前言

AOP，也就是面向切面編程，它可以將公共的代碼抽離出來(lái)，動(dòng)態(tài)的織入到目標(biāo)類(lèi)、目標(biāo)方法中，大大提高我們編程的效率，也使程序變得更加優(yōu)雅。如事務(wù)、操作日志等都可以使用AOP實(shí)現(xiàn)。這種織入可以是在運(yùn)行期動(dòng)態(tài)生成代理對(duì)象實(shí)現(xiàn)，也可以在編譯期、類(lèi)加載時(shí)期靜態(tài)織入到代碼中。而Spring正是通過(guò)第一種方法實(shí)現(xiàn)，且在代理類(lèi)的生成上也有兩種方式：JDK Proxy和CGLIB，默認(rèn)當(dāng)類(lèi)實(shí)現(xiàn)了接口時(shí)使用前者，否則使用后者；另外Spring AOP只能實(shí)現(xiàn)對(duì)方法的增強(qiáng)。

正文

基本概念

AOP的術(shù)語(yǔ)很多，雖然不清楚術(shù)語(yǔ)我們也能很熟練地使用AOP，但是要理解分析源碼，術(shù)語(yǔ)就需要深刻體會(huì)其含義。

增強(qiáng)（Advice）：就是我們想要額外增加的功能

目標(biāo)對(duì)象（Target）：就是我們想要增強(qiáng)的目標(biāo)類(lèi)，如果沒(méi)有AOP，我們需要在每個(gè)目標(biāo)對(duì)象中實(shí)現(xiàn)日志、事務(wù)管理等非業(yè)務(wù)邏輯

連接點(diǎn)（JoinPoint）：程序執(zhí)行時(shí)的特定時(shí)機(jī)，如方法執(zhí)行前、后以及拋出異常后等等。

切點(diǎn)（Pointcut）：連接點(diǎn)的導(dǎo)航，我們?nèi)绾握业侥繕?biāo)對(duì)象呢？切點(diǎn)的作用就在于此，在Spring中就是匹配表達(dá)式。

引介（Introduction）：引介是一種特殊的增強(qiáng)，它為類(lèi)添加一些屬性和方法。這樣，即使一個(gè)業(yè)務(wù)類(lèi)原本沒(méi)有實(shí)現(xiàn)某個(gè)接口，通過(guò)AOP的引介功能，我們可以動(dòng)態(tài)地為該業(yè)務(wù)類(lèi)添加接口的實(shí)現(xiàn)邏輯，讓業(yè)務(wù)類(lèi)成為這個(gè)接口的實(shí)現(xiàn)類(lèi)。

織入（Weaving）：即如何將增強(qiáng)添加到目標(biāo)對(duì)象的連接點(diǎn)上，有動(dòng)態(tài)（運(yùn)行期生成代理）、靜態(tài)（編譯期、類(lèi)加載時(shí)期）兩種方式。

代理（Proxy）：目標(biāo)對(duì)象被織入增強(qiáng)后，就會(huì)產(chǎn)生一個(gè)代理對(duì)象，該對(duì)象可能是和原對(duì)象實(shí)現(xiàn)了同樣的一個(gè)接口（JDK），也可能是原對(duì)象的子類(lèi)（CGLIB）。

切面（Aspect、Advisor）：切面由切點(diǎn)和增強(qiáng)組成，包含了這兩者的定義。

代理對(duì)象的創(chuàng)建

在熟悉了AOP術(shù)語(yǔ)后，下面就來(lái)看看Spring是如何創(chuàng)建代理對(duì)象的，是否還記得上一篇提到的AOP的入口呢？在AbstractAutowireCapableBeanFactory類(lèi)的applyBeanPostProcessorsAfterInitialization方法中循環(huán)調(diào)用了BeanPostProcessor的postProcessAfterInitialization方法，其中一個(gè)就是我們創(chuàng)建代理對(duì)象的入口。

這里是Bean實(shí)例化完成去創(chuàng)建代理對(duì)象，理所當(dāng)然應(yīng)該這樣，但實(shí)際上在Bean實(shí)例化之前調(diào)用了一個(gè)resolveBeforeInstantiation方法，這里實(shí)際上我們也是有機(jī)會(huì)可以提前創(chuàng)建代理對(duì)象的，這里放到最后來(lái)分析，先來(lái)看主入口，進(jìn)入到AbstractAutoProxyCreator類(lèi)中：

public Object postProcessAfterInitialization(@Nullable Object bean, String beanName) { if (bean != null) { Object cacheKey = getCacheKey(bean.getClass(), beanName); if (!this.earlyProxyReferences.contains(cacheKey)) { return wrapIfNecessary(bean, beanName, cacheKey); } } return bean; } protected Object wrapIfNecessary(Object bean, String beanName, Object cacheKey) { //創(chuàng)建當(dāng)前bean的代理，如果這個(gè)bean有advice的話(huà)，重點(diǎn)看 // Create proxy if we have advice. Object[] specificInterceptors = getAdvicesAndAdvisorsForBean(bean.getClass(), beanName, null); //如果有切面，則生成該bean的代理 if (specificInterceptors != DO_NOT_PROXY) { this.advisedBeans.put(cacheKey, Boolean.TRUE); //把被代理對(duì)象bean實(shí)例封裝到SingletonTargetSource對(duì)象中 Object proxy = createProxy( bean.getClass(), beanName, specificInterceptors, new SingletonTargetSource(bean)); this.proxyTypes.put(cacheKey, proxy.getClass()); return proxy; } this.advisedBeans.put(cacheKey, Boolean.FALSE); return bean; }

先從緩存中拿，沒(méi)有則調(diào)用wrapIfNecessary方法創(chuàng)建。在這個(gè)方法里面主要看兩個(gè)地方：getAdvicesAndAdvisorsForBean和createProxy。簡(jiǎn)單一句話(huà)概括就是先掃描后創(chuàng)建，問(wèn)題是掃描什么呢？你可以先結(jié)合上面的概念思考下，換你會(huì)怎么做。進(jìn)入到子類(lèi)AbstractAdvisorAutoProxyCreator的getAdvicesAndAdvisorsForBean方法中：

protected Object[] getAdvicesAndAdvisorsForBean( Class<?> beanClass, String beanName, @Nullable TargetSource targetSource) { //找到合格的切面 List<Advisor> advisors = findEligibleAdvisors(beanClass, beanName); if (advisors.isEmpty()) { return DO_NOT_PROXY; } return advisors.toArray(); } protected List<Advisor> findEligibleAdvisors(Class<?> beanClass, String beanName) { //找到候選的切面,其實(shí)就是一個(gè)尋找有@Aspectj注解的過(guò)程，把工程中所有有這個(gè)注解的類(lèi)封裝成Advisor返回 List<Advisor> candidateAdvisors = findCandidateAdvisors(); //判斷候選的切面是否作用在當(dāng)前beanClass上面，就是一個(gè)匹配過(guò)程。現(xiàn)在就是一個(gè)匹配 List<Advisor> eligibleAdvisors = findAdvisorsThatCanApply(candidateAdvisors, beanClass, beanName); extendAdvisors(eligibleAdvisors); if (!eligibleAdvisors.isEmpty()) { //對(duì)有@Order@Priority進(jìn)行排序 eligibleAdvisors = sortAdvisors(eligibleAdvisors); } return eligibleAdvisors; }

在findEligibleAdvisors方法中可以看到有兩個(gè)步驟，第一先找到所有的切面，即掃描所有帶有@Aspect注解的類(lèi)，并將其中的切點(diǎn)（表達(dá)式）和增強(qiáng)封裝為切面，掃描完成后，自然是要判斷哪些切面能夠連接到當(dāng)前Bean實(shí)例上。下面一步步來(lái)分析，首先是掃描過(guò)程，進(jìn)入到AnnotationAwareAspectJAutoProxyCreator類(lèi)中：

protected List<Advisor> findCandidateAdvisors() { // 先通過(guò)父類(lèi)AbstractAdvisorAutoProxyCreator掃描，這里不重要 List<Advisor> advisors = super.findCandidateAdvisors(); // 主要看這里 if (this.aspectJAdvisorsBuilder != null) { advisors.addAll(this.aspectJAdvisorsBuilder.buildAspectJAdvisors()); } return advisors; }

