本篇文章为你整理了Spring AOP中增强Advice的执行顺序(spring增强类型并说明特点)的详细内容,包含有springaop增强处理类有哪些 spring增强类型并说明特点 spring增强器 springboot增强 Spring AOP中增强Advice的执行顺序,希望能帮助你了解 Spring AOP中增强Advice的执行顺序。
本文主要验证Spring AOP中Advice的执行顺序问题。(Spring版本: 5.3.23)
Spring AOP中Advice分类
Spring AOP中Advice可分为如下五类:
@Around
@Before
@AfterReturning
@AfterThrowing
@After
public static void main(String[] args) {
SpringApplication.run(MySpringApplication.class, args);
定义一个用于测试的Controller类
package sakura.springinaction.controller;
@Controller
@Slf4j
public class IndexController {
@GetMapping("/time")
@ResponseBody
public String time() {
LocalDateTime now = LocalDateTime.now();
String nowTime = now.format(DateTimeFormatter.ISO_LOCAL_DATE_TIME);
log.info("Current time: " + nowTime);
return nowTime;
定义一个声明式切面 Apsect1
@Slf4j
@Component
@Aspect
public class Aspect1 {
// 定义 Point Cut 切面
@Pointcut("execution(public * sakura.springinaction.controller.*.*(..))")
public void controllerLayer() {
// 定义Advice
@Before("controllerLayer()")
private void beforeAdvice2() {
log.info("Aspect_1 # @Before");
@After("controllerLayer() @annotation(getMapping)")
private void afterAdvice1(GetMapping getMapping) {
log.info("Aspect_1 # @afterAdvice" + " path: " + Arrays.toString(getMapping.value()));
@AfterReturning(pointcut = "controllerLayer()", returning = "val")
private void afterReturningAdvice(Object val) {
log.info("Aspect_1 # @AfterReturning" + " returnValue: " + val);
@AfterThrowing(pointcut = "controllerLayer()", throwing = "thrower")
private void afterThrowingAdvice(Throwable thrower) {
log.info("Aspect_1 # @AfterThrowing" + " thrower: " + thrower.getClass().getName());
@Around("controllerLayer() @annotation(getMapping)")
private Object aroundAdvice(ProceedingJoinPoint pjp, GetMapping getMapping) throws Throwable {
// Around 前置处理
Stopwatch stopwatch = Stopwatch.createStarted();
log.info("Aspect_1 # @Around-Before" + " methodName: " + pjp.getSignature().getName() + ", path: " + Arrays.toString(getMapping.value()));
Object result = pjp.proceed();
// Around 后置处理
log.info("Aspect_1 # @Around-After" + " methodName: " + pjp.getSignature().getName() + ", runTime: " + stopwatch.elapsed(TimeUnit.NANOSECONDS));
return result;
在 发起请求(http://localhost:8080/time) 后,日志输出如图:
在同一个切面(Apsect)定义中对于同一个Join Point而言,不同类型的Advice执行先后顺序依次是:
@Around 前置处理
@Before
@AfterReturning/@AfterThrowing
@After
@Around 后置置处理
对于进入Join Point的Advice而言(比如: @Around 前置处理,@Before),优先级越高,越先执行;
对于从Join Point出来的Advice而言(比如: @Around 后置处理,@After),优先级越高,越后执行;
优先级从高到低依次为: @Around, @Before,@After,@AfterReturning,@AfterThrowing;
PS:
如果在同一个切面(Apsect)中定义了两个同类型的Advice(比如定义两个@Before), 对于某个Join Point而言这两个Advice都匹配,那么这两个Advice执行的先后顺序是无法确定的。
不同Aspect中Advice执行顺序
问: 当不同的Aspect中的Advice 都匹配到了同一个Join Point,那么那个Aspect中的Advice 先执行,那个后执行呢?
答: 不确定 ,但是可以通过在class上添加注解@Order指定优先级确定执行顺序(参考文档)
实验一: Aspect1为高优先级,Aspect2为低优先级
与Aspect1 类似,再定义一个切面类Aspect2,如下
package sakura.springinaction.advice;
import org.springframework.core.annotation.Order;
@Slf4j
@Component
@Aspect
@Order(2)
public class Aspect2 {
// 定义Advice
@Before("sakura.springinaction.advice.Aspect1.controllerLayer()")
private void beforeAdvice2() {
log.info("Aspect_2 # @Before");
@After("sakura.springinaction.advice.Aspect1.controllerLayer() @annotation(getMapping)")
private void afterAdvice1(GetMapping getMapping) {
log.info("Aspect_2 # @afterAdvice" + " path: " + Arrays.toString(getMapping.value()));
@AfterReturning(pointcut = "sakura.springinaction.advice.Aspect1.controllerLayer()", returning = "val")
private void afterReturningAdvice(Object val) {
log.info("Aspect_2 # @AfterReturning" + " returnValue: " + val);
@AfterThrowing(pointcut = "sakura.springinaction.advice.Aspect1.controllerLayer()", throwing = "thrower")
private void afterThrowingAdvice(Throwable thrower) {
log.info("Aspect_2 # @AfterThrowing" + " thrower: " + thrower.getClass().getName());
@Around("sakura.springinaction.advice.Aspect1.controllerLayer() @annotation(getMapping)")
private Object aroundAdvice(ProceedingJoinPoint pjp, GetMapping getMapping) throws Throwable {
Stopwatch stopwatch = Stopwatch.createStarted();
log.info("Aspect_2 # @Around-Before" + " methodName: " + pjp.getSignature().getName() + ", path: " + Arrays.toString(getMapping.value()));
Object result = pjp.proceed();
log.info("Aspect_2 # @Around-After" + " methodName: " + pjp.getSignature().getName() + ", runTime: " + stopwatch.elapsed(TimeUnit.NANOSECONDS));
return result;
Aspect1 添加@Order注解指定优先级,如下
@Slf4j
@Component
@Aspect
@Order(1)
public class Aspect1 {
//...
此时,Aspect1的优先级比Aspect2的优先级高。
实验结果如下:
说明:
高优先级的Aspect1中的@Around前置处理和@Before先于低优先级的Aspect2执行,而@AfterReturning,@After和@Around后置处理,则低优先级的Aspect2先执行。
实验二: Aspect1为低优先级,Aspect2为高优先级
更改两个Aspect中@Order注解优先级,如下:
@Slf4j
@Component
@Aspect
@Order(2)
public class Aspect1 {
//...
@Slf4j
@Component
@Aspect
@Order(1)
public class Aspect2 {
//...
实验结果如下:
当不同的Aspect中的Advice 都匹配到了同一个Join Point,不同Aspect中的Advice 执行顺序不确定。
通过在Aspect类上添加注解@Order指定优先级,确定执行顺序,执行顺序满足如下规律
对于@Around前置处理 和@Before两种Advice而言,所在的Aspect优先级越高,越先执行
对于@AfterReturning,@AfterThrowing,@After和@Around后置处理 类型的Advice而言,所在的Aspect优先级越高,越后执行
以上就是Spring AOP中增强Advice的执行顺序(spring增强类型并说明特点)的详细内容,想要了解更多 Spring AOP中增强Advice的执行顺序的内容,请持续关注盛行IT软件开发工作室。
郑重声明:本文由网友发布,不代表盛行IT的观点,版权归原作者所有,仅为传播更多信息之目的,如有侵权请联系,我们将第一时间修改或删除,多谢。