第七十三篇从电影院售票到停车场计数：生活场景解析Java原子类精髓

- 一、原子类基础：电影院售票系统
- - 1.1 传统售票的并发问题
  - 1.2 原子类解决方案
- 二、原子类家族：超市收银系统
- - 2.1 基础类型原子类
  - 2.2 数组类型原子类
- 三、CAS机制深度解析：停车场管理系统
- - 3.1 CAS工作原理
  - 3.2 车位计数器实现
- 四、高性能实践：银行账户系统
- - 4.1 账户余额更新
  - 4.2 性能对比测试
- 五、原子类进阶：电影院选座系统
- - 5.1 座位状态管理
  - 5.2 批量更新优化

想象电影院售票处：多个窗口同时售票却不会售出重复座位；停车场入口自动计数系统实时更新车位状态——这正是Java原子类的完美生活映射。本文将用日常场景揭秘高并发环境下的无锁编程奥秘。

一、原子类基础：电影院售票系统

1.1 传统售票的并发问题

// 非原子操作导致超卖
public class TicketCounter {private int totalTickets = 100;public void sellTicket() {if(totalTickets > 0) {// 此处可能被其他线程中断totalTickets--; System.out.println("售出1张票，剩余：" + totalTickets);}}
}

1.2 原子类解决方案

public class AtomicTicketCounter {private final AtomicInteger tickets = new AtomicInteger(100);public void safeSellTicket() {int current;do {current = tickets.get();if(current <= 0) {System.out.println("票已售罄");return;}} while(!tickets.compareAndSet(current, current - 1));System.out.println("安全售出，剩余：" + tickets.get());}
}

二、原子类家族：超市收银系统

2.1 基础类型原子类

// 收银台计数器
AtomicLong dailySales = new AtomicLong(0); // 每笔交易更新
void processTransaction(double amount) {// 累加销售额dailySales.addAndGet((long)(amount * 100));// 更新最高单笔交易AtomicReference<Double> maxTransaction = new AtomicReference<>(0.0);maxTransaction.accumulateAndGet(amount, Math::max);
}

2.2 数组类型原子类

// 货架商品库存
AtomicIntegerArray shelfStock = new AtomicIntegerArray(10); // 补充第3号货架商品
void restockShelf(int shelfId, int quantity) {shelfStock.addAndGet(shelfId, quantity);
}// 顾客购买商品
boolean purchaseItem(int shelfId) {int current;do {current = shelfStock.get(shelfId);if(current <= 0) return false;} while(!shelfStock.compareAndSet(shelfId, current, current-1));return true;
}

三、CAS机制深度解析：停车场管理系统

3.1 CAS工作原理

3.2 车位计数器实现

public class ParkingLot {private final AtomicInteger availableSpots;private final int totalSpots;public ParkingLot(int capacity) {this.totalSpots = capacity;this.availableSpots = new AtomicInteger(capacity);}// 车辆进入public boolean carEnter() {int current;do {current = availableSpots.get();if(current <= 0) return false;} while(!availableSpots.compareAndSet(current, current - 1));updateDisplay();return true;}// 车辆离开public void carExit() {availableSpots.incrementAndGet();updateDisplay();}private void updateDisplay() {System.out.println("当前车位: " + availableSpots + "/" + totalSpots);}
}

四、高性能实践：银行账户系统

4.1 账户余额更新

public class BankAccount {private final AtomicLong balance;public BankAccount(long initial) {this.balance = new AtomicLong(initial);}// 无锁存款public void deposit(long amount) {balance.addAndGet(amount);}// 安全取款（避免透支）public boolean withdraw(long amount) {long current;do {current = balance.get();if(current < amount) return false;} while(!balance.compareAndSet(current, current - amount));return true;}// 转账操作public boolean transferTo(BankAccount target, long amount) {if(this.withdraw(amount)) {target.deposit(amount);return true;}return false;}
}

4.2 性能对比测试

操作方式	100万次操作耗时	适用场景
synchronized	450ms	复杂事务处理
ReentrantLock	380ms	需要高级功能的场景
AtomicLong	120ms	简单计数器场景

五、原子类进阶：电影院选座系统

5.1 座位状态管理

public class SeatManager {// 使用AtomicReference数组管理座位状态private final AtomicReference<SeatStatus>[] seats;public SeatManager(int capacity) {seats = new AtomicReference[capacity];// 初始化所有座位为空闲Arrays.fill(seats, new AtomicReference<>(SeatStatus.AVAILABLE));}// 预订座位public boolean bookSeat(int seatId) {return seats[seatId].compareAndSet(SeatStatus.AVAILABLE, SeatStatus.OCCUPIED);}// 释放座位public void releaseSeat(int seatId) {seats[seatId].set(SeatStatus.AVAILABLE);}enum SeatStatus { AVAILABLE, OCCUPIED }
}

5.2 批量更新优化

// 使用LongAdder优化高频计数器
public class TicketCounter {private final LongAdder totalSales = new LongAdder();private final LongAdder vipSales = new LongAdder();public void sellTicket(boolean isVip) {totalSales.increment();if(isVip) vipSales.increment();}// 生成报表public void generateReport() {System.out.println("总售票: " + totalSales.sum());System.out.println("VIP票: " + vipSales.sum());}
}