這里委托給了BeanFactoryAspectJAdvisorsBuilderAdapter類(lèi)，并調(diào)用其父類(lèi)的buildAspectJAdvisors方法創(chuàng)建切面對(duì)象：

public List<Advisor> buildAspectJAdvisors() { List<String> aspectNames = this.aspectBeanNames; if (aspectNames == null) { synchronized (this) { aspectNames = this.aspectBeanNames; if (aspectNames == null) { List<Advisor> advisors = new ArrayList<>(); aspectNames = new ArrayList<>(); //獲取spring容器中的所有bean的名稱(chēng)BeanName String[] beanNames = BeanFactoryUtils.beanNamesForTypeIncludingAncestors( this.beanFactory, Object.class, true, false); for (String beanName : beanNames) { if (!isEligibleBean(beanName)) { continue; } Class<?> beanType = this.beanFactory.getType(beanName); if (beanType == null) { continue; } //判斷類(lèi)上是否有@Aspect注解 if (this.advisorFactory.isAspect(beanType)) { aspectNames.add(beanName); AspectMetadata amd = new AspectMetadata(beanType, beanName); if (amd.getAjType().getPerClause().getKind() == PerClauseKind.SINGLETON) { // 當(dāng)@Aspect的value屬性為''時(shí)才會(huì)進(jìn)入到這里 // 創(chuàng)建獲取有@Aspect注解類(lèi)的實(shí)例工廠(chǎng)，負(fù)責(zé)獲取有@Aspect注解類(lèi)的實(shí)例 MetadataAwareAspectInstanceFactory factory = new BeanFactoryAspectInstanceFactory(this.beanFactory, beanName); //創(chuàng)建切面advisor對(duì)象 List<Advisor> classAdvisors = this.advisorFactory.getAdvisors(factory); if (this.beanFactory.isSingleton(beanName)) { this.advisorsCache.put(beanName, classAdvisors); } else { this.aspectFactoryCache.put(beanName, factory); } advisors.addAll(classAdvisors); } else { MetadataAwareAspectInstanceFactory factory = new PrototypeAspectInstanceFactory(this.beanFactory, beanName); this.aspectFactoryCache.put(beanName, factory); advisors.addAll(this.advisorFactory.getAdvisors(factory)); } } } this.aspectBeanNames = aspectNames; return advisors; } } } return advisors; }

這個(gè)方法里面首先從IOC中拿到所有Bean的名稱(chēng)，并循環(huán)判斷該類(lèi)上是否帶有@Aspect注解，如果有則將BeanName和Bean的Class類(lèi)型封裝到BeanFactoryAspectInstanceFactory中，并調(diào)用ReflectiveAspectJAdvisorFactory.getAdvisors創(chuàng)建切面對(duì)象：

public List<Advisor> getAdvisors(MetadataAwareAspectInstanceFactory aspectInstanceFactory) { //從工廠(chǎng)中獲取有@Aspect注解的類(lèi)Class Class<?> aspectClass = aspectInstanceFactory.getAspectMetadata().getAspectClass(); //從工廠(chǎng)中獲取有@Aspect注解的類(lèi)的名稱(chēng) String aspectName = aspectInstanceFactory.getAspectMetadata().getAspectName(); validate(aspectClass); // 創(chuàng)建工廠(chǎng)的裝飾類(lèi)，獲取實(shí)例只會(huì)獲取一次 MetadataAwareAspectInstanceFactory lazySingletonAspectInstanceFactory = new LazySingletonAspectInstanceFactoryDecorator(aspectInstanceFactory); List<Advisor> advisors = new ArrayList<>(); //這里循環(huán)沒(méi)有@Pointcut注解的方法 for (Method method : getAdvisorMethods(aspectClass)) { //非常重要重點(diǎn)看看 Advisor advisor = getAdvisor(method, lazySingletonAspectInstanceFactory, advisors.size(), aspectName); if (advisor != null) { advisors.add(advisor); } } if (!advisors.isEmpty() && lazySingletonAspectInstanceFactory.getAspectMetadata().isLazilyInstantiated()) { Advisor instantiationAdvisor = new SyntheticInstantiationAdvisor(lazySingletonAspectInstanceFactory); advisors.add(0, instantiationAdvisor); } //判斷屬性上是否有引介注解，這里可以不看 for (Field field : aspectClass.getDeclaredFields()) { //判斷屬性上是否有DeclareParents注解，如果有返回切面 Advisor advisor = getDeclareParentsAdvisor(field); if (advisor != null) { advisors.add(advisor); } } return advisors; } private List<Method> getAdvisorMethods(Class<?> aspectClass) { final List<Method> methods = new ArrayList<>(); ReflectionUtils.doWithMethods(aspectClass, method -> { // Exclude pointcuts if (AnnotationUtils.getAnnotation(method, Pointcut.class) == null) { methods.add(method); } }); methods.sort(METHOD_COMPARATOR); return methods; }

根據(jù)Aspect的Class拿到所有不帶@Pointcut注解的方法對(duì)象（為什么是不帶@Pointcut注解的方法？仔細(xì)想想不難理解），另外要注意這里對(duì)method進(jìn)行了排序，看看這個(gè)METHOD_COMPARATOR比較器：

private static final Comparator<Method> METHOD_COMPARATOR; static { Comparator<Method> adviceKindComparator = new ConvertingComparator<>( new InstanceComparator<>( Around.class, Before.class, After.class, AfterReturning.class, AfterThrowing.class), (Converter<Method, Annotation>) method -> { AspectJAnnotation<?> annotation = AbstractAspectJAdvisorFactory.findAspectJAnnotationOnMethod(method); return (annotation != null ? annotation.getAnnotation() : null); }); Comparator<Method> methodNameComparator = new ConvertingComparator<>(Method::getName); METHOD_COMPARATOR = adviceKindComparator.thenComparing(methodNameComparator); }

關(guān)注InstanceComparator構(gòu)造函數(shù)參數(shù)，記住它們的順序，這就是AOP鏈?zhǔn)秸{(diào)用中同一個(gè)@Aspect類(lèi)中Advice的執(zhí)行順序。接著往下看，在getAdvisors方法中循環(huán)獲取到的methods，分別調(diào)用getAdvisor方法，也就是根據(jù)方法逐個(gè)去創(chuàng)建切面：

public Advisor getAdvisor(Method candidateAdviceMethod, MetadataAwareAspectInstanceFactory aspectInstanceFactory, int declarationOrderInAspect, String aspectName) { validate(aspectInstanceFactory.getAspectMetadata().getAspectClass()); //獲取pointCut對(duì)象，最重要的是從注解中獲取表達(dá)式 AspectJExpressionPointcut expressionPointcut = getPointcut( candidateAdviceMethod, aspectInstanceFactory.getAspectMetadata().getAspectClass()); if (expressionPointcut == null) { return null; } //創(chuàng)建Advisor切面類(lèi)，這才是真正的切面類(lèi)，一個(gè)切面類(lèi)里面肯定要有1、pointCut 2、advice //這里pointCut是expressionPointcut， advice 增強(qiáng)方法是 candidateAdviceMethod return new InstantiationModelAwarePointcutAdvisorImpl(expressionPointcut, candidateAdviceMethod, this, aspectInstanceFactory, declarationOrderInAspect, aspectName); } private static final Class<?>[] ASPECTJ_ANNOTATION_CLASSES = new Class<?>[] { Pointcut.class, Around.class, Before.class, After.class, AfterReturning.class, AfterThrowing.class}; private AspectJExpressionPointcut getPointcut(Method candidateAdviceMethod, Class<?> candidateAspectClass) { //從候選的增強(qiáng)方法里面 candidateAdviceMethod 找有有注解 //Pointcut.class, Around.class, Before.class, After.class, AfterReturning.class, AfterThrowing.class //并把注解信息封裝成AspectJAnnotation對(duì)象 AspectJAnnotation<?> aspectJAnnotation = AbstractAspectJAdvisorFactory.findAspectJAnnotationOnMethod(candidateAdviceMethod); if (aspectJAnnotation == null) { return null; } //創(chuàng)建一個(gè)PointCut類(lèi)，并且把前面從注解里面解析的表達(dá)式設(shè)置進(jìn)去 AspectJExpressionPointcut ajexp = new AspectJExpressionPointcut(candidateAspectClass, new String[0], new Class<?>[0]); ajexp.setExpression(aspectJAnnotation.getPointcutExpression()); if (this.beanFactory != null) { ajexp.setBeanFactory(this.beanFactory); } return ajexp; }

