由《Spring的IOC的源碼解析(三)》繼續分析! 首先介紹一下容器命名空間句柄 ContextNamespaceHandler 位置:spring-context-*.jar的META-INF/spring-handlers 容器命名空間句柄: http\:// www.springframework.org/schema/context=org.springframework.context.config.ContextNamespaceHandler ContextNamespaceHandler的init方法實作如下, registerBeanDefinitionParser("property-placeholder", new PropertyPlaceholderBeanDefinitionParser()); registerBeanDefinitionParser("property-override", new PropertyOverrideBeanDefinitionParser()); registerBeanDefinitionParser("annotation-config", new AnnotationConfigBeanDefinitionParser()); registerBeanDefinitionParser("component-scan", new ComponentScanBeanDefinitionParser()); registerBeanDefinitionParser("load-time-weaver", new LoadTimeWeaverBeanDefinitionParser()); registerBeanDefinitionParser("spring-configured", new SpringConfiguredBeanDefinitionParser()); registerBeanDefinitionParser("mbean-export", new MBeanExportBeanDefinitionParser()); registerBeanDefinitionParser("mbean-server", new MBeanServerBeanDefinitionParser());
當 ContextNamespaceHandler在解析标簽的時候,會調用每種标簽的解析器,解析過程就會調用 AopNamespaceUtils為目前類型的标簽 嘗試 注冊相應的BPP;下邊給出一些示例,格式為: {标簽-标簽解析器-注冊的BPP} config/ ConfigBeanDefinitionParser/ AspectJAwareAdvisorAutoProxyCreator aspectj-autoproxy/ AspectJAutoProxyBeanDefinitionParser/ AnnotationAwareAspectJAutoProxyCreator load-time-weaver/ LoadTimeWeaverBeanDefinitionParser/沒有 component-scan/ ComponentScanBeanDefinitionParser/沒有
繼續分析AOP Proxy對象的生成過程! (1)擷取能夠處理目标類的advisor 首先是調用AbstractAdvisorAutoProxyCreator. getAdvicesAndAdvisorsForBean()方法,然後調用: protected List<Advisor> findEligibleAdvisors(Class<?> beanClass, String beanName) { List<Advisor> candidateAdvisors = findCandidateAdvisors(); //這個方法就是去beanFactory中查詢全部Advisor類型的bean,然後初始化;參見(a) List<Advisor> eligibleAdvisors = findAdvisorsThatCanApply(candidateAdvisors, beanClass, beanName);//參看(b) extendAdvisors(eligibleAdvisors); if (!eligibleAdvisors.isEmpty()) { eligibleAdvisors = sortAdvisors(eligibleAdvisors);//排序 } return eligibleAdvisors; } (a) 擷取并初始化Advisor,實作邏輯在 BeanFactoryAdvisorRetrievalHelper.findAdvisorBeans(),其核心代碼是: advisorNames = BeanFactoryUtils.beanNamesForTypeIncludingAncestors(this.beanFactory, Advisor.class, true, false); //查詢所有注冊在工廠中的Advisor的bean name or id List<Advisor> advisors = new LinkedList<>(); advisors.add(this.beanFactory.getBean(name, Advisor.class)); //初始化Advisor,同時也會初始化Advisor中的advice,生成一個advice執行個體;
(b) 往後又調用 AopUtils.findAdvisorsThatCanApply()方法; findAdvisorsThatCanApply調用的核心函數就是: public static boolean canApply(Advisor advisor, Class<?> targetClass, boolean hasIntroductions) { if (advisor instanceof IntroductionAdvisor) { return ((IntroductionAdvisor) advisor).getClassFilter().matches(targetClass); //advisor是 IntroductionAdvisor的比對過程 } else if (advisor instanceof PointcutAdvisor) {//正常使用AOP,都是這種advisor PointcutAdvisor pca = (PointcutAdvisor) advisor; return canApply(pca.getPointcut(), targetClass, hasIntroductions); //advisor是 PointcutAdvisor的比對過程 } else { // It doesn't have a pointcut so we assume it applies. return true; //其他的 advisor 預設要處理 } } pointcut的比對過程如下: public static boolean canApply(Pointcut pc, Class<?> targetClass, boolean hasIntroductions) { Assert.notNull(pc, "Pointcut must not be null"); if (!pc.getClassFilter().matches(targetClass)) { //擷取advisor中的pointcut對象,關于pointcut的使用,請參看《 spring中的pointcut 》;這裡相當于把所有的advisor中的pointcut都拿來比對一遍,看看是否滿足 return false; }
MethodMatcher methodMatcher = pc.getMethodMatcher(); 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; }
