深入源码分析SpringAOP实现原理

深入源码分析SpringAOP实现原理

阅读这篇文章前,最好有代理模式的基础,以及了解关于Spring扩展点例如BeanPostProcessor和如何使用自定义标签集成Spring,这些文章在我的博客里都能找到。当然,也最好有使用AOP的经验,这篇文章不会讲解如何使用AOP。

1. AOP简介

说到AOP,其实这是一个面向方面的编程思想 ,它解决了OOP的一些弊端,例如我们需要为多个不具有继承关系的类引入一个公共行为, 比如说日志、权限验证、事务管理等等,我们需要将这些代码重复的添加到一系列的类中,将产生大量的重复代码,如果需要修改,将在每个类中去进行修改,不便于维护,代码的侵入性极高。所以就有了AOP这样面向方面编程的编程思想,其功能可以为每个需要的类加入共同的行为 ,如果需要修改,只需要修改切面中的代码,改一处等于改多处,并且便于编程,写一个切面类即可达到在每个类中加入重复代码的目的。

阅读此篇文章,你将了解Spring是如何实现AOP(前置通知、后置通知、环绕通知),由于Spring中的事务管理是基于AOP的功能来做的,所以你将更好的能理解Spring是如何将事务统一管理起来的。

2. 自定义标签开启AOP

只要用过AOP都知道,如果需要使用AOP,需要在配置文件中写这样一段配置:

<aop:aspectj-autoproxy />

只有写了这段配置才可以开启AOP功能,那么这个自定义标签又做了什么呢?在上一篇讲解自定义标签的文章中详细讲到了,此时我们需要关注其标签头aop去寻找对应的命名空间:

xmlns:aop="http://www.springframework.org/schema/aop"

全局搜索命名空间http\://www.springframework.org/schema/aop,注意http后加一个 “\” ,可以找到spring.handlers文件中对应的handler类:

http\://www.springframework.org/schema/aop=org.springframework.aop.config.AopNamespaceHandler

这样就找到了命名空间对应的handler:

public class AopNamespaceHandler extends NamespaceHandlerSupport {

	/**
	 * Register the {@link BeanDefinitionParser BeanDefinitionParsers} for the
	 * '{@code config}', '{@code spring-configured}', '{@code aspectj-autoproxy}'
	 * and '{@code scoped-proxy}' tags.
	 */
	@Override
	public void init() {
		// In 2.0 XSD as well as in 2.1 XSD.
		registerBeanDefinitionParser("config", new ConfigBeanDefinitionParser());
		registerBeanDefinitionParser("aspectj-autoproxy", new AspectJAutoProxyBeanDefinitionParser());
		registerBeanDefinitionDecorator("scoped-proxy", new ScopedProxyBeanDefinitionDecorator());

		// Only in 2.0 XSD: moved to context namespace as of 2.1
		registerBeanDefinitionParser("spring-configured", new SpringConfiguredBeanDefinitionParser());
	}
}

回到开头配置自定义标签,我们使用了aspectj-autoproxy这个Parser。在init方法中,我们找到aspectj-autoproxy对应的Parser是AspectJAutoProxyBeanDefinitionParser这个类:

@Override
@Nullable
//我们只关注解析的主方法,parse方法
public BeanDefinition parse(Element element, ParserContext parserContext) {
    //注册一个类到IOC容器中
    AopNamespaceUtils.registerAspectJAnnotationAutoProxyCreatorIfNecessary(parserContext, element);
    extendBeanDefinition(element, parserContext);
    return null;
}

3. 注册AnnotationAwareAspectJAutoProxyCreator

AnnotationAwareAspectJAutoProxyCreator 是实现AOP功能的主要类,我们先来看看这个类的结构:
在这里插入图片描述
此类实现了BeanPostProcessor ,稍后将关注其后置处理Bean的方法postProcessAfterInitialization ,并且实现了BeanFactorAware接口,此类将取得并存有一个BeanFactory实例对象。

回到主线,关注注册此类的方法:

public static void registerAspectJAnnotationAutoProxyCreatorIfNecessary(
    ParserContext parserContext, Element sourceElement) {
//将一个类作为Bean注册到IOC容器中
    BeanDefinition beanDefinition = AopConfigUtils.registerAspectJAnnotationAutoProxyCreatorIfNecessary(
        parserContext.getRegistry(), parserContext.extractSource(sourceElement));
    //处理proxy-target-class与expose-proxy属性
    useClassProxyingIfNecessary(parserContext.getRegistry(), sourceElement);
    //注册组件并通知
    registerComponentIfNecessary(beanDefinition, parserContext);
}

其中,在注册这个类的过程中主要完成了3件事:

  1. 注册AnnotationAwareAspectJAutoProxyCreator

    @Nullable
    public static BeanDefinition registerAspectJAnnotationAutoProxyCreatorIfNecessary(BeanDefinitionRegistry registry,
                                                                                      @Nullable Object source) {
    //将AnnotationAwareAspectJAutoProxyCreator这个类注册到IOC容器中
        return registerOrEscalateApcAsRequired(AnnotationAwareAspectJAutoProxyCreator.class, registry, source);
    }
    
    @Nullable
    private static BeanDefinition registerOrEscalateApcAsRequired(Class<?> cls, BeanDefinitionRegistry registry,
                                                                  @Nullable Object source) {
    
        Assert.notNull(registry, "BeanDefinitionRegistry must not be null");
    
        //如果IOC容器中已经存在了此类型的Bean,则需要判断优先级
        if (registry.containsBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME)) {
            //获取此类的BeanDefinition信息
            BeanDefinition apcDefinition = registry.getBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME);
            //如果此Bean的ClassName与AnnotationAwareAspectJAutoProxyCreator类的
            //ClassName不同的话,判断优先级
            if (!cls.getName().equals(apcDefinition.getBeanClassName())) {
                int currentPriority = findPriorityForClass(apcDefinition.getBeanClassName());
                int requiredPriority = findPriorityForClass(cls);
                //如果已存在Bean优先级小于Creator的优先级
                if (currentPriority < requiredPriority) {
                    //将ClassName替换成Creator
                    apcDefinition.setBeanClassName(cls.getName());
                }
            }
            //不进行注册,因为已经注册了
            return null;
        }
    
        //如果到这里,说明IOC容器中没有配置对应Creator
        //使用Crearir的Class构造一个BeanDefinition
        RootBeanDefinition beanDefinition = new RootBeanDefinition(cls);
        beanDefinition.setSource(source);
        //配置依赖属性order,将其设置为最高优先级
        beanDefinition.getPropertyValues().add("order", Ordered.HIGHEST_PRECEDENCE);
        beanDefinition.setRole(BeanDefinition.ROLE_INFRASTRUCTURE);
        //将设置好属性的BeanDefinition注册进IOC容器中
        registry.registerBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME, beanDefinition);
        return beanDefinition;
    }
    
  2. 处理proxy-target-classexpose-proxy属性

    private static void useClassProxyingIfNecessary(BeanDefinitionRegistry registry, @Nullable Element sourceElement) {
        if (sourceElement != null) {
            boolean proxyTargetClass = Boolean.valueOf(sourceElement.getAttribute(PROXY_TARGET_CLASS_ATTRIBUTE));
            //处理proxy-target-class属性
            if (proxyTargetClass) {
                AopConfigUtils.forceAutoProxyCreatorToUseClassProxying(registry);
            }
            boolean exposeProxy = Boolean.valueOf(sourceElement.getAttribute(EXPOSE_PROXY_ATTRIBUTE));
            //处理expose-proxy属性
            if (exposeProxy) {
                AopConfigUtils.forceAutoProxyCreatorToExposeProxy(registry);
            }
        }
    }
    

    其中设置属性的过程:

    public static void forceAutoProxyCreatorToUseClassProxying(BeanDefinitionRegistry registry) {
        if (registry.containsBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME)) {
            //根据之前注册的BeanName取出Creator
            BeanDefinition definition = registry.getBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME);
            //将Creator的BeanDefinition的属性proxyTargetClass设置为true
            definition.getPropertyValues().add("proxyTargetClass", Boolean.TRUE);
        }
    }
    
    public static void forceAutoProxyCreatorToExposeProxy(BeanDefinitionRegistry registry) {
        if (registry.containsBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME)) {
            //根据之前注册的BeanName取出Creator
            BeanDefinition definition = registry.getBeanDefinition(AUTO_PROXY_CREATOR_BEAN_NAME);
            //将Creator的BeanDefinition的属性exposeProxy设置为true
            definition.getPropertyValues().add("exposeProxy", Boolean.TRUE);
        }
    }
    
    • proxy-target-class:在Spring的AOP中,默认的如果目标类实现了至少一个接口,将使用JDK动态代理实现AOP,否则使用CGLib动态代理实现AOP,如果希望AOP都使用CGLib实现,你就可以设置proxy-target-class属性为true,但要注意几个问题:

      • 无法对final的方法进行动态代理 ,原因很简单,CGLib使用继承实现,final方法无法重写,所以final的方法不能应用AOP。
      • 需要配置CGLib的JAR包
    • expose-proxy:在讲解事务的那篇文章中有提到,如果一个类中的事务A方法调用了同一个类中的事务B方法,B方法将没有事务,这个道理在AOP中也是这样的,相同类下的不同方法互相调用,内部方法将无法被应用通知(无法进行AOP),此时你需要将expose-proxy属性设置为true,暴露一个代理类(此属性的原理在下面会有详细讲解),然后在A方法中需要调用B方法的话需要这样写:

      public class Service{
          public void A(){
              ((Service)AopContext.currentProxy()).B();
          }
      
          public void B(){
      		//do something...
          }
      }
      

      这样,B方法就算再A方法内也可以被AOP。其中AopContext 是存放线程变量的类,形象的称之为AOP的上下文

4. 实现AOP代理

4.1 创建AOP代理

上面,自定义标签的配置完成了对Creator类的自动注册,我们可以知道,此类实现了BeanPostProcessor接口,将会在IOC容器初始化每个Bean时都调用此类的postProcessAfterInitialization 方法,此方法即为AOP代理的入口,此方法在抽象父类AbstractAutoProxyCreator实现:

@Override
public Object postProcessAfterInitialization(@Nullable Object bean, String beanName) throws BeansException {
    if (bean != null) {
        //先从缓存中获取Key,由要代理的Bean的Class与benaName组成
        Object cacheKey = getCacheKey(bean.getClass(), beanName);
        //判断是否是过早暴露的Bean,此概念在讲IOC解决循环依赖中有提到
        //如果是过早暴露的Bean,则此时连依赖注入都没有完成,则不对其进行代理
        //待其真正初始化之后再尝试代理
        if (!this.earlyProxyReferences.contains(cacheKey)) {
            //如果符合条件进行AOP代理
            return wrapIfNecessary(bean, beanName, cacheKey);
        }
    }
    return bean;
}
protected Object wrapIfNecessary(Object bean, String beanName, Object cacheKey) {
    //如果先前已经处理过的,不进行处理
    if (StringUtils.hasLength(beanName) && this.targetSourcedBeans.contains(beanName)) {
        return bean;
    }
    //如果此Bean已经被标记为无法代理,不进行处理
    if (Boolean.FALSE.equals(this.advisedBeans.get(cacheKey))) {
        return bean;
    }
    //如果Bean为AOP类的类型,或是需要跳过的类型,不进行处理
    if (isInfrastructureClass(bean.getClass()) || shouldSkip(bean.getClass(), beanName)) {
        //标记为不代理
        this.advisedBeans.put(cacheKey, Boolean.FALSE);
        return bean;
    }

    // Create proxy if we have advice.
    //寻找符合此Bean的增强方法(通知方法)
    Object[] specificInterceptors = getAdvicesAndAdvisorsForBean(bean.getClass(), beanName, null);
    //如果寻找到的增强方法列表不为空,也就是不为DO_NOT_PROXY
    if (specificInterceptors != DO_NOT_PROXY) {
        //标记为已代理
        this.advisedBeans.put(cacheKey, Boolean.TRUE);
        //根据找到的增强方法,对此Bean进行动态代理
        Object proxy = createProxy(
            bean.getClass(), beanName, specificInterceptors, new SingletonTargetSource(bean));
        this.proxyTypes.put(cacheKey, proxy.getClass());
        //将代理对象作为Bean返回给IOC容器
        return proxy;
    }
	//如果走到这里,说明代理失败,标记为代理失败
    this.advisedBeans.put(cacheKey, Boolean.FALSE);
    return bean;
}

文章到了这里,就已经基本完成AOP的实现了,剩下我们需要关注的就是两件事:

  1. 如何寻找符合Bean的增强器
  2. 如何对Bean创建动态代理

4.2 寻找所有的增强器

@Override
@Nullable
protected Object[] getAdvicesAndAdvisorsForBean(Class<?> beanClass, String beanName, @Nullable TargetSource targetSource) {
    //寻找适合的Advisor
    List<Advisor> advisors = findEligibleAdvisors(beanClass, beanName);
    if (advisors.isEmpty()) {
        return DO_NOT_PROXY;
    }
    return advisors.toArray();
}

这里有一个Advisor的概念,其中Advisor封装了切点信息与advise通知方法等等信息

protected List<Advisor> findEligibleAdvisors(Class<?> beanClass, String beanName) {
    //寻找所有适用的Advisor
    List<Advisor> candidateAdvisors = findCandidateAdvisors();
    //从所有Advisor中选出适合被当前Bean使用的Advisor
    List<Advisor> eligibleAdvisors = findAdvisorsThatCanApply(candidateAdvisors, beanClass, beanName);
    extendAdvisors(eligibleAdvisors);
    if (!eligibleAdvisors.isEmpty()) {
        eligibleAdvisors = sortAdvisors(eligibleAdvisors);
    }
    return eligibleAdvisors;
}

4.2.1 寻找已存在的Advisor

首先,执行下面的方法寻找合适的Advisor(此方法在子类Creator中得到实现):

@Override
protected List<Advisor> findCandidateAdvisors() {
    // Add all the Spring advisors found according to superclass rules.
    //首先调用父类findCandidateAdvisors的方法寻找在IOC容器中的Advisor类型的Bean
    List<Advisor> advisors = super.findCandidateAdvisors();
    // Build Advisors for all AspectJ aspects in the bean factory.
    if (this.aspectJAdvisorsBuilder != null) {
        //寻找注解的Advisor
        advisors.addAll(this.aspectJAdvisorsBuilder.buildAspectJAdvisors());
    }
    return advisors;
}

先调用父类的findCandidateAdvisors方法寻找所有的Advisor

protected List<Advisor> findCandidateAdvisors() {
    Assert.state(this.advisorRetrievalHelper != null, "No BeanFactoryAdvisorRetrievalHelper available");
    //委派Helper类去寻找
    return this.advisorRetrievalHelper.findAdvisorBeans();
}
public List<Advisor> findAdvisorBeans() {
    // Determine list of advisor bean names, if not cached already.
    String[] advisorNames = null;
    synchronized (this) {
        //先从缓存中取
        advisorNames = this.cachedAdvisorBeanNames;
        //缓存中若没有再去IOC容器中取
        if (advisorNames == null) {
            // Do not initialize FactoryBeans here: We need to leave all regular beans
            // uninitialized to let the auto-proxy creator apply to them!
            //从IOC容器中寻找所有Advisor类型的BeanName
            advisorNames = BeanFactoryUtils.beanNamesForTypeIncludingAncestors(
                this.beanFactory, Advisor.class, true, false);
            //放入缓存
            this.cachedAdvisorBeanNames = advisorNames;
        }
    }
    //没有找到,返回空
    if (advisorNames.length == 0) {
        return new LinkedList<>();
    }

    List<Advisor> advisors = new LinkedList<>();
    //遍历上面得到的所有BeanName
    for (String name : advisorNames) {
        if (isEligibleBean(name)) {
            if (this.beanFactory.isCurrentlyInCreation(name)) {
                if (logger.isDebugEnabled()) {
                    logger.debug("Skipping currently created advisor '" + name + "'");
                }
            }
            else {
                try {
                    //根据BeanName从IOC获取Bean实例并存入List中
                    advisors.add(this.beanFactory.getBean(name, Advisor.class));
                }
                //catch方法 略...
            }
        }
    }
    return advisors;
}

注意此时仅仅是只获取IOC容器中那些Advisor类型的Bean作为Advisor,而在我们注解配置AOP的方式中并不是配置Advisor类的,下面会说到如何获取所有打了注解的切面,这里只是尝试去IOC容器中找是否存在这样的类,有的话也不会漏掉。

4.2.2 寻找所有注解的Advisor

现在我们拿到了部分的Advisor,回到我们主类AnnotationAwareAspectJAutoProxyCreatorSfindCandidateAdvisors 方法,此时将委派BeanFactoryAspectJAdvisorsBuilder类去执行buildAspectJAdvisors 方法,将继续获取被注解了的Advisor:

@Override
protected List<Advisor> findCandidateAdvisors() {
    // Add all the Spring advisors found according to superclass rules.
    List<Advisor> advisors = super.findCandidateAdvisors();
    // Build Advisors for all AspectJ aspects in the bean factory.
    if (this.aspectJAdvisorsBuilder != null) {
        //委派BeanFactoryAspectJAdvisorsBuilder去寻找注解Advisor
        advisors.addAll(this.aspectJAdvisorsBuilder.buildAspectJAdvisors());
    }
    return advisors;
}
public List<Advisor> buildAspectJAdvisors() {
    List<String> aspectNames = this.aspectBeanNames;

    if (aspectNames == null) {
        synchronized (this) {
            aspectNames = this.aspectBeanNames;
            //双重加锁保证在并发情况下不会寻找两次
            if (aspectNames == null) {
                List<Advisor> advisors = new LinkedList<>();
                aspectNames = new LinkedList<>();
                //从IOC容器中获取所有的BeanName
                String[] beanNames = BeanFactoryUtils.beanNamesForTypeIncludingAncestors(
                    this.beanFactory, Object.class, true, false);
                //遍历所有的BeanName
                for (String beanName : beanNames) {
                    if (!isEligibleBean(beanName)) {
                        continue;
                    }
                    // We must be careful not to instantiate beans eagerly as in this case they
                    // would be cached by the Spring container but would not have been weaved.
                    Class<?> beanType = this.beanFactory.getType(beanName);
                    if (beanType == null) {
                        continue;
                    }
                    //判断此时的Bean的类上是否打了@Aspect注解
                    if (this.advisorFactory.isAspect(beanType)) {
                        //如果是,判断此类将是一个Advisor
                        aspectNames.add(beanName);
                        AspectMetadata amd = new AspectMetadata(beanType, beanName);
                        //如果是单例,说明可以缓存下来
                        if (amd.getAjType().getPerClause().getKind() == PerClauseKind.SINGLETON) {
                            //封装成一个对象
                            MetadataAwareAspectInstanceFactory factory =
                                new BeanFactoryAspectInstanceFactory(this.beanFactory, beanName);
                            //委派advisorFactory工厂针对以上封装信息创建Advisor
                            List<Advisor> classAdvisors = this.advisorFactory.getAdvisors(factory);
                            //如果是单例,缓存
                            if (this.beanFactory.isSingleton(beanName)) {
                                this.advisorsCache.put(beanName, classAdvisors);
                            }
                            //如果不是,只缓存factory,待下一次进入取出缓存的factory
                            //然后再用advisorFactory创建一次Advisor,省去寻找Bean与创建factory的麻烦
                            else {
                                this.aspectFactoryCache.put(beanName, factory);
                            }
                            advisors.addAll(classAdvisors);
                        }
                        else {
                            // Per target or per this.
                            if (this.beanFactory.isSingleton(beanName)) {
                                throw new IllegalArgumentException("Bean with name '" + beanName +
                                                                   "' is a singleton, but aspect instantiation model is not singleton");
                            }
                            MetadataAwareAspectInstanceFactory factory =
                                new PrototypeAspectInstanceFactory(this.beanFactory, beanName);
                            this.aspectFactoryCache.put(beanName, factory);
                            advisors.addAll(this.advisorFactory.getAdvisors(factory));
                        }
                    }
                }
                this.aspectBeanNames = aspectNames;
                return advisors;
            }
        }
    }

    //如果走到这里,说明已经找过一遍了,这里从缓存获取信息
    if (aspectNames.isEmpty()) {
        return Collections.emptyList();
    }
    List<Advisor> advisors = new LinkedList<>();
    //遍历所有缓存的切面名
    for (String aspectName : aspectNames) {
        //根据切面名从缓存拿Advisor
        List<Advisor> cachedAdvisors = this.advisorsCache.get(aspectName);
        //如果可以拿到,直接获取
        if (cachedAdvisors != null) {
            advisors.addAll(cachedAdvisors);
        }
        //如果拿不到,说明此时缓存的是factory
        //根据factory使用advisorFactory创建Advisor
        else {
            MetadataAwareAspectInstanceFactory factory = this.aspectFactoryCache.get(aspectName);
            advisors.addAll(this.advisorFactory.getAdvisors(factory));
        }
    }
    return advisors;
}

到这里,我们已经完成了对所有Advisor的获取,这里值得关注的是工厂(advisorFactory)是如何创建Advisor的呢?进入ReflectiveAspectJAdvisorFactory类的getAdvisors 方法:

@Override
public List<Advisor> getAdvisors(MetadataAwareAspectInstanceFactory aspectInstanceFactory) {
    //获取切面类类型
    Class<?> aspectClass = aspectInstanceFactory.getAspectMetadata().getAspectClass();
    //获取切面Name
    String aspectName = aspectInstanceFactory.getAspectMetadata().getAspectName();
    //验证
    validate(aspectClass);

    // We need to wrap the MetadataAwareAspectInstanceFactory with a decorator
    // so that it will only instantiate once.
    MetadataAwareAspectInstanceFactory lazySingletonAspectInstanceFactory =
        new LazySingletonAspectInstanceFactoryDecorator(aspectInstanceFactory);

    List<Advisor> advisors = new LinkedList<>();
    //遍历所有除了被打上@Pointcut注解的方法
    for (Method method : getAdvisorMethods(aspectClass)) {
        //在其每个方法中获取Advisor
        Advisor advisor = getAdvisor(method, lazySingletonAspectInstanceFactory, advisors.size(), aspectName);
        if (advisor != null) {
            advisors.add(advisor);
        }
    }