之前就說(shuō)過(guò)切面的定義，是切點(diǎn)和增強(qiáng)的組合，所以這里首先通過(guò)getPointcut獲取到注解對(duì)象，然后new了一個(gè)Pointcut對(duì)象，并將表達(dá)式設(shè)置進(jìn)去。然后在getAdvisor方法中最后new了一個(gè)InstantiationModelAwarePointcutAdvisorImpl對(duì)象：

public InstantiationModelAwarePointcutAdvisorImpl(AspectJExpressionPointcut declaredPointcut, Method aspectJAdviceMethod, AspectJAdvisorFactory aspectJAdvisorFactory, MetadataAwareAspectInstanceFactory aspectInstanceFactory, int declarationOrder, String aspectName) { 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; //這個(gè)方法重點(diǎn)看看，創(chuàng)建advice對(duì)象 this.instantiatedAdvice = instantiateAdvice(this.declaredPointcut); } }

這個(gè)就是我們的切面類(lèi)，在其構(gòu)造方法的最后通過(guò)instantiateAdvice創(chuàng)建了Advice對(duì)象。注意這里傳進(jìn)來(lái)的declarationOrder參數(shù)，它就是循環(huán)method時(shí)的序號(hào)，其作用就是賦值給這里的declarationOrder屬性以及Advice的declarationOrder屬性，在后面排序時(shí)就會(huì)通過(guò)這個(gè)序號(hào)來(lái)比較，因此Advice的執(zhí)行順序是固定的，至于為什么要固定，后面分析完AOP鏈?zhǔn)秸{(diào)用過(guò)程自然就明白了。

public Advice getAdvice(Method candidateAdviceMethod, AspectJExpressionPointcut expressionPointcut, MetadataAwareAspectInstanceFactory aspectInstanceFactory, int declarationOrder, String aspectName) { //獲取有@Aspect注解的類(lèi) Class<?> candidateAspectClass = aspectInstanceFactory.getAspectMetadata().getAspectClass(); validate(candidateAspectClass); //找到candidateAdviceMethod方法上面的注解，并且包裝成AspectJAnnotation對(duì)象，這個(gè)對(duì)象中就有注解類(lèi)型 AspectJAnnotation<?> aspectJAnnotation = AbstractAspectJAdvisorFactory.findAspectJAnnotationOnMethod(candidateAdviceMethod); if (aspectJAnnotation == null) { return null; } AbstractAspectJAdvice springAdvice; //根據(jù)不同的注解類(lèi)型創(chuàng)建不同的advice類(lèi)實(shí)例 switch (aspectJAnnotation.getAnnotationType()) { case AtPointcut: if (logger.isDebugEnabled()) { logger.debug('Processing pointcut ’' + candidateAdviceMethod.getName() + '’'); } return null; case AtAround: //實(shí)現(xiàn)了MethodInterceptor接口 springAdvice = new AspectJAroundAdvice( candidateAdviceMethod, expressionPointcut, aspectInstanceFactory); break; case AtBefore: //實(shí)現(xiàn)了MethodBeforeAdvice接口，沒(méi)有實(shí)現(xiàn)MethodInterceptor接口 springAdvice = new AspectJMethodBeforeAdvice( candidateAdviceMethod, expressionPointcut, aspectInstanceFactory); break; case AtAfter: //實(shí)現(xiàn)了MethodInterceptor接口 springAdvice = new AspectJAfterAdvice( candidateAdviceMethod, expressionPointcut, aspectInstanceFactory); break; case AtAfterReturning: //實(shí)現(xiàn)了AfterReturningAdvice接口，沒(méi)有實(shí)現(xiàn)MethodInterceptor接口 springAdvice = new AspectJAfterReturningAdvice( candidateAdviceMethod, expressionPointcut, aspectInstanceFactory); AfterReturning afterReturningAnnotation = (AfterReturning) aspectJAnnotation.getAnnotation(); if (StringUtils.hasText(afterReturningAnnotation.returning())) { springAdvice.setReturningName(afterReturningAnnotation.returning()); } break; case AtAfterThrowing: //實(shí)現(xiàn)了MethodInterceptor接口 springAdvice = new AspectJAfterThrowingAdvice( candidateAdviceMethod, expressionPointcut, aspectInstanceFactory); AfterThrowing afterThrowingAnnotation = (AfterThrowing) aspectJAnnotation.getAnnotation(); if (StringUtils.hasText(afterThrowingAnnotation.throwing())) { springAdvice.setThrowingName(afterThrowingAnnotation.throwing()); } break; default: throw new UnsupportedOperationException( 'Unsupported advice type on method: ' + candidateAdviceMethod); } // Now to configure the advice... springAdvice.setAspectName(aspectName); springAdvice.setDeclarationOrder(declarationOrder); String[] argNames = this.parameterNameDiscoverer.getParameterNames(candidateAdviceMethod); if (argNames != null) { springAdvice.setArgumentNamesFromStringArray(argNames); } //計(jì)算argNames和類(lèi)型的對(duì)應(yīng)關(guān)系 springAdvice.calculateArgumentBindings(); return springAdvice; }

這里邏輯很清晰，就是拿到方法上的注解類(lèi)型，根據(jù)類(lèi)型創(chuàng)建不同的增強(qiáng)Advice對(duì)象：AspectJAroundAdvice、AspectJMethodBeforeAdvice、AspectJAfterAdvice、AspectJAfterReturningAdvice、AspectJAfterThrowingAdvice。完成之后通過(guò)calculateArgumentBindings方法進(jìn)行參數(shù)綁定，感興趣的可自行研究。這里主要看看幾個(gè)Advice的繼承體系：

可以看到有兩個(gè)Advice是沒(méi)有實(shí)現(xiàn)MethodInterceptor接口的：AspectJMethodBeforeAdvice和AspectJAfterReturningAdvice。而MethodInterceptor有一個(gè)invoke方法，這個(gè)方法就是鏈?zhǔn)秸{(diào)用的核心方法，但那兩個(gè)沒(méi)有實(shí)現(xiàn)該方法的Advice怎么處理呢？稍后會(huì)分析。

到這里切面對(duì)象就創(chuàng)建完成了，接下來(lái)就是判斷當(dāng)前創(chuàng)建的Bean實(shí)例是否和這些切面匹配以及對(duì)切面排序。匹配過(guò)程比較復(fù)雜，對(duì)理解主流程也沒(méi)什么幫助，所以這里就不展開(kāi)分析，感興趣的自行分析（AbstractAdvisorAutoProxyCreator.findAdvisorsThatCanApply()）。

下面看看排序的過(guò)程，回到AbstractAdvisorAutoProxyCreator.findEligibleAdvisors方法：

protected List<Advisor> findEligibleAdvisors(Class<?> beanClass, String beanName) { //找到候選的切面,其實(shí)就是一個(gè)尋找有@Aspectj注解的過(guò)程，把工程中所有有這個(gè)注解的類(lèi)封裝成Advisor返回 List<Advisor> candidateAdvisors = findCandidateAdvisors(); //判斷候選的切面是否作用在當(dāng)前beanClass上面，就是一個(gè)匹配過(guò)程。。現(xiàn)在就是一個(gè)匹配 List<Advisor> eligibleAdvisors = findAdvisorsThatCanApply(candidateAdvisors, beanClass, beanName); extendAdvisors(eligibleAdvisors); if (!eligibleAdvisors.isEmpty()) { //對(duì)有@Order@Priority進(jìn)行排序 eligibleAdvisors = sortAdvisors(eligibleAdvisors); } return eligibleAdvisors; }

sortAdvisors方法就是排序，但這個(gè)方法有兩個(gè)實(shí)現(xiàn)：當(dāng)前類(lèi)AbstractAdvisorAutoProxyCreator和子類(lèi)AspectJAwareAdvisorAutoProxyCreator，應(yīng)該走哪個(gè)呢？

通過(guò)類(lèi)圖我們可以肯定是進(jìn)入的AspectJAwareAdvisorAutoProxyCreator類(lèi)，因?yàn)锳nnotationAwareAspectJAutoProxyCreator的父類(lèi)是它。

protected List<Advisor> sortAdvisors(List<Advisor> advisors) { List<PartiallyComparableAdvisorHolder> partiallyComparableAdvisors = new ArrayList<>(advisors.size()); for (Advisor element : advisors) { partiallyComparableAdvisors.add( new PartiallyComparableAdvisorHolder(element, DEFAULT_PRECEDENCE_COMPARATOR)); } List<PartiallyComparableAdvisorHolder> sorted = PartialOrder.sort(partiallyComparableAdvisors); if (sorted != null) { List<Advisor> result = new ArrayList<>(advisors.size()); for (PartiallyComparableAdvisorHolder pcAdvisor : sorted) { result.add(pcAdvisor.getAdvisor()); } return result; } else { return super.sortAdvisors(advisors); } }

這里排序主要是委托給PartialOrder進(jìn)行的，而在此之前將所有的切面都封裝成了PartiallyComparableAdvisorHolder對(duì)象，注意傳入的DEFAULT_PRECEDENCE_COMPARATOR參數(shù)，這個(gè)就是比較器對(duì)象：

private static final Comparator<Advisor> DEFAULT_PRECEDENCE_COMPARATOR = new AspectJPrecedenceComparator();

所以我們直接看這個(gè)比較器的compare方法：

public int compare(Advisor o1, Advisor o2) { int advisorPrecedence = this.advisorComparator.compare(o1, o2); if (advisorPrecedence == SAME_PRECEDENCE && declaredInSameAspect(o1, o2)) { advisorPrecedence = comparePrecedenceWithinAspect(o1, o2); } return advisorPrecedence; } private final Comparator<? super Advisor> advisorComparator; public AspectJPrecedenceComparator() { this.advisorComparator = AnnotationAwareOrderComparator.INSTANCE; }

第一步先通過(guò)AnnotationAwareOrderComparator去比較，點(diǎn)進(jìn)去看可以發(fā)現(xiàn)是對(duì)實(shí)現(xiàn)了PriorityOrdered和Ordered接口以及標(biāo)記了Priority和Order注解的非同一個(gè)@Aspect類(lèi)中的切面進(jìn)行排序。這個(gè)和之前分析BeanFacotryPostProcessor類(lèi)是一樣的原理。而對(duì)同一個(gè)@Aspect類(lèi)中的切面排序主要是comparePrecedenceWithinAspect方法：

private int comparePrecedenceWithinAspect(Advisor advisor1, Advisor advisor2) { boolean oneOrOtherIsAfterAdvice = (AspectJAopUtils.isAfterAdvice(advisor1) || AspectJAopUtils.isAfterAdvice(advisor2)); int adviceDeclarationOrderDelta = getAspectDeclarationOrder(advisor1) - getAspectDeclarationOrder(advisor2); if (oneOrOtherIsAfterAdvice) { // the advice declared last has higher precedence if (adviceDeclarationOrderDelta < 0) { // advice1 was declared before advice2 // so advice1 has lower precedence return LOWER_PRECEDENCE; } else if (adviceDeclarationOrderDelta == 0) { return SAME_PRECEDENCE; } else { return HIGHER_PRECEDENCE; } } else { // the advice declared first has higher precedence if (adviceDeclarationOrderDelta < 0) { // advice1 was declared before advice2 // so advice1 has higher precedence return HIGHER_PRECEDENCE; } else if (adviceDeclarationOrderDelta == 0) { return SAME_PRECEDENCE; } else { return LOWER_PRECEDENCE; } } } private int getAspectDeclarationOrder(Advisor anAdvisor) { AspectJPrecedenceInformation precedenceInfo = AspectJAopUtils.getAspectJPrecedenceInformationFor(anAdvisor); if (precedenceInfo != null) { return precedenceInfo.getDeclarationOrder(); } else { return 0; } }

這里就是通過(guò)precedenceInfo.getDeclarationOrder拿到在創(chuàng)建InstantiationModelAwarePointcutAdvisorImpl對(duì)象時(shí)設(shè)置的declarationOrder屬性，這就驗(yàn)證了之前的說(shuō)法（實(shí)際上這里排序過(guò)程非常復(fù)雜，不是簡(jiǎn)單的按照這個(gè)屬性進(jìn)行排序）。

當(dāng)上面的一切都進(jìn)行完成后，就該創(chuàng)建代理對(duì)象了，回到AbstractAutoProxyCreator.wrapIfNecessary，看關(guān)鍵部分代碼：

//如果有切面，則生成該bean的代理 if (specificInterceptors != DO_NOT_PROXY) { this.advisedBeans.put(cacheKey, Boolean.TRUE); //把被代理對(duì)象bean實(shí)例封裝到SingletonTargetSource對(duì)象中 Object proxy = createProxy( bean.getClass(), beanName, specificInterceptors, new SingletonTargetSource(bean)); this.proxyTypes.put(cacheKey, proxy.getClass()); return proxy; }

注意這里將被代理對(duì)象封裝成了一個(gè)SingletonTargetSource對(duì)象，它是TargetSource的實(shí)現(xiàn)類(lèi)。

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); } //創(chuàng)建代理工廠(chǎng) ProxyFactory proxyFactory = new ProxyFactory(); proxyFactory.copyFrom(this); if (!proxyFactory.isProxyTargetClass()) { if (shouldProxyTargetClass(beanClass, beanName)) { //proxyTargetClass 是否對(duì)類(lèi)進(jìn)行代理，而不是對(duì)接口進(jìn)行代理，設(shè)置為true時(shí)，使用CGLib代理。 proxyFactory.setProxyTargetClass(true); } else { evaluateProxyInterfaces(beanClass, proxyFactory); } } //把a(bǔ)dvice類(lèi)型的增強(qiáng)包裝成advisor切面 Advisor[] advisors = buildAdvisors(beanName, specificInterceptors); proxyFactory.addAdvisors(advisors); proxyFactory.setTargetSource(targetSource); customizeProxyFactory(proxyFactory); 用來(lái)控制代理工廠(chǎng)被配置后，是否還允許修改代理的配置,默認(rèn)為false proxyFactory.setFrozen(this.freezeProxy); if (advisorsPreFiltered()) { proxyFactory.setPreFiltered(true); } //獲取代理實(shí)例 return proxyFactory.getProxy(getProxyClassLoader()); }

這里通過(guò)ProxyFactory對(duì)象去創(chuàng)建代理實(shí)例，這是工廠(chǎng)模式的體現(xiàn)，但在創(chuàng)建代理對(duì)象之前還有幾個(gè)準(zhǔn)備動(dòng)作：需要判斷是JDK代理還是CGLIB代理以及通過(guò)buildAdvisors方法將擴(kuò)展的Advice封裝成Advisor切面。準(zhǔn)備完成則通過(guò)getProxy創(chuàng)建代理對(duì)象：

public Object getProxy(@Nullable ClassLoader classLoader) { //根據(jù)目標(biāo)對(duì)象是否有接口來(lái)判斷采用什么代理方式，cglib代理還是jdk動(dòng)態(tài)代理 return createAopProxy().getProxy(classLoader); } protected final synchronized AopProxy createAopProxy() { if (!this.active) { activate(); } return getAopProxyFactory().createAopProxy(this); } public AopProxy createAopProxy(AdvisedSupport config) throws AopConfigException { 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); } }

首先通過(guò)配置拿到對(duì)應(yīng)的代理類(lèi)：ObjenesisCglibAopProxy和JdkDynamicAopProxy，然后再通過(guò)getProxy創(chuàng)建Bean的代理，這里以JdkDynamicAopProxy為例：

public Object getProxy(@Nullable ClassLoader classLoader) { //advised是代理工廠(chǎng)對(duì)象 Class<?>[] proxiedInterfaces = AopProxyUtils.completeProxiedInterfaces(this.advised, true); findDefinedEqualsAndHashCodeMethods(proxiedInterfaces); return Proxy.newProxyInstance(classLoader, proxiedInterfaces, this); }

這里的代碼你應(yīng)該不陌生了，就是JDK的原生API，newProxyInstance方法傳入的InvocationHandler對(duì)象是this，因此，最終AOP代理的調(diào)用就是從該類(lèi)中的invoke方法開(kāi)始。至此，代理對(duì)象的創(chuàng)建就完成了，下面來(lái)看下整個(gè)過(guò)程的時(shí)序圖：

小結(jié)

代理對(duì)象的創(chuàng)建過(guò)程整體來(lái)說(shuō)并不復(fù)雜，首先找到所有帶有@Aspect注解的類(lèi)，并獲取其中沒(méi)有@Pointcut注解的方法，循環(huán)創(chuàng)建切面，而創(chuàng)建切面需要切點(diǎn)和增強(qiáng)兩個(gè)元素，其中切點(diǎn)可簡(jiǎn)單理解為我們寫(xiě)的表達(dá)式，增強(qiáng)則是根據(jù)@Before、@Around、@After等注解創(chuàng)建的對(duì)應(yīng)的Advice類(lèi)。切面創(chuàng)建好后則需要循環(huán)判斷哪些切面能對(duì)當(dāng)前的Bean實(shí)例的方法進(jìn)行增強(qiáng)并排序，最后通過(guò)ProxyFactory創(chuàng)建代理對(duì)象。

AOP鏈?zhǔn)秸{(diào)用

熟悉JDK動(dòng)態(tài)代理的都知道通過(guò)代理對(duì)象調(diào)用方法時(shí)，會(huì)進(jìn)入到InvocationHandler對(duì)象的invoke方法，所以我們直接從JdkDynamicAopProxy的這個(gè)方法開(kāi)始：

public Object invoke(Object proxy, Method method, Object[] args) throws Throwable { MethodInvocation invocation; Object oldProxy = null; boolean setProxyContext = false; //從代理工廠(chǎng)中拿到TargetSource對(duì)象，該對(duì)象包裝了被代理實(shí)例bean TargetSource targetSource = this.advised.targetSource; Object target = null; try { //被代理對(duì)象的equals方法和hashCode方法是不能被代理的，不會(huì)走切面 ....... Object retVal; // 可以從當(dāng)前線(xiàn)程中拿到代理對(duì)象 if (this.advised.exposeProxy) { // Make invocation available if necessary. oldProxy = AopContext.setCurrentProxy(proxy); setProxyContext = true; } //這個(gè)target就是被代理實(shí)例 target = targetSource.getTarget(); Class<?> targetClass = (target != null ? target.getClass() : null); //從代理工廠(chǎng)中拿過(guò)濾器鏈 Object是一個(gè)MethodInterceptor類(lèi)型的對(duì)象，其實(shí)就是一個(gè)advice對(duì)象 List<Object> chain = this.advised.getInterceptorsAndDynamicInterceptionAdvice(method, targetClass); //如果該方法沒(méi)有執(zhí)行鏈，則說(shuō)明這個(gè)方法不需要被攔截，則直接反射調(diào)用 if (chain.isEmpty()) { Object[] argsToUse = AopProxyUtils.adaptArgumentsIfNecessary(method, args); retVal = AopUtils.invokeJoinpointUsingReflection(target, method, argsToUse); } else { invocation = new ReflectiveMethodInvocation(proxy, target, method, args, targetClass, 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())) { retVal = proxy; } return retVal; } finally { if (target != null && !targetSource.isStatic()) { // Must have come from TargetSource. targetSource.releaseTarget(target); } if (setProxyContext) { // Restore old proxy. AopContext.setCurrentProxy(oldProxy); } } }

這段代碼比較長(zhǎng)，我刪掉了不關(guān)鍵的地方。首先來(lái)看this.advised.exposeProxy這個(gè)屬性，這在@EnableAspectJAutoProxy注解中可以配置，當(dāng)為true時(shí)，會(huì)將該代理對(duì)象設(shè)置到當(dāng)前線(xiàn)程的ThreadLocal對(duì)象中，這樣就可以通過(guò)AopContext.currentProxy拿到代理對(duì)象。這個(gè)有什么用呢？我相信有經(jīng)驗(yàn)的Java開(kāi)發(fā)都遇到過(guò)這樣一個(gè)BUG，在Service實(shí)現(xiàn)類(lèi)中調(diào)用本類(lèi)中的另一個(gè)方法時(shí)，事務(wù)不會(huì)生效，這是因?yàn)橹苯油ㄟ^(guò)this調(diào)用就不會(huì)調(diào)用到代理對(duì)象的方法，而是原對(duì)象的，所以事務(wù)切面就沒(méi)有生效。因此這種情況下就可以從當(dāng)前線(xiàn)程的ThreadLocal對(duì)象拿到代理對(duì)象，不過(guò)實(shí)際上直接使用@Autowired注入自己本身也可以拿到代理對(duì)象。

接下來(lái)就是通過(guò)getInterceptorsAndDynamicInterceptionAdvice拿到執(zhí)行鏈，看看具體做了哪些事情：

public List<Object> getInterceptorsAndDynamicInterceptionAdvice( Advised config, Method method, @Nullable Class<?> targetClass) { AdvisorAdapterRegistry registry = GlobalAdvisorAdapterRegistry.getInstance(); //從代理工廠(chǎng)中獲得該被代理類(lèi)的所有切面advisor，config就是代理工廠(chǎng)對(duì)象 Advisor[] advisors = config.getAdvisors(); List<Object> interceptorList = new ArrayList<>(advisors.length); Class<?> actualClass = (targetClass != null ? targetClass : method.getDeclaringClass()); Boolean hasIntroductions = null; for (Advisor advisor : advisors) { //大部分走這里 if (advisor instanceof PointcutAdvisor) { // Add it conditionally. PointcutAdvisor pointcutAdvisor = (PointcutAdvisor) advisor; //如果切面的pointCut和被代理對(duì)象是匹配的，說(shuō)明是切面要攔截的對(duì)象 if (config.isPreFiltered() || pointcutAdvisor.getPointcut().getClassFilter().matches(actualClass)) { MethodMatcher mm = pointcutAdvisor.getPointcut().getMethodMatcher(); boolean match; if (mm instanceof IntroductionAwareMethodMatcher) { if (hasIntroductions == null) { hasIntroductions = hasMatchingIntroductions(advisors, actualClass); } match = ((IntroductionAwareMethodMatcher) mm).matches(method, actualClass, hasIntroductions); } else { //接下來(lái)判斷方法是否是切面pointcut需要攔截的方法 match = mm.matches(method, actualClass); } //如果類(lèi)和方法都匹配 if (match) { //獲取到切面advisor中的advice，并且包裝成MethodInterceptor類(lèi)型的對(duì)象 MethodInterceptor[] interceptors = registry.getInterceptors(advisor); if (mm.isRuntime()) { for (MethodInterceptor interceptor : interceptors) { interceptorList.add(new InterceptorAndDynamicMethodMatcher(interceptor, mm)); } } else { interceptorList.addAll(Arrays.asList(interceptors)); } } } } //如果是引介切面 else if (advisor instanceof IntroductionAdvisor) { IntroductionAdvisor ia = (IntroductionAdvisor) advisor; if (config.isPreFiltered() || ia.getClassFilter().matches(actualClass)) { Interceptor[] interceptors = registry.getInterceptors(advisor); interceptorList.addAll(Arrays.asList(interceptors)); } } else { Interceptor[] interceptors = registry.getInterceptors(advisor); interceptorList.addAll(Arrays.asList(interceptors)); } } return interceptorList; }

這也是個(gè)長(zhǎng)方法，看關(guān)鍵的部分，因?yàn)橹拔覀儎?chuàng)建的基本上都是InstantiationModelAwarePointcutAdvisorImpl對(duì)象，該類(lèi)是PointcutAdvisor的實(shí)現(xiàn)類(lèi)，所以會(huì)進(jìn)入第一個(gè)if判斷里，這里首先進(jìn)行匹配，看切點(diǎn)和當(dāng)前對(duì)象以及該對(duì)象的哪些方法匹配，如果能匹配上，則調(diào)用getInterceptors獲取執(zhí)行鏈：

private final List<AdvisorAdapter> adapters = new ArrayList<>(3); public DefaultAdvisorAdapterRegistry() { registerAdvisorAdapter(new MethodBeforeAdviceAdapter()); registerAdvisorAdapter(new AfterReturningAdviceAdapter()); registerAdvisorAdapter(new ThrowsAdviceAdapter()); } public MethodInterceptor[] getInterceptors(Advisor advisor) throws UnknownAdviceTypeException { List<MethodInterceptor> interceptors = new ArrayList<>(3); Advice advice = advisor.getAdvice(); //如果是MethodInterceptor類(lèi)型的，如：AspectJAroundAdvice //AspectJAfterAdvice //AspectJAfterThrowingAdvice if (advice instanceof MethodInterceptor) { interceptors.add((MethodInterceptor) advice); } //處理 AspectJMethodBeforeAdvice AspectJAfterReturningAdvice for (AdvisorAdapter adapter : this.adapters) { if (adapter.supportsAdvice(advice)) { interceptors.add(adapter.getInterceptor(advisor)); } } if (interceptors.isEmpty()) { throw new UnknownAdviceTypeException(advisor.getAdvice()); } return interceptors.toArray(new MethodInterceptor[0]); }

這里我們可以看到如果是MethodInterceptor的實(shí)現(xiàn)類(lèi)，則直接添加到鏈中，如果不是，則需要通過(guò)適配器去包裝后添加，剛好這里有MethodBeforeAdviceAdapter和AfterReturningAdviceAdapter兩個(gè)適配器對(duì)應(yīng)上文兩個(gè)沒(méi)有實(shí)現(xiàn)MethodInterceptor接口的類(lèi)。最后將Interceptors返回。

if (chain.isEmpty()) { Object[] argsToUse = AopProxyUtils.adaptArgumentsIfNecessary(method, args); retVal = AopUtils.invokeJoinpointUsingReflection(target, method, argsToUse);}else { // We need to create a method invocation... invocation = new ReflectiveMethodInvocation(proxy, target, method, args, targetClass, chain); // Proceed to the joinpoint through the interceptor chain. retVal = invocation.proceed();}

返回到invoke方法后，如果執(zhí)行鏈為空，說(shuō)明該方法不需要被增強(qiáng)，所以直接反射調(diào)用原對(duì)象的方法（注意傳入的是TargetSource封裝的被代理對(duì)象）；反之，則通過(guò)ReflectiveMethodInvocation類(lèi)進(jìn)行鏈?zhǔn)秸{(diào)用，關(guān)鍵方法就是proceed：

private int currentInterceptorIndex = -1; public Object proceed() throws Throwable { //如果執(zhí)行鏈中的advice全部執(zhí)行完，則直接調(diào)用joinPoint方法，就是被代理方法 if (this.currentInterceptorIndex == this.interceptorsAndDynamicMethodMatchers.size() - 1) { return invokeJoinpoint(); } Object interceptorOrInterceptionAdvice = this.interceptorsAndDynamicMethodMatchers.get(++this.currentInterceptorIndex); if (interceptorOrInterceptionAdvice instanceof InterceptorAndDynamicMethodMatcher) { InterceptorAndDynamicMethodMatcher dm = (InterceptorAndDynamicMethodMatcher) interceptorOrInterceptionAdvice; Class<?> targetClass = (this.targetClass != null ? this.targetClass : this.method.getDeclaringClass()); if (dm.methodMatcher.matches(this.method, targetClass, this.arguments)) { return dm.interceptor.invoke(this); } else { return proceed(); } } else { //調(diào)用MethodInterceptor中的invoke方法 return ((MethodInterceptor) interceptorOrInterceptionAdvice).invoke(this); } }

這個(gè)方法的核心就在兩個(gè)地方：invokeJoinpoint和interceptorOrInterceptionAdvice.invoke(this)。當(dāng)增強(qiáng)方法調(diào)用完后就會(huì)通過(guò)前者調(diào)用到被代理的方法，否則則是依次調(diào)用Interceptor的invoke方法。下面就分別看看每個(gè)Interceptor是怎么實(shí)現(xiàn)的。

AspectJAroundAdvice public Object invoke(MethodInvocation mi) throws Throwable { if (!(mi instanceof ProxyMethodInvocation)) { throw new IllegalStateException('MethodInvocation is not a Spring ProxyMethodInvocation: ' + mi); } ProxyMethodInvocation pmi = (ProxyMethodInvocation) mi; ProceedingJoinPoint pjp = lazyGetProceedingJoinPoint(pmi); JoinPointMatch jpm = getJoinPointMatch(pmi); return invokeAdviceMethod(pjp, jpm, null, null); }

MethodBeforeAdviceInterceptor -> AspectJMethodBeforeAdvice public Object invoke(MethodInvocation mi) throws Throwable { this.advice.before(mi.getMethod(), mi.getArguments(), mi.getThis()); return mi.proceed(); } public void before(Method method, Object[] args, @Nullable Object target) throws Throwable { invokeAdviceMethod(getJoinPointMatch(), null, null); }

AspectJAfterAdvice public Object invoke(MethodInvocation mi) throws Throwable { try { return mi.proceed(); } finally { invokeAdviceMethod(getJoinPointMatch(), null, null); } }

AfterReturningAdviceInterceptor -> AspectJAfterReturningAdvice public Object invoke(MethodInvocation mi) throws Throwable { Object retVal = mi.proceed(); this.advice.afterReturning(retVal, mi.getMethod(), mi.getArguments(), mi.getThis()); return retVal; } public void afterReturning(@Nullable Object returnValue, Method method, Object[] args, @Nullable Object target) throws Throwable { if (shouldInvokeOnReturnValueOf(method, returnValue)) { invokeAdviceMethod(getJoinPointMatch(), returnValue, null); } }

AspectJAfterThrowingAdvice public Object invoke(MethodInvocation mi) throws Throwable { try { return mi.proceed(); } catch (Throwable ex) { if (shouldInvokeOnThrowing(ex)) { invokeAdviceMethod(getJoinPointMatch(), null, ex); } throw ex; } }

這里的調(diào)用順序是怎樣的呢？其核心就是通過(guò)proceed方法控制流程，每執(zhí)行完一個(gè)Advice就會(huì)回到proceed方法中調(diào)用下一個(gè)Advice。可以思考一下，怎么才能讓調(diào)用結(jié)果滿(mǎn)足如下圖的執(zhí)行順序。

