详解Spring循环依赖的解决方案
spring针对Bean之间的循环依赖，有自己的处理方案。关键点就是三级缓存。当然这种方案不能解决所有的问题，他只能解决Bean单例模式下非构造函数的循环依赖。
我们就从A->B->C-A这个初始化顺序，也就是A的Bean中需要B的实例，B的Bean中需要C的实例，C的Bean中需要A的实例，当然这种需要不是构造函数那种依赖。前提条件有了，我们就可以开始了。毫无疑问，我们会先初始化A.初始化的方法是org.springframework.beans.factory.support.AbstractBeanFactory#doGetBean
protectedTdoGetBean( finalStringname,finalClass requiredType,finalObject[]args,booleantypeCheckOnly) throwsBeansException{ finalStringbeanName=transformedBeanName(name); Objectbean; //Eagerlychecksingletoncacheformanuallyregisteredsingletons. ObjectsharedInstance=getSingleton(beanName);//关注点1 if(sharedInstance!=null&&args==null){ if(logger.isDebugEnabled()){ if(isSingletonCurrentlyInCreation(beanName)){ logger.debug("Returningeagerlycachedinstanceofsingletonbean'"+beanName+ "'thatisnotfullyinitializedyet-aconsequenceofacircularreference"); } else{ logger.debug("Returningcachedinstanceofsingletonbean'"+beanName+"'"); } } bean=getObjectForBeanInstance(sharedInstance,name,beanName,null); } else{ //Failifwe'realreadycreatingthisbeaninstance: //We'reassumablywithinacircularreference. if(isPrototypeCurrentlyInCreation(beanName)){ thrownewBeanCurrentlyInCreationException(beanName); } //Checkifbeandefinitionexistsinthisfactory. BeanFactoryparentBeanFactory=getParentBeanFactory(); if(parentBeanFactory!=null&&!containsBeanDefinition(beanName)){ //Notfound->checkparent. StringnameToLookup=originalBeanName(name); if(args!=null){ //Delegationtoparentwithexplicitargs. return(T)parentBeanFactory.getBean(nameToLookup,args); } else{ //Noargs->delegatetostandardgetBeanmethod. returnparentBeanFactory.getBean(nameToLookup,requiredType); } } if(!typeCheckOnly){ markBeanAsCreated(beanName); } try{ finalRootBeanDefinitionmbd=getMergedLocalBeanDefinition(beanName); checkMergedBeanDefinition(mbd,beanName,args); //Guaranteeinitializationofbeansthatthecurrