.setCreateTime(optDate);
history.setCreateBy(userId);
history.setUpdateBy(userId);
history.setUpdateTime(optDate);
addHistoryList.add(history);
});
//取消订单
int rows = orderMapper.cancelBatch(orderList);
if (rows < 1) {
throw new RuntimeException("更改订单状态失败");
}
orderItems.stream().collect(Collectors.groupingBy(it -> it.getSkuId())).forEach((k, v) -> {
AtomicReference<Integer> totalCount = new AtomicReference<>(0);
v.forEach(it -> totalCount.updateAndGet(v1 -> v1 + it.getQuantity()));
skuMapper.updateStockById(k, optDate, -1 * totalCount.get());
});
//创建订单操作记录
boolean flag = orderOperateHistoryService.saveBatch(addHistoryList);
if (!flag) {
throw new RuntimeException("创建订单操作记录失败");
}
// 根据order 退还积分
orderUsePointsService.refundOrderUsePoints(order.getId());
return "取消订单成功";
}
方法2:使用jdk自带的阻塞队列
实现一个简单的队列,每隔一定时间执行队列。
/**
* (30分钟扫描三十分钟内需要发送的订单)
*/
@Scheduled(cron = "0 0/30 * * * ?")
public void checkOrderStatus() {
DelayQueue<ItemVo<Order>> queue = new DelayQueue<ItemVo<Order>>();
try {
// 插入订单
new Thread(new PutOrder(queue)).start();
} catch (Exception e) {
e.printStackTrace();
}
}
这里使用队列的优势可以跟前端时间匹配上,前端读秒几秒这里就什么时候取消
import com.alibaba.fastjson.JSON;
import com.baomidou.mybatisplus.toolkit.CollectionUtils;
import com.kxmall.market.biz.BeanContext;
import com.kxmall.market.data.dto.PrivatePlanAndDetailDO;
import com.kxmall.market.data.mapper.PrivatePlanMapper;
import com.kxmall.market.data.util.DateUtil;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.springframework.beans.factory.annotation.Autowired;
import java.util.Date;
import java.util.List;
import java.util.concurrent.DelayQueue;
/**
* 模拟订单插入的功能
*/
public class PutOrder implements Runnable {
private static final Logger logger = LoggerFactory.getLogger(PutOrder.class);
// public static PutOrder putOrder;
@Autowired
private PrivatePlanMapper privatePlanMapper;
// 使用DelayQueue:一个使用优先级队列实现的无界阻塞队列。
private DelayQueue<ItemVo<Order>> queue;
public PutOrder(DelayQueue<ItemVo<Order>> queue) {
super();
this.queue = queue;
}
@Override
public void run() {
Date startTime = new Date();
privatePlanMapper = BeanContext.getApplicationContext().getBean(PrivatePlanMapper.class);
// 每隔半小时获取半小时内需要取消的
List<PrivatePlanAndDetailDO> privatePlanDOS = privatePlanMapper.getPrivatePlanDetailList();
logger.info("待取消清单->{}", JSON.toJSONString(privatePlanDOS));
if (CollectionUtils.isNotEmpty(privatePlanDOS)) {
privatePlanDOS.forEach(s -> {
long count = DateUtil.calLastedTime(startTime,s.getTodoTime() )*1000;
Order tbOrder = new Order(s.getId().toString(), 0.0);
ItemVo<Order> itemVoTb = new ItemVo<Order>(count, tbOrder);
queue.offer(itemVoTb);
logger.info("订单{}将在->{}秒后取消", s.getId().toString(),count/1000);
});
// 取出过期订单的线程
new Thread(new FetchOrder(queue)).start();
}else {
logger.info("没有待发送订单->");
}
}
}
import java.util.concurrent.Delayed;
import java.util.concurrent.TimeUnit;
/**
* 存到队列里的元素
* 支持延时获取的元素的阻塞队列,元素必须要实现Delayed接口。
* 根据订单有效时间作为队列的优先级
* @param <T>
*/
public class ItemVo<T> implements Delayed{
// 到期时间 单位:ms
private long activeTime;
// 订单实体(使用泛型是因为后续扩展其他业务共用此业务类)
private T data;
public ItemVo(long activeTime, T data) {
super();
// 将传入的时间转换为超时的时刻
this.activeTime = TimeUnit.NANOSECONDS.convert(activeTime, TimeUnit.MILLISECONDS)
+ System.nanoTime();
this.data = data;
}
public long getActiveTime() {
return activeTime;
}
public T getData() {
return data;
}
// 按照剩余时间进行排序
@Override
public int compareTo(Delayed o) {
// 订单剩余时间-当前传入的时间= 实际剩余时间(单位纳秒)
long d = getDelay(TimeUnit.NANOSECONDS) - o.getDelay(TimeUnit.NANOSECONDS);
// 根据剩余时间判断等于0 返回1 不等于0
// 有可能大于0 有可能小于0 大于0返回1 小于返回-1
return (d == 0) ? 0 : ((d > 0) ? 1 : -1);
}
// 获取剩余时间
@Override
public long getDelay(TimeUnit unit) {
// 剩余时间= 到期时间-当前系统时间,系统一般是纳秒级的,所以这里做一次转换
long d = unit.convert(activeTime-System.nanoTime(), TimeUnit.NANOSECONDS);
return d;
}
}
方法3:分布式场景(mq队列)
使用mq队列,消费消息。如果消息到达30分钟没有付款,那么就取消
方法4:分布式场景(redis)
使用redis商品下单,设置过期时间 30分钟,并且写一个redis监听器,监听过期需要操作的key,然后判单是否过期