以上就是AOP的鏈?zhǔn)秸{(diào)用過(guò)程，但是這只是只有一個(gè)切面類(lèi)的情況，如果有多個(gè)@Aspect類(lèi)呢，這個(gè)調(diào)用過(guò)程又是怎樣的？其核心思想和“棧”一樣，就是“先進(jìn)后出，后進(jìn)先出”。

AOP擴(kuò)展知識(shí)

一、自定義全局?jǐn)r截器Interceptor

在上文創(chuàng)建代理對(duì)象的時(shí)候有這樣一個(gè)方法：

protected Advisor[] buildAdvisors(@Nullable String beanName, @Nullable Object[] specificInterceptors) { //自定義MethodInterceptor.拿到setInterceptorNames方法注入的Interceptor對(duì)象 Advisor[] commonInterceptors = resolveInterceptorNames(); List<Object> allInterceptors = new ArrayList<>(); if (specificInterceptors != null) { allInterceptors.addAll(Arrays.asList(specificInterceptors)); if (commonInterceptors.length > 0) { if (this.applyCommonInterceptorsFirst) { allInterceptors.addAll(0, Arrays.asList(commonInterceptors)); } else { allInterceptors.addAll(Arrays.asList(commonInterceptors)); } } } Advisor[] advisors = new Advisor[allInterceptors.size()]; for (int i = 0; i < allInterceptors.size(); i++) { //對(duì)自定義的advice要進(jìn)行包裝，把a(bǔ)dvice包裝成advisor對(duì)象，切面對(duì)象 advisors[i] = this.advisorAdapterRegistry.wrap(allInterceptors.get(i)); } return advisors; }

這個(gè)方法的作用就在于我們可以擴(kuò)展我們自己的Interceptor，首先通過(guò)resolveInterceptorNames方法獲取到通過(guò)setInterceptorNames方法設(shè)置的Interceptor，然后調(diào)用DefaultAdvisorAdapterRegistry.wrap方法將其包裝為DefaultPointcutAdvisor對(duì)象并返回：

public Advisor wrap(Object adviceObject) throws UnknownAdviceTypeException { if (adviceObject instanceof Advisor) { return (Advisor) adviceObject; } if (!(adviceObject instanceof Advice)) { throw new UnknownAdviceTypeException(adviceObject); } Advice advice = (Advice) adviceObject; if (advice instanceof MethodInterceptor) { return new DefaultPointcutAdvisor(advice); } for (AdvisorAdapter adapter : this.adapters) { if (adapter.supportsAdvice(advice)) { return new DefaultPointcutAdvisor(advice); } } throw new UnknownAdviceTypeException(advice); } public DefaultPointcutAdvisor(Advice advice) { this(Pointcut.TRUE, advice); }

需要注意DefaultPointcutAdvisor構(gòu)造器里面?zhèn)魅肓艘粋€(gè)Pointcut.TRUE，表示這種擴(kuò)展的Interceptor是全局的攔截器。下面來(lái)看看如何使用：

public class MyMethodInterceptor implements MethodInterceptor { @Override public Object invoke(MethodInvocation invocation) throws Throwable { System.out.println('自定義攔截器'); return invocation.proceed(); }}

首先寫(xiě)一個(gè)類(lèi)實(shí)現(xiàn)MethodInterceptor 接口，在invoke方法中實(shí)現(xiàn)我們的攔截邏輯，然后通過(guò)下面的方式測(cè)試，只要UserService 有AOP攔截就會(huì)發(fā)現(xiàn)自定義的MyMethodInterceptor也生效了。

public void costomInterceptorTest() { AnnotationAwareAspectJAutoProxyCreator bean = applicationContext.getBean(AnnotationAwareAspectJAutoProxyCreator.class); bean.setInterceptorNames('myMethodInterceptor '); UserService userService = applicationContext.getBean(UserService.class); userService.queryUser('dark'); }

但是如果換個(gè)順序，像下面這樣：

public void costomInterceptorTest() { UserService userService = applicationContext.getBean(UserService.class); AnnotationAwareAspectJAutoProxyCreator bean = applicationContext.getBean(AnnotationAwareAspectJAutoProxyCreator.class); bean.setInterceptorNames('myMethodInterceptor '); userService.queryUser('dark'); }

這時(shí)自定義的全局?jǐn)r截器就沒(méi)有作用了，這是為什么呢？因?yàn)楫?dāng)執(zhí)行g(shù)etBean的時(shí)候，如果有切面匹配就會(huì)通過(guò)ProxyFactory去創(chuàng)建代理對(duì)象，注意Interceptor是存到這個(gè)Factory對(duì)象中的，而這個(gè)對(duì)象和代理對(duì)象是一一對(duì)應(yīng)的，因此調(diào)用getBean時(shí)，還沒(méi)有myMethodInterceptor這個(gè)對(duì)象，自定義攔截器就沒(méi)有效果了，也就是說(shuō)要想自定義攔截器生效，就必須在代理對(duì)象生成之前注冊(cè)進(jìn)去。

二、循環(huán)依賴(lài)三級(jí)緩存存在的必要性

在上一篇文章我分析了Spring是如何通過(guò)三級(jí)緩存來(lái)解決循環(huán)依賴(lài)的問(wèn)題的，但你是否考慮過(guò)第三級(jí)緩存為什么要存在？我直接將bean存到二級(jí)不就行了么，為什么還要存一個(gè)ObjectFactory對(duì)象到第三級(jí)緩存中？一個(gè)是因?yàn)椴皇敲總€(gè)Bean都會(huì)出現(xiàn)循環(huán)依賴(lài)，所以三級(jí)緩存只存了一個(gè)工廠(chǎng)對(duì)象；二是我們?cè)贎Autowired對(duì)象時(shí)，想要注入的不一定是Bean本身，而是想要注入一個(gè)修改過(guò)后的對(duì)象，如代理對(duì)象。在AbstractAutowireCapableBeanFactory.getEarlyBeanReference方法中循環(huán)調(diào)用了SmartInstantiationAwareBeanPostProcessor.getEarlyBeanReference方法，AbstractAutoProxyCreator對(duì)象就實(shí)現(xiàn)了該方法：

public Object getEarlyBeanReference(Object bean, String beanName) { Object cacheKey = getCacheKey(bean.getClass(), beanName); if (!this.earlyProxyReferences.contains(cacheKey)) { this.earlyProxyReferences.add(cacheKey); } // 創(chuàng)建代理對(duì)象 return wrapIfNecessary(bean, beanName, cacheKey); }

因此，當(dāng)我們想要對(duì)循壞依賴(lài)的Bean做出修改時(shí)，就可以像AOP這樣做。

三、如何在Bean創(chuàng)建之前提前創(chuàng)建代理對(duì)象

Spring的代理對(duì)象基本上都是在Bean實(shí)例化完成之后創(chuàng)建的，但在文章開(kāi)始我就說(shuō)過(guò)，Spring也提供了一個(gè)機(jī)會(huì)在創(chuàng)建Bean對(duì)象之前就創(chuàng)建代理對(duì)象，在AbstractAutowireCapableBeanFactory.resolveBeforeInstantiation方法中：

protected Object resolveBeforeInstantiation(String beanName, RootBeanDefinition mbd) { Object bean = null; if (!Boolean.FALSE.equals(mbd.beforeInstantiationResolved)) { // Make sure bean class is actually resolved at this point. if (!mbd.isSynthetic() && hasInstantiationAwareBeanPostProcessors()) { Class<?> targetType = determineTargetType(beanName, mbd); if (targetType != null) { bean = applyBeanPostProcessorsBeforeInstantiation(targetType, beanName); if (bean != null) { bean = applyBeanPostProcessorsAfterInitialization(bean, beanName); } } } mbd.beforeInstantiationResolved = (bean != null); } return bean; } protected Object applyBeanPostProcessorsBeforeInstantiation(Class<?> beanClass, String beanName) { for (BeanPostProcessor bp : getBeanPostProcessors()) { if (bp instanceof InstantiationAwareBeanPostProcessor) { InstantiationAwareBeanPostProcessor ibp = (InstantiationAwareBeanPostProcessor) bp; Object result = ibp.postProcessBeforeInstantiation(beanClass, beanName); if (result != null) { return result; } } } return null; }

主要是InstantiationAwareBeanPostProcessor.postProcessBeforeInstantiation方法中，這里又會(huì)進(jìn)入到AbstractAutoProxyCreator類(lèi)中：

public Object postProcessBeforeInstantiation(Class<?> beanClass, String beanName) { TargetSource targetSource = getCustomTargetSource(beanClass, beanName); if (targetSource != null) { if (StringUtils.hasLength(beanName)) { this.targetSourcedBeans.add(beanName); } Object[] specificInterceptors = getAdvicesAndAdvisorsForBean(beanClass, beanName, targetSource); Object proxy = createProxy(beanClass, beanName, specificInterceptors, targetSource); this.proxyTypes.put(cacheKey, proxy.getClass()); return proxy; } return null; } protected TargetSource getCustomTargetSource(Class<?> beanClass, String beanName) { // We can’t create fancy target sources for directly registered singletons. if (this.customTargetSourceCreators != null && this.beanFactory != null && this.beanFactory.containsBean(beanName)) { for (TargetSourceCreator tsc : this.customTargetSourceCreators) { TargetSource ts = tsc.getTargetSource(beanClass, beanName); if (ts != null) { return ts; } } } // No custom TargetSource found. return null; }

看到這里大致應(yīng)該明白了，先是獲取到一個(gè)自定義的TargetSource對(duì)象，然后創(chuàng)建代理對(duì)象，所以我們首先需要自己實(shí)現(xiàn)一個(gè)TargetSource類(lèi)，這里直接繼承一個(gè)抽象類(lèi)，getTarget方法則返回原始對(duì)象：

public class MyTargetSource extends AbstractBeanFactoryBasedTargetSource { @Override public Object getTarget() throws Exception { return getBeanFactory().getBean(getTargetBeanName()); }}

但這還不夠，上面首先判斷了customTargetSourceCreators!=null，而這個(gè)屬性是個(gè)數(shù)組，可以通過(guò)下面這個(gè)方法設(shè)置進(jìn)來(lái)：

public void setCustomTargetSourceCreators(TargetSourceCreator... targetSourceCreators) { this.customTargetSourceCreators = targetSourceCreators; }

所以我們還要實(shí)現(xiàn)一個(gè)TargetSourceCreator類(lèi)，同樣繼承一個(gè)抽象類(lèi)實(shí)現(xiàn)，并只對(duì)userServiceImpl對(duì)象進(jìn)行攔截：

public class MyTargetSourceCreator extends AbstractBeanFactoryBasedTargetSourceCreator { @Override protected AbstractBeanFactoryBasedTargetSource createBeanFactoryBasedTargetSource(Class<?> beanClass, String beanName) { if (getBeanFactory() instanceof ConfigurableListableBeanFactory) { if(beanName.equalsIgnoreCase('userServiceImpl')) { return new MyTargetSource(); } } return null; }}

createBeanFactoryBasedTargetSource方法是在AbstractBeanFactoryBasedTargetSourceCreator.getTargetSource中調(diào)用的，而getTargetSource就是在上面getCustomTargetSource中調(diào)用的。以上工作做完后，還需要將其設(shè)置到AnnotationAwareAspectJAutoProxyCreator對(duì)象中，因此需要我們注入這個(gè)對(duì)象：

@Configurationpublic class TargetSourceCreatorBean { @Autowired private BeanFactory beanFactory; @Bean public AnnotationAwareAspectJAutoProxyCreator annotationAwareAspectJAutoProxyCreator() { AnnotationAwareAspectJAutoProxyCreator creator = new AnnotationAwareAspectJAutoProxyCreator(); MyTargetSourceCreator myTargetSourceCreator = new MyTargetSourceCreator(); myTargetSourceCreator.setBeanFactory(beanFactory); creator.setCustomTargetSourceCreators(myTargetSourceCreator); return creator; }}

這樣，當(dāng)我們通過(guò)getBean獲取userServiceImpl的對(duì)象時(shí)，就會(huì)優(yōu)先生成代理對(duì)象，然后在調(diào)用執(zhí)行鏈的過(guò)程中再通過(guò)TargetSource.getTarget獲取到被代理對(duì)象。但是，為什么我們?cè)趃etTarget方法中調(diào)用getBean就能拿到被代理對(duì)象呢？

繼續(xù)探究，通過(guò)斷點(diǎn)我發(fā)現(xiàn)從getTarget進(jìn)入時(shí)，在resolveBeforeInstantiation方法中返回的bean就是null了，而getBeanPostProcessors方法返回的Processors中也沒(méi)有了AnnotationAwareAspectJAutoProxyCreator對(duì)象，也就是沒(méi)有進(jìn)入到AbstractAutoProxyCreator.postProcessBeforeInstantiation方法中，所以不會(huì)再次獲取到代理對(duì)象，那AnnotationAwareAspectJAutoProxyCreator對(duì)象是在什么時(shí)候移除的呢？

帶著問(wèn)題，我開(kāi)始反推，發(fā)現(xiàn)在AbstractBeanFactoryBasedTargetSourceCreator類(lèi)中有這樣一個(gè)方法buildInternalBeanFactory：

protected DefaultListableBeanFactory buildInternalBeanFactory(ConfigurableBeanFactory containingFactory) { DefaultListableBeanFactory internalBeanFactory = new DefaultListableBeanFactory(containingFactory); // Required so that all BeanPostProcessors, Scopes, etc become available. internalBeanFactory.copyConfigurationFrom(containingFactory); // Filter out BeanPostProcessors that are part of the AOP infrastructure, // since those are only meant to apply to beans defined in the original factory. internalBeanFactory.getBeanPostProcessors().removeIf(beanPostProcessor -> beanPostProcessor instanceof AopInfrastructureBean); return internalBeanFactory; }

在這里移除掉了所有AopInfrastructureBean的子類(lèi)，而AnnotationAwareAspectJAutoProxyCreator就是其子類(lèi)，那這個(gè)方法是在哪里調(diào)用的呢？繼續(xù)反推：

protected DefaultListableBeanFactory getInternalBeanFactoryForBean(String beanName) { synchronized (this.internalBeanFactories) { DefaultListableBeanFactory internalBeanFactory = this.internalBeanFactories.get(beanName); if (internalBeanFactory == null) { internalBeanFactory = buildInternalBeanFactory(this.beanFactory); this.internalBeanFactories.put(beanName, internalBeanFactory); } return internalBeanFactory; } } public final TargetSource getTargetSource(Class<?> beanClass, String beanName) { AbstractBeanFactoryBasedTargetSource targetSource = createBeanFactoryBasedTargetSource(beanClass, beanName); // 創(chuàng)建完targetSource后就移除掉AopInfrastructureBean類(lèi)型的BeanPostProcessor對(duì)象，如AnnotationAwareAspectJAutoProxyCreator DefaultListableBeanFactory internalBeanFactory = getInternalBeanFactoryForBean(beanName); ...... return targetSource; }

至此，關(guān)于TargetSource接口擴(kuò)展的原理就搞明白了。

總結(jié)

本篇篇幅比較長(zhǎng)，主要搞明白Spring代理對(duì)象是如何創(chuàng)建的以及AOP鏈?zhǔn)秸{(diào)用過(guò)程，而后面的擴(kuò)展則是對(duì)AOP以及Bean創(chuàng)建過(guò)程中一些疑惑的補(bǔ)充，可根據(jù)實(shí)際情況學(xué)習(xí)掌握。希望大家多多支持好吧啦網(wǎng)。