Set<Class<?>> classes = new LinkedHashSet<>(ClassUtils.getAllInterfacesForClassAsSet(targetClass)); classes.add(targetClass); for (Class<?> clazz : classes) { Method[] methods = ReflectionUtils.getAllDeclaredMethods(clazz); for (Method method : methods) { if ((introductionAwareMethodMatcher != null && introductionAwareMethodMatcher.matches(method, targetClass, hasIntroductions)) || methodMatcher.matches(method, targetClass)) { return true; } } }
return false; }
最終,傳回所有适合代理類的advisor
(2)建立AOP代理 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); // 為bean設定屬性org.springframework.aop.framework.autoproxy.AutoProxyUtils.originalTargetClass,值是beanClass } ProxyFactory proxyFactory = new ProxyFactory(); //代理工廠 proxyFactory.copyFrom(this); //為代理工廠設定proxyTargetClass,是否初始化等參數
if (!proxyFactory.isProxyTargetClass()) { //預設應該條件為true,進入if if (shouldProxyTargetClass(beanClass, beanName)) { //判斷是代理類,還是代理類的接口;判斷的邏輯就是看bean的定義中,屬性名為 org.springframework.aop.framework.autoproxy.AutoProxyUtils.preserveTargetClass的屬性的值是否設定為了true。暫時隻知道 ConfigurationClassPostProcessor可以配置這個屬性為true;預設是false,即代理接口 proxyFactory.setProxyTargetClass(true); //代理目标類,标記 proxyTargetClass=true } else { evaluateProxyInterfaces(beanClass, proxyFactory); //代理目标類的接口;如果是可以代理的接口,就把接口加入到ProxyFactory中,見(a) } } Advisor[] advisors = buildAdvisors(beanName, specificInterceptors); //合并共用和指定攔截器到advisor,可以把advisor叫做增強器;見(b) proxyFactory.addAdvisors(advisors); //這是增強器 proxyFactory.setTargetSource(targetSource);//設定執行個體對象,用于代理調用代理方法是使用 customizeProxyFactory(proxyFactory);
proxyFactory.setFrozen(this.freezeProxy); if (advisorsPreFiltered()) { proxyFactory.setPreFiltered(true); } return proxyFactory.getProxy(getProxyClassLoader()); //生成代理對象,見(c) }
(a) 可代理的接口的定義 至少有一個方法的接口 或者 protected boolean isInternalLanguageInterface(Class<?> ifc) { return (ifc.getName().equals("groovy.lang.GroovyObject") || ifc.getName().endsWith(".cglib.proxy.Factory") || ifc.getName().endsWith(".bytebuddy.MockAccess")); } 傳回false的接口 或者 protected boolean isConfigurationCallbackInterface(Class<?> ifc) { return (InitializingBean.class == ifc || DisposableBean.class == ifc || Closeable.class == ifc || AutoCloseable.class == ifc || ObjectUtils.containsElement(ifc.getInterfaces(), Aware.class)); } 傳回false的接口
(b)合并共用和指定攔截器到advisor,預設沒有共用攔截器 protected Advisor[] buildAdvisors(@Nullable String beanName, @Nullable Object[] specificInterceptors) { Advisor[] commonInterceptors = resolveInterceptorNames(); //把定義的攔截器包裝成advisor,傳回 List<Object> allInterceptors = new ArrayList<>(); if (specificInterceptors != null) { allInterceptors.addAll(Arrays.asList(specificInterceptors)); if (commonInterceptors.length > 0) { if (this.applyCommonInterceptorsFirst) { //預設把攔截器的執行放在最前邊執行 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++) { advisors[i] = this.advisorAdapterRegistry.wrap(allInterceptors.get(i)); //包裝advisor,适配方法是 DefaultAdvisorAdapterRegistry.wrap() } return advisors; }
适配過程: public Advisor wrap(Object adviceObject) throws UnknownAdviceTypeException { if (adviceObject instanceof Advisor) { return (Advisor) adviceObject; //如果是advisor對象,不用處理 } if (!(adviceObject instanceof Advice)) { //隻能是advicor or advice throw new UnknownAdviceTypeException(adviceObject); } Advice advice = (Advice) adviceObject; //advice對象都包裝成 DefaultPointcutAdvisor對象 if (advice instanceof MethodInterceptor) { // So well-known it doesn't even need an adapter. return new DefaultPointcutAdvisor(advice); } for (AdvisorAdapter adapter : this.adapters) { // Check that it is supported. if (adapter.supportsAdvice(advice)) { return new DefaultPointcutAdvisor(advice); } } throw new UnknownAdviceTypeException(advice); }
(c)代理工廠生成代理對象 這裡一直沒有讀懂aopProxyFactory對象的執行個體話,這裡假定aopProxyFactory就是一個DefaultAopProxyFactory對象; DefaultAopProxyFactory中生成代理對象的方法: 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); //類代理,分析過程見【1】 } else { return new JdkDynamicAopProxy(config); //接口代理, 分析過程見 【2】 } } 如果不是代理接口,就使用CGLIB位元組碼動态代理;否則使用JDK的動态代理。
【1】 jdk動态代理 JdkDynamicAopProxy.getProxy()方法: public Object getProxy(@Nullable ClassLoader classLoader) { Class<?>[] proxiedInterfaces = AopProxyUtils.completeProxiedInterfaces(this.advised , true); findDefinedEqualsAndHashCodeMethods(proxiedInterfaces); return Proxy.newProxyInstance(classLoader, proxiedInterfaces, this); } 是以,傳回的代理對象就是JDK的Proxy.newProxyInstance方法生成的,而且參數需要的 InvocationHandler對象就是目前的JdkDynamicAopProxy對象; 我們知道,JDK動态代理的實作原理就是代理對象在執行目标方法的時候,會檢查目前方法是否被代理?如果被代理,那麼就執行 InvocationHandler的invoke方法代替;
下邊看看JDK動态代理模式下的,代理對象真正執行方法時,切面織入方法的功能是怎樣實作的。 JdkDynamicAopProxy實作 InvocationHandler接口,invoke方法的核心邏輯是: List<Object> chain = this.advised.getInterceptorsAndDynamicInterceptionAdvice(method, targetClass); // chain是 InterceptorAndDynamicMethodMatcher連結清單; 參見{1} if (chain.isEmpty()) { //跳過熱交換,隻是執行目标類的方法,method.invoke Object[] argsToUse = AopProxyUtils.adaptArgumentsIfNecessary(method, args); retVal = AopUtils.invokeJoinpointUsingReflection(target, method, argsToUse); } else { //建立 method invocation invocation = new ReflectiveMethodInvocation(proxy, target, method, args, targetClass, chain); //反射方法處理過程,target參數就是被代理類的執行個體 //執行攔截器鍊和方法,參看{4} retVal = invocation.proceed(); }
{1}把增強器advisors包裝成mathodMatcher鍊的過程 調用DefaultAdvisorChainFactory.getInterceptorsAndDynamicInterceptionAdvice()方法, 核心代碼如下: MethodInterceptor[] interceptors = registry.getInterceptors(advisor); // registry是 DefaultAdvisorAdapterRegistry執行個體,這裡就是把advisor中的advice包裝成 MethodInterceptor,見{2} MethodMatcher mm = pointcutAdvisor.getPointcut().getMethodMatcher(); //獲得比對方法;參見{3} if (MethodMatchers.matches(mm, method, actualClass, hasIntroductions)) { if (mm.isRuntime()) { for (MethodInterceptor interceptor : interceptors) { interceptorList.add(new InterceptorAndDynamicMethodMatcher(interceptor, mm)); //采用組合模式,将攔截器對象和比對方法對象放在 InterceptorAndDynamicMethodMatcher對象中,得到傳回值對象; } } else { interceptorList.addAll(Arrays.asList(interceptors)); } }
{2}例如:下邊是預設包裝關系 MethodBeforeAdvice MethodBeforeAdviceInterceptor AfterReturningAdvice AfterReturningAdviceInterceptor ThrowsAdvice ThrowsAdviceInterceptor 根據advice的解析過程可知,before 類型的advice的目标類就是MethodBeforeAdvice類的子類;advice的pointcut依賴指定的目标類是AspectJExpressionPointcut
{3}由{2}的說明可知,MethodMatcher方法對象的擷取過程就是擷取AspectJExpressionPointcut對象;在pointcut執行個體化時會解析expression表達式,邏輯如下: private PointcutExpression buildPointcutExpression(@Nullable ClassLoader classLoader) { PointcutParser parser = initializePointcutParser(classLoader); PointcutParameter[] pointcutParameters = new PointcutParameter[this.pointcutParameterNames.length]; for (int i = 0; i < pointcutParameters.length; i++) { pointcutParameters[i] = parser.createPointcutParameter( this.pointcutParameterNames[i], this.pointcutParameterTypes[i]); } return parser.parsePointcutExpression(replaceBooleanOperators(resolveExpression()), this.pointcutDeclarationScope, pointcutParameters); //expression的對象類是 PointcutExpressionImpl } 需要說明的是,AspectJExpressionPointcut類實作IntroductionAwareMethodMatcher、MethodMatcher接口;