    // If it's a per target aspect, emit the dummy instantiating aspect.
    if (!advisors.isEmpty() && lazySingletonAspectInstanceFactory.getAspectMetadata().isLazilyInstantiated()) {
        Advisor instantiationAdvisor = new SyntheticInstantiationAdvisor(lazySingletonAspectInstanceFactory);
        advisors.add(0, instantiationAdvisor);
    }

    // Find introduction fields.
    //获取DeclareParents注解的filed
    for (Field field : aspectClass.getDeclaredFields()) {
        Advisor advisor = getDeclareParentsAdvisor(field);
        if (advisor != null) {
            advisors.add(advisor);
        }
    }

    return advisors;
}

我们需要关注两点:

  1. getAdvisorMethods:此方法获取类上所有的AdvisorMethods,那么是如何进行的:

    private List<Method> getAdvisorMethods(Class<?> aspectClass) {
        final List<Method> methods = new LinkedList<>();
        ReflectionUtils.doWithMethods(aspectClass, method -> {
            // Exclude pointcuts
            //获取方法上的注解,如果是Pointcut注解不处理
            if (AnnotationUtils.getAnnotation(method, Pointcut.class) == null) {
                methods.add(method);
            }
        });
        Collections.sort(methods, METHOD_COMPARATOR);
        return methods;
    }
    

    我们可以知道,其将不是Pointcut的注解的方法加入到方法集合中作为AdvisorMethod

  2. getAdvisor:从上面获得的方法中,提取出Advisor

    @Override
    @Nullable
    public Advisor getAdvisor(Method candidateAdviceMethod, MetadataAwareAspectInstanceFactory aspectInstanceFactory,
                              int declarationOrderInAspect, String aspectName) {
    
        validate(aspectInstanceFactory.getAspectMetadata().getAspectClass());
    
        //切点信息的获取
        AspectJExpressionPointcut expressionPointcut = getPointcut(
            candidateAdviceMethod, aspectInstanceFactory.getAspectMetadata().getAspectClass());
        if (expressionPointcut == null) {
            return null;
        }
    
        //根据获得的切点信息封装增强器
        return new InstantiationModelAwarePointcutAdvisorImpl(expressionPointcut, candidateAdviceMethod,
                                                              this, aspectInstanceFactory, declarationOrderInAspect, aspectName);
    }
    

    我们依然需要关注两个点:

    1. getPointcut:切点信息的获取
    2. InstantiationModelAwarePointcutAdvisorImpl:根据切点信息封装成增强器

4.2.3 获取切点信息

依然是在创建Advisor的工厂中的方法getPointcut

@Nullable
private AspectJExpressionPointcut getPointcut(Method candidateAdviceMethod, Class<?> candidateAspectClass) {
    //获取该方法的注解
    AspectJAnnotation<?> aspectJAnnotation =
        AbstractAspectJAdvisorFactory.findAspectJAnnotationOnMethod(candidateAdviceMethod);
    if (aspectJAnnotation == null) {
        return null;
    }

    //封装信息
    AspectJExpressionPointcut ajexp =
        new AspectJExpressionPointcut(candidateAspectClass, new String[0], new Class<?>[0]);
    //设置切点表达式
    //例如:@Before("test()")上的test()
    ajexp.setExpression(aspectJAnnotation.getPointcutExpression());
    if (this.beanFactory != null) {
        ajexp.setBeanFactory(this.beanFactory);
    }
    return ajexp;
}

需要关注的是获取方法的注解findAspectJAnnotationOnMethod

@SuppressWarnings("unchecked")
@Nullable
protected static AspectJAnnotation<?> findAspectJAnnotationOnMethod(Method method) {
    //需要关注的类型
    Class<?>[] classesToLookFor = new Class<?>[] {
        Before.class, Around.class, After.class, AfterReturning.class, AfterThrowing.class, Pointcut.class};
    for (Class<?> c : classesToLookFor) {
        //将每个需要关注的类型都与方法进行匹配
        AspectJAnnotation<?> foundAnnotation = findAnnotation(method, (Class<Annotation>) c);
        //如果找到,返回
        if (foundAnnotation != null) {
            return foundAnnotation;
        }
    }
    return null;
}

findAnnotation

@Nullable
private static <A extends Annotation> AspectJAnnotation<A> findAnnotation(Method method, Class<A> toLookFor) {
    //根据指定的类型寻找方法上的注解
    A result = AnnotationUtils.findAnnotation(method, toLookFor);
    if (result != null) {
        //如果找到了,将其封装为AspectJAnnotation对象返回
        return new AspectJAnnotation<>(result);
    }
    else {
        return null;
    }
}

注意,在寻找注解的时候仅仅是获取了例如@Before(“test()”)中的test()这样的切点方法名信息,这个过程在封装AspectJAnnotation对象时在构造函数完成

public AspectJAnnotation(A annotation) {
    this.annotation = annotation;
    this.annotationType = determineAnnotationType(annotation);
    // We know these methods exist with the same name on each object,
    // but need to invoke them reflectively as there isn't a common interface.
    try {
        //处理注解上的pointcut信息
        this.pointcutExpression = resolveExpression(annotation);
        this.argumentNames = (String) annotation.getClass().getMethod("argNames").invoke(annotation);
    }
    catch (Exception ex) {
        throw new IllegalArgumentException(annotation + " cannot be an AspectJ annotation", ex);
    }
}
private String resolveExpression(A annotation) throws Exception {
    //遍历获取注解中的两个方法名:value和pointcut
    for (String methodName : EXPRESSION_PROPERTIES) {
        Method method;
        try {
            //获取method对象
            method = annotation.getClass().getDeclaredMethod(methodName);
        }
        catch (NoSuchMethodException ex) {
            method = null;
        }
        if (method != null) {
            //获取切点方法名,也就是注解上设置的,例如@Before("test()")
            //此时获取"test()"这样的字符串
            String candidate = (String) method.invoke(annotation);
            if (StringUtils.hasText(candidate)) {
                return candidate;
            }
        }
    }
    throw new IllegalStateException("Failed to resolve expression: " + annotation);
}

4.2.4 根据切点信息,封装成Advisor

根据切点信息生成增强器,所有的增强都由InstantiationModelAwarePointcutAdvisorImpl 这个类来封装,当然,这个类是一个Advisor。进入此类的构造函数:

public InstantiationModelAwarePointcutAdvisorImpl(AspectJExpressionPointcut declaredPointcut,
                                                  Method aspectJAdviceMethod, AspectJAdvisorFactory aspectJAdvisorFactory,
                                                  MetadataAwareAspectInstanceFactory aspectInstanceFactory, int declarationOrder, String aspectName) {

    //将获得的信息都封装到这个对象的属性上
    this.declaredPointcut = declaredPointcut;
    this.declaringClass = aspectJAdviceMethod.getDeclaringClass();
    this.methodName = aspectJAdviceMethod.getName();
    this.parameterTypes = aspectJAdviceMethod.getParameterTypes();
    this.aspectJAdviceMethod = aspectJAdviceMethod;
    this.aspectJAdvisorFactory = aspectJAdvisorFactory;
    this.aspectInstanceFactory = aspectInstanceFactory;
    this.declarationOrder = declarationOrder;
    this.aspectName = aspectName;

    if (aspectInstanceFactory.getAspectMetadata().isLazilyInstantiated()) {
        // Static part of the pointcut is a lazy type.
        Pointcut preInstantiationPointcut = Pointcuts.union(
            aspectInstanceFactory.getAspectMetadata().getPerClausePointcut(), this.declaredPointcut);

        // Make it dynamic: must mutate from pre-instantiation to post-instantiation state.
        // If it's not a dynamic pointcut, it may be optimized out
        // by the Spring AOP infrastructure after the first evaluation.
        this.pointcut = new PerTargetInstantiationModelPointcut(
            this.declaredPointcut, preInstantiationPointcut, aspectInstanceFactory);
        this.lazy = true;
    }
    else {
        // A singleton aspect.
        this.pointcut = this.declaredPointcut;
        this.lazy = false;
        //解析当前切点适用的Advice,并保存到instantiatedAdvice属性上
        this.instantiatedAdvice = instantiateAdvice(this.declaredPointcut);
    }
}

在封装的过程中只是简单的将信息放入类的属性中,而值得关注的是最后一行的instantiateAdvice 方法,此方法将根据注解类型选择不同的Advise,例如@Before、@After等等都是不同的Advice,它们需要前置或是后置通知,所体现的增强的逻辑是不同的,所以就需要不同的Advice来完成:

private Advice instantiateAdvice(AspectJExpressionPointcut pointcut) {
    //委派别的类去获取Advice
    Advice advice = this.aspectJAdvisorFactory.getAdvice(this.aspectJAdviceMethod, pointcut,
                                                         this.aspectInstanceFactory, this.declarationOrder, this.aspectName);
    return (advice != null ? advice : EMPTY_ADVICE);
}

此类又委派了aspectJAdvisorFactory去获取Advice:

@Override
@Nullable
public Advice getAdvice(Method candidateAdviceMethod, AspectJExpressionPointcut expressionPointcut,
                        MetadataAwareAspectInstanceFactory aspectInstanceFactory, int declarationOrder, String aspectName) {

    //根据之前的信息获取切面类类型
    Class<?> candidateAspectClass = aspectInstanceFactory.getAspectMetadata().getAspectClass();
    validate(candidateAspectClass);

    //寻找方法上的注解
    AspectJAnnotation<?> aspectJAnnotation =
        AbstractAspectJAdvisorFactory.findAspectJAnnotationOnMethod(candidateAdviceMethod);
    if (aspectJAnnotation == null) {
        return null;
    }

    // If we get here, we know we have an AspectJ method.
    // Check that it's an AspectJ-annotated class
    if (!isAspect(candidateAspectClass)) {
        throw new AopConfigException("Advice must be declared inside an aspect type: " +
                                     "Offending method '" + candidateAdviceMethod + "' in class [" +
                                     candidateAspectClass.getName() + "]");
    }

    if (logger.isDebugEnabled()) {
        logger.debug("Found AspectJ method: " + candidateAdviceMethod);
    }

    AbstractAspectJAdvice springAdvice;

    //根据刚刚从方法上获得的注解,解析成枚举的类型进行swich
    switch (aspectJAnnotation.getAnnotationType()) {
            //@Before注解将会走这里
        case AtBefore:
            springAdvice = new AspectJMethodBeforeAdvice(
                candidateAdviceMethod, expressionPointcut, aspectInstanceFactory);
            break;
            //@After注解
        case AtAfter:
            springAdvice = new AspectJAfterAdvice(
                candidateAdviceMethod, expressionPointcut, aspectInstanceFactory);
            break;
            //@AfterReturning
        case AtAfterReturning:
            springAdvice = new AspectJAfterReturningAdvice(
                candidateAdviceMethod, expressionPointcut, aspectInstanceFactory);
            AfterReturning afterReturningAnnotation = (AfterReturning) aspectJAnnotation.getAnnotation();
            if (StringUtils.hasText(afterReturningAnnotation.returning())) {
                springAdvice.setReturningName(afterReturningAnnotation.returning());
            }
            break;
            //@AfterThrowing
        case AtAfterThrowing:
            springAdvice = new AspectJAfterThrowingAdvice(
                candidateAdviceMethod, expressionPointcut, aspectInstanceFactory);
            AfterThrowing afterThrowingAnnotation = (AfterThrowing) aspectJAnnotation.getAnnotation();
            if (StringUtils.hasText(afterThrowingAnnotation.throwing())) {
                springAdvice.setThrowingName(afterThrowingAnnotation.throwing());
            }
            break;
            //@Around
        case AtAround:
            springAdvice = new AspectJAroundAdvice(
                candidateAdviceMethod, expressionPointcut, aspectInstanceFactory);
            break;
            //@Pointcut
        case AtPointcut:
            if (logger.isDebugEnabled()) {
                logger.debug("Processing pointcut '" + candidateAdviceMethod.getName() + "'");
            }
            return null;
        default:
            throw new UnsupportedOperationException(
                "Unsupported advice type on method: " + candidateAdviceMethod);
    }

    // Now to configure the advice...
    //配置得到的Advice
    springAdvice.setAspectName(aspectName);
    springAdvice.setDeclarationOrder(declarationOrder);
    String[] argNames = this.parameterNameDiscoverer.getParameterNames(candidateAdviceMethod);
    if (argNames != null) {
        springAdvice.setArgumentNamesFromStringArray(argNames);
    }
    springAdvice.calculateArgumentBindings();
    return springAdvice;
}

从上面我们可以知道,不同的通知对应不同的Advice,那么这些不同的advice的执行逻辑有哪些不同呢?这个疑问留着后面的执行过程会进行解答,这里只需要知道我们获取的Advisor封装了advice与其切点信息

4.3 寻找适用的增强器

我们的思路回到主线,此时我们已经寻找到所有的增强器了,现在开始从中筛选出适用的增强器:

protected List<Advisor> findEligibleAdvisors(Class<?> beanClass, String beanName) {
    //寻找所有适用的Advisor
    //这一步在4.2中进行
    List<Advisor> candidateAdvisors = findCandidateAdvisors();
    //从所有Advisor中选出适合被当前Bean使用的Advisor
    List<Advisor> eligibleAdvisors = findAdvisorsThatCanApply(candidateAdvisors, beanClass, beanName);
    extendAdvisors(eligibleAdvisors);
    if (!eligibleAdvisors.isEmpty()) {
        eligibleAdvisors = sortAdvisors(eligibleAdvisors);
    }
    return eligibleAdvisors;
}

此时执行findAdvisorsThatCanApply方法寻找适用的增强器:

protected List<Advisor> findAdvisorsThatCanApply(
    List<Advisor> candidateAdvisors, Class<?> beanClass, String beanName) {

    ProxyCreationContext.setCurrentProxiedBeanName(beanName);
    try {
        //委派AopUtils去做
        return AopUtils.findAdvisorsThatCanApply(candidateAdvisors, beanClass);
    }
    finally {
        ProxyCreationContext.setCurrentProxiedBeanName(null);
    }
}

这里委派了AopUtils去做事情:

public static List<Advisor> findAdvisorsThatCanApply(List<Advisor> candidateAdvisors, Class<?> clazz) {
    if (candidateAdvisors.isEmpty()) {
        return candidateAdvisors;
    }
    List<Advisor> eligibleAdvisors = new LinkedList<>();
    //遍历之前找到的所有的Advisor
    for (Advisor candidate : candidateAdvisors) {
        //在这里会优先处理引介增强,并且判断Advisor是否适合此Bean
        if (candidate instanceof IntroductionAdvisor && canApply(candidate, clazz)) {
            //优先加入List中
            eligibleAdvisors.add(candidate);
        }
    }
    boolean hasIntroductions = !eligibleAdvisors.isEmpty();
    for (Advisor candidate : candidateAdvisors) {
        if (candidate instanceof IntroductionAdvisor) {
            // already processed
            //已经处理过引介增强
            continue;
        }
        //剩下的Advisor再进行判断
        if (canApply(candidate, clazz, hasIntroductions)) {
            //加入List中
            eligibleAdvisors.add(candidate);
        }
    }
    //返回适合的Advisor List
    return eligibleAdvisors;
}

我们重点关注canApply方法,看看是如何判断是否适合的:

public static boolean canApply(Advisor advisor, Class<?> targetClass, boolean hasIntroductions) {
    if (advisor instanceof IntroductionAdvisor) {
        //调用引介增强器的ClassFilter去匹配当前Bean是否适合
        return ((IntroductionAdvisor) advisor).getClassFilter().matches(targetClass);
    }
    else if (advisor instanceof PointcutAdvisor) {
        //如果是普通的Advisor,转换成PointcutAdvisor
        //PointcutAdvisor接口定义了获取切点的方法getPointcut
        PointcutAdvisor pca = (PointcutAdvisor) advisor;
        //根据切点信息,寻找是否适合
        return canApply(pca.getPointcut(), targetClass, hasIntroductions);
    }
    else {
        // It doesn't have a pointcut so we assume it applies.
        //如果没有切点,就假设它适合
        return true;
    }
}

这里我们先忽略引介增强器,关注普通的Advisor。这里进入canApply方法,根据切点信息去判断是否合适:

public static boolean canApply(Pointcut pc, Class<?> targetClass, boolean hasIntroductions) {
    Assert.notNull(pc, "Pointcut must not be null");
    //获取切点类过滤器,如果不符合直接返回false
    //如果符合进入下一步判断
    if (!pc.getClassFilter().matches(targetClass)) {
        return false;
    }

    //获取切点的方法匹配器
    MethodMatcher methodMatcher = pc.getMethodMatcher();
    //如果此时methodMatcher是一个MethodMatcher.TRUE,说明匹配任何方法,直接返回true
    if (methodMatcher == MethodMatcher.TRUE) {
        // No need to iterate the methods if we're matching any method anyway...
        return true;
    }

    IntroductionAwareMethodMatcher introductionAwareMethodMatcher = null;
    if (methodMatcher instanceof IntroductionAwareMethodMatcher) {
        introductionAwareMethodMatcher = (IntroductionAwareMethodMatcher) methodMatcher;
    }

    //拿到目标类的所有父类(包括接口)
    Set<Class<?>> classes = new LinkedHashSet<>(ClassUtils.getAllInterfacesForClassAsSet(targetClass));
    //包括本类
    classes.add(targetClass);
    //遍历父类与本类
    for (Class<?> clazz : classes) {
        //获取本class的方法对象
        Method[] methods = ReflectionUtils.getAllDeclaredMethods(clazz);
        //遍历方法
        for (Method method : methods) {
            //两个验证器只要其中一个匹配,就返回true
            if ((introductionAwareMethodMatcher != null &&
                 introductionAwareMethodMatcher.matches(method, targetClass, hasIntroductions)) ||
                methodMatcher.matches(method, targetClass)) {
                return true;
            }
        }
    }

    return false;
}

到这里,就完成了适用的增强器的查找,从上面我们可以知道,当一个Bean中的任何一个方法(方法包括在父类中的方法)匹配Advisor中的切点信息,就认为当前的Advisor是匹配当前Bean的,就将此时的Advisor加入此类的Advisors列表中,注意,一个Bean中是可以有多个Advisor的,如果不能理解为什么一个Bean对应多个Advisor,你应该还没有明白Advisor的概念,Advisor中包含advice与切点信息,也就是说,一个通知方法例如前置通知@Before是对应一个Advisor的,如果一个类中既有前置通知又有后置通知,那么这个类中的Advisor是会匹配两个的。

4.4 创建代理

4.4.1 获取代理类型(JDK或CGLIB)

再次回到最开始的地方,还记得4.1的wrapIfNecessary方法吗?此时我们已经寻找完符合此Bean的增强器了(getAdvicesAndAdvisorsForBean方法),就像刚刚所说,如果有两个通知匹配的话,现在增强器列表应该会有两个Advisor,也就是说增强列表不为空,将进行下一步,创建代理的过程createProxy

protected Object createProxy(Class<?> beanClass, @Nullable String beanName,
                             @Nullable Object[] specificInterceptors, TargetSource targetSource) {

    if (this.beanFactory instanceof ConfigurableListableBeanFactory) {
        AutoProxyUtils.exposeTargetClass((ConfigurableListableBeanFactory) this.beanFactory, beanName, beanClass);
    }

    ProxyFactory proxyFactory = new ProxyFactory();
    //复制当前Creator的一些属性例如proxyTargetClass或是exposeProxy等
    proxyFactory.copyFrom(this);

    //如果proxyTargetClass是false才会进入判断
    //如果proxyTargetClass是true就直接用CGLib代理了,不需要判断
    if (!proxyFactory.isProxyTargetClass()) {
        //判断是否需要代理本类
        if (shouldProxyTargetClass(beanClass, beanName)) {
            //也就是说,如果需要代理本类,将使用CGLib方式代理
            proxyFactory.setProxyTargetClass(true);
        }
        //如果不需要,则获取所有该类的接口,设置到proxyFactory对象中
        else {
            evaluateProxyInterfaces(beanClass, proxyFactory);
        }
    }

    //处理所有得到的适用的拦截器转换成Advisor
    Advisor[] advisors = buildAdvisors(beanName, specificInterceptors);
    //将Advisors加入proxyFactory对象中
    proxyFactory.addAdvisors(advisors);
    //将原本的Bean对象(原生未代理)加入proxyFactory对象中
    proxyFactory.setTargetSource(targetSource);
    //定制代理
    customizeProxyFactory(proxyFactory);

    proxyFactory.setFrozen(this.freezeProxy);
    if (advisorsPreFiltered()) {
        proxyFactory.setPreFiltered(true);
    }

    //设置好属性之后,就调用proxyFactory对象生成代理对象
    return proxyFactory.getProxy(getProxyClassLoader());
}

以上过程都在给proxyFactory这个对象设置属性,然后调用proxyFactory根据设置的一些属性生成代理对象,最后执行关键的创建代理方法getProxy

public Object getProxy(@Nullable ClassLoader classLoader) {
    return createAopProxy().getProxy(classLoader);
}
protected final synchronized AopProxy createAopProxy() {
    if (!this.active) {
        activate();
    }
    //使用aopProxyFactory根据本类信息创建
    return getAopProxyFactory().createAopProxy(this);
}

其中aopProxyFactory是在其父类的构造函数中赋值的:

public ProxyCreatorSupport() {
    this.aopProxyFactory = new DefaultAopProxyFactory();
}

进入其创建方法createAopProxy

@Override
public AopProxy createAopProxy(AdvisedSupport config) throws AopConfigException {
    //判断ProxyTargetClass属性和是否有代理接口等等判断使用何种方式做代理
    if (config.isOptimize() || config.isProxyTargetClass() || hasNoUserSuppliedProxyInterfaces(config)) {
        Class<?> targetClass = config.getTargetClass();
        if (targetClass == null) {
            throw new AopConfigException("TargetSource cannot determine target class: " +
                                         "Either an interface or a target is required for proxy creation.");
        }
        if (targetClass.isInterface() || Proxy.isProxyClass(targetClass)) {
            return new JdkDynamicAopProxy(config);
        }
        return new ObjenesisCglibAopProxy(config);
    }
    else {
        return new JdkDynamicAopProxy(config);
    }
}

注意此时config对象是我们之前设置了各种信息的proxyFactory对象,在代理的构造函数中存放了这个属性,也就是说代理对象持有一个Advisor列表。

这段代码清晰明了,主要就是在判断使用何种方式做代理

  • optimize:这个属性适用于CGLib,对JDK动态代理无效。主要是控制CGLib创建的代理是否使用激进的优化策略。
  • proxyTargetClass:文章开头也有提到,如果属性为true,将会使用CGLib进行代理。
  • hasNoUserSuppliedProxyInterfaces:是否存在代理接口。

总结一下JDK于CGLIB如何选择:

  • 如果目标对象实现了接口,默认情况下会采用JDK代理
  • 如果目标对象实现了接口,但可以使用proxyTargetClass属性来强制使用CGLIB代理
  • 如果对象没用实现接口,必须使用CGLIB代理

4.4.2 获取代理

思路回到开头获取代理的方法中:

public Object getProxy(@Nullable ClassLoader classLoader) {
    return createAopProxy().getProxy(classLoader);
}

此时我们完成了createAopProxy,拿到了JDK代理对象(我们这里假设用JDK做代理)接着调用其getProxy方法:

@Override
public Object getProxy(@Nullable ClassLoader classLoader) {
    if (logger.isDebugEnabled()) {
        logger.debug("Creating JDK dynamic proxy: target source is " + this.advised.getTargetSource());
    }
    //拿到被代理类的接口
    Class<?>[] proxiedInterfaces = AopProxyUtils.completeProxiedInterfaces(this.advised, true);
    //检查上面得到的接口是否定义了equals、hashCode方法
    findDefinedEqualsAndHashCodeMethods(proxiedInterfaces);
    //创建代理对象
    return Proxy.newProxyInstance(classLoader, proxiedInterfaces, this);
}

注意此时的advised就是我们上面所说的config对象也就是封装半天信息的proxyFactory对象,里面存有一系列信息例如Advisor、切点之类

这里不对动态代理进行解释,如果熟悉动态代理的看到这里应该已经懂了,其将当前对象作为实现代理的主要对象。毫无疑问,当前对象一定实现了InvocationHandler接口,我们需要关注的即为它的invoke方法:

@Override
@Nullable
public Object invoke(Object proxy, Method method, Object[] args) throws Throwable {
    MethodInvocation invocation;
    Object oldProxy = null;
    boolean setProxyContext = false;

    //原生对象
    TargetSource targetSource = this.advised.targetSource;
    Object target = null;

    try {
        //eqauls()方法,如果目标对象未实现此方法
        if (!this.equalsDefined && AopUtils.isEqualsMethod(method)) {
            // The target does not implement the equals(Object) method itself.
            return equals(args[0]);
        }
        //hashCode()方法,如果目标对象未实现此方法
        else if (!this.hashCodeDefined && AopUtils.isHashCodeMethod(method)) {
            // The target does not implement the hashCode() method itself.
            return hashCode();
        }
        else if (method.getDeclaringClass() == DecoratingProxy.class) {
            // There is only getDecoratedClass() declared -> dispatch to proxy config.
            return AopProxyUtils.ultimateTargetClass(this.advised);
        }
        //Advised接口或者其父接口中定义的方法,直接反射调用,不应用通知
        else if (!this.advised.opaque && method.getDeclaringClass().isInterface() &&
                 method.getDeclaringClass().isAssignableFrom(Advised.class)) {
            // Service invocations on ProxyConfig with the proxy config...
            return AopUtils.invokeJoinpointUsingReflection(this.advised, method, args);
        }

        Object retVal;

        //这里就是先前提到的如果exposeProxy属性为true,将会暴露一个proxy代理对象
        //给AOP上下文对象,存在线程变量中
        if (this.advised.exposeProxy) {
            // Make invocation available if necessary.
            oldProxy = AopContext.setCurrentProxy(proxy);
            setProxyContext = true;
        }

        // Get as late as possible to minimize the time we "own" the target,
        // in case it comes from a pool.
        //获得目标对象的类
        target = targetSource.getTarget();
        Class<?> targetClass = (target != null ? target.getClass() : null);

        // Get the interception chain for this method.
        //获取可以应用到此方法上的Interceptor列表
        List<Object> chain = this.advised.getInterceptorsAndDynamicInterceptionAdvice(method, targetClass);

        // Check whether we have any advice. If we don't, we can fallback on direct
        // reflective invocation of the target, and avoid creating a MethodInvocation.
        //如果没有可以应用到此方法的通知(Interceptor),此直接反射调用 method.invoke(target, args)
        if (chain.isEmpty()) {
            // We can skip creating a MethodInvocation: just invoke the target directly
            // Note that the final invoker must be an InvokerInterceptor so we know it does
            // nothing but a reflective operation on the target, and no hot swapping or fancy proxying.
            Object[] argsToUse = AopProxyUtils.adaptArgumentsIfNecessary(method, args);
            retVal = AopUtils.invokeJoinpointUsingReflection(target, method, argsToUse);
        }
        else {
            // We need to create a method invocation...
            //走到这里,说明该方法符合被通知的条件,创建MethodInvocation
            //执行其proceed方法
            invocation = new ReflectiveMethodInvocation(proxy, target, method, args, targetClass, chain);
            // Proceed to the joinpoint through the interceptor chain.
            retVal = invocation.proceed();
        }

        // Massage return value if necessary.
        Class<?> returnType = method.getReturnType();
        if (retVal != null && retVal == target &&
            returnType != Object.class && returnType.isInstance(proxy) &&
            !RawTargetAccess.class.isAssignableFrom(method.getDeclaringClass())) {
            // Special case: it returned "this" and the return type of the method
            // is type-compatible. Note that we can't help if the target sets
            // a reference to itself in another returned object.
            retVal = proxy;
        }
        else if (retVal == null && returnType != Void.TYPE && returnType.isPrimitive()) {
            throw new AopInvocationException(
                "Null return value from advice does not match primitive return type for: " + method);
        }
        return retVal;
    }
    finally {
        if (target != null && !targetSource.isStatic()) {
            // Must have come from TargetSource.
            targetSource.releaseTarget(target);
        }
        if (setProxyContext) {
            // Restore old proxy.
            AopContext.setCurrentProxy(oldProxy);
        }
    }
}

也就是说,每次调用代理对象的方法,将执行以下步骤:

  1. 判断是否是equals或hashCode或Advised的方法:如果是执行对应判断。
  2. 判断是否需要暴露代理对象(将代理对象存入AopContext):在文章开头就有提到这个属性,还是之前的例子,在A方法中调用B方法,此时B方法其实是原生对象的B方法,但如果B方法需要被AOP,需要执行的是代理对象的B方法,而不是原生对象的B方法,原生对象的B方法是没有被代理AOP的。所以这里需要暴露出代理对象,从AopContext上下文中取出代理对象,将其强转成对应类,执行B方法,此时的B方法即可是被AOP的方法了。值得一提的是AopContext暴露出来的代理对象是线程变量。
  3. 获取当前方法对应的拦截器链:主要是使用advised去对当前方法进行判断是否符合切点如果符合取出对应需要执行的链。如果此时链为空,代表执行的该方法是不需要被AOP的,正常反射执行即可,如果链不为空,将链封装成ReflectiveMethodInvocation对象执行其proceed方法,该对象的该方法是执行AOP的核心,下面详细讲解。

4.5 执行AOP通知

到此我们创建好了代理对象,并返回出去作为一个Bean存放在IOC容器中,当我们取出此Bean并执行方法时,如果碰到了需要被AOP的方法时,会找到一个拦截器链,然后将其传入ReflectiveMethodInvocation对象封装,接着不会执行原方法而是执行其proceed方法,所以该方法是执行AOP通知的核心方法:

//此属性默认为-1
private int currentInterceptorIndex = -1;

//存放interceptors的列表
protected final List<?> interceptorsAndDynamicMethodMatchers;

@Override
@Nullable
public Object proceed() throws Throwable {
    //	We start with an index of -1 and increment early.
    //如果Interceptor执行完了,则执行joinPoint
    if (this.currentInterceptorIndex == this.interceptorsAndDynamicMethodMatchers.size() - 1) {
        return invokeJoinpoint();
    }

    //根据游标获取对应的Advice
    Object interceptorOrInterceptionAdvice =
        this.interceptorsAndDynamicMethodMatchers.get(++this.currentInterceptorIndex);
    //如果要动态匹配joinPoint
    if (interceptorOrInterceptionAdvice instanceof InterceptorAndDynamicMethodMatcher) {
        // Evaluate dynamic method matcher here: static part will already have
        // been evaluated and found to match.
        InterceptorAndDynamicMethodMatcher dm =
            (InterceptorAndDynamicMethodMatcher) interceptorOrInterceptionAdvice;
        //动态匹配:运行时参数是否满足匹配条件
        if (dm.methodMatcher.matches(this.method, this.targetClass, this.arguments)) {
            return dm.interceptor.invoke(this);
        }
        else {
            // Dynamic matching failed.
            // Skip this interceptor and invoke the next in the chain.
            //动态匹配失败时,略过当前Intercetpor,调用下一个Interceptor
            return proceed();
        }
    }
    else {
        // It's an interceptor, so we just invoke it: The pointcut will have
        // been evaluated statically before this object was constructed.
        //执行当前IntercetporAdvice
        //注意此时传了this对象,是为了执行链的保持
        return ((MethodInterceptor) interceptorOrInterceptionAdvice).invoke(this);
    }
}

到这里,需要解释几点:

  1. 如何判断执行结束?在开头我们说到currentInterceptorIndex属性是为-1的,interceptorsAndDynamicMethodMatchers的size是执行链的数量,刚进来的时候执行链一定大于等于1,所以其减1也不会等于-1,不会执行invokeJoinpoint方法,走到下面执行这样一行代码

    Object interceptorOrInterceptionAdvice =
            this.interceptorsAndDynamicMethodMatchers.get(++this.currentInterceptorIndex);
    

    注意此时currentInterceptorIndex游标变成了0,也就是取执行链的第一个执行器取执行,同时游标变成0。下面会执行Advice的invoke方法,这里举例Before类型的Adviceinvoke方法:

    @Override
    public Object invoke(MethodInvocation mi) throws Throwable {
        this.advice.before(mi.getMethod(), mi.getArguments(), mi.getThis() );
        return mi.proceed();
    }
    

    注意此时mi对象就是上面传进来的this,也就是ReflectiveMethodInvocation对象,他在通知方法执行完成之后,又执行了ReflectiveMethodInvocationproceed方法,然后第二次进入proceed时,游标已经变成了0,我们假设此时拦截器数量为一个,那么此时currentInterceptorIndex=0,interceptorsAndDynamicMethodMatchers.size() - 1 = 0:

    if (this.currentInterceptorIndex == this.interceptorsAndDynamicMethodMatchers.size() - 1) {
        return invokeJoinpoint();
    }
    

    将中止执行下面的内容,直接执行invokeJoinpoint方法:

    @Nullable
    protected Object invokeJoinpoint() throws Throwable {
        return AopUtils.invokeJoinpointUsingReflection(this.target, this.method, this.arguments);
    }
    

    这里底层就是通过反射执行原方法,到此结束,也就是说其利用游标索引与循环调用proceed方法来判断当前执行链的结束,真是个聪明的方法。

  2. 为什么@Before的Advice的invoke可以实现前置通知效果,同样的@After如何实现后置通知效果?在这里我们将补上上面没有详细讲解的一些Advice类

    • MethodBeforeAdviceInterceptor前置通知:

      public class MethodBeforeAdviceInterceptor implements MethodInterceptor, Serializable {
      
          private MethodBeforeAdvice advice;
      
          /**
      	 * Create a new MethodBeforeAdviceInterceptor for the given advice.
      	 * @param advice the MethodBeforeAdvice to wrap
      	 */
          //构造器初始化保存了advice对象
          public MethodBeforeAdviceInterceptor(MethodBeforeAdvice advice) {
              Assert.notNull(advice, "Advice must not be null");
              this.advice = advice;
          }
      
          @Override
          public Object invoke(MethodInvocation mi) throws Throwable {
              this.advice.before(mi.getMethod(), mi.getArguments(), mi.getThis() );
              return mi.proceed();
          }
      }
      
    • AfterReturningAdviceInterceptor后置通知:

      public class AfterReturningAdviceInterceptor implements MethodInterceptor, AfterAdvice, Serializable {
      
          private final AfterReturningAdvice advice;
      
      
          /**
      	 * Create a new AfterReturningAdviceInterceptor for the given advice.
      	 * @param advice the AfterReturningAdvice to wrap
      	 */
          public AfterReturningAdviceInterceptor(AfterReturningAdvice advice) {
              Assert.notNull(advice, "Advice must not be null");
              this.advice = advice;
          }
      
          @Override
          public Object invoke(MethodInvocation mi) throws Throwable {
              Object retVal = mi.proceed();
              this.advice.afterReturning(retVal, mi.getMethod(), mi.getArguments(), mi.getThis());
              return retVal;
          }
      
      }
      

      从上面两个不同的Advice可以看出,其invoke方法的逻辑只是顺序不同而已,但为什么可以达到前置后置通知的呢?其实前置通知很好理解,在invoke的逻辑中只要先执行我们指定的前置通知逻辑,然后继续往下执行proceed即可实现在执行原方法前加一段逻辑的功能,我们主要分析后置通知是如何实现的。还是进行一个假设,此时该方法只有一个执行链后置通知@After,那么刚开始进入proceed方法时,获取这个Advice执行invoke方法,然后又去执行proceed方法了,由上面的分析此时的游标显示执行链已经执行完毕,所以此时会直接反射执行原方法,然后proceed方法执行完毕,最后将会执行后置通知Advice的invoke方法的下一步,advice.afterReturning,也就是用户自定义的后置方法逻辑,实现了在方法最后插入一段代码逻辑的功能,如果是多个通知例如前置后置都有的话,留给读者思考,相信懂得以上逻辑之后推敲起来并不是很难。

5. 总结

到这里为止,我们分析了AOP是从自定义标签开始,自定义标签注册了一个Creator类,而此Creator类是一个BeanPostProcessor,也就是说每个Bean都将调用Creator实现的方法。

来到了如何实现AOP代理的分析中,在此方法中判断Bean是否需要被代理,然后进一步判断此Bean是否可以获得增强器,在这个过程中会去拿所有的Advisor,然后在所有的Advisor列表中筛选出适用的增强器,如果适用的增强器Advisor为空证明此Bean不用被代理,如果不为空则会进入创建代理的流程。

所以我们开始分析创建代理是怎样的过程 ,在什么情况下会进行JDK动态代理,在什么情况下会进行CGLIB动态代理,然后创建代理结束之后。

必不可少的也需要分析是如何执行AOP通知 的,其核心类是ReflectiveMethodInvocationproceed方法,轮询执行拦截器,并且不同的Advice根据其特有的逻辑执行调用顺序,来完成AOP通知的功能。

猜你喜欢

转载自blog.csdn.net/qq_41737716/article/details/85130473