{4}循環處理攔截器鍊,使用matcher比對,然後調用攔截器的invoke方法 if (this.currentInterceptorIndex == this.interceptorsAndDynamicMethodMatchers.size() - 1) { //所有 advice都調用完之後,調用這裡,但是after advice是逆向調用的,所有代理方法會在after adviec之前之前執行 return invokeJoinpoint(); } //other code if (dm.methodMatcher.matches(this.method, this.targetClass, this.arguments)) { //比對;見{5}分析 return dm.interceptor.invoke(this); //調用;參見{6} } else { // Dynamic matching failed. // Skip this interceptor and invoke the next in the chain. return proceed(); //如果比對失敗,跳過目前advisor,繼續下一個 }
{5}可知比對過程就是AspectJExpressionPointcut的matches方法邏輯過程; {6}看看典型advice的處理過程 before advice的 MethodBeforeAdviceInterceptor調用過程如下: public Object invoke(MethodInvocation mi) throws Throwable { this.advice.before(mi.getMethod(), mi.getArguments(), mi.getThis() ); // AspectJMethodBeforeAdvice.before方法,這裡就是調用before通知在切面裡定義的執行方法 return mi.proceed(); //跳轉到 ReflectiveMethodInvocation. proceed()方法,實作調用多個advice的功能 }
AspectJMethodBeforeAdvice.before(){ invokeAdviceMethod(getJoinPointMatch(), null, null); }
after-returning的AfterReturningAdviceInterceptor調用過程如下: public Object invoke(MethodInvocation mi) throws Throwable { Object retVal = mi.proceed(); //跳轉到 ReflectiveMethodInvocation. proceed()方法,實作調用多個advice的功能 this.advice.afterReturning(retVal, mi.getMethod(), mi.getArguments(), mi.getThis()); //調用後處理方法 return retVal; }
AspectJAfterReturningAdvice.afterReturning(){ if (shouldInvokeOnReturnValueOf(method, returnValue)) { invokeAdviceMethod(getJoinPointMatch(), returnValue, null); } } 可見,所有的advice都是先執行before advice 再執行after-returning advice ;所有的advice都執行完之後,執行被代理的方法;因為after advice是逆向調用的,是以被代理的方法會在after advice之前調用。
around advice通知是怎麼實作的呢? 看到DefaultAdvisorAdapterRegistry類的wrap方法: if (adviceObject instanceof Advisor) { //advice 不是advisor,跳過 return (Advisor) adviceObject; } if (!(adviceObject instanceof Advice)) { throw new UnknownAdviceTypeException(adviceObject); } Advice advice = (Advice) adviceObject; if (advice instanceof MethodInterceptor) { // So well-known it doesn't even need an adapter. return new DefaultPointcutAdvisor(advice); //advice 都在這裡處理,被适配成 DefaultPointcutAdvisor對象 }
再看攔截器方法鍊——也就是advice的方法封裝,也是DefaultAdvisorAdapterRegistry的方法: public MethodInterceptor[] getInterceptors(Advisor advisor) throws UnknownAdviceTypeException { List<MethodInterceptor> interceptors = new ArrayList<>(3); Advice advice = advisor.getAdvice(); if (advice instanceof MethodInterceptor) { //around advice的處理類 AspectJAroundAdvice實作了 MethodInterceptor接口,是以就是around advice的方法攔截器就是 AspectJAroundAdvice對象 interceptors.add((MethodInterceptor) advice); } 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[interceptors.size()]); } 下邊看方法攔截器的調用過程: 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); //傳回一個 MethodInvocationProceedingJoinPoint對象,這個對象就是在切面中定義的around方法的實參 JoinPointMatch jpm = getJoinPointMatch(pmi); return invokeAdviceMethod(pjp, jpm, null, null); //調用父類的方法;參見{7} }
{7}advice方法的調用,核心就是下邊這一句 this.aspectJAdviceMethod.invoke(this.aspectInstanceFactory.getAspectInstance(), actualArgs); 解釋: this.aspectJAdviceMethod擷取advice指定方法的Method對象; this.aspectInstanceFactory.getAspectInstance()擷取一個advice所屬切面aspect的執行個體; actualArgs是方法參數; 也就是method.invoke(object,args)而已;
before advice,after-returning advice在調用advice方法時,會調用下一個advice的方法,但是around advice沒有這樣做;事實上,around advice調用下一個advice方法的語句在,切面實作的方法中,從這一點上将,隻有around advice的方法參數是必須要被使用的;而在around advice方法中,可以在point.proceed();語句前後加增強語句,正是“環繞”的意義所在;需要注意的是,如果有多個around advice被執行,那麼point.proceed();語句後面的增強語句的執行過程像堆棧調用,即先被調用的後被執行;
到此,基于JDK的動态代理實作的AOP過程就分析完成了!
【2】 位元組碼代理 ObjenesisCglibAopProxy. getProxy()方法,暫時不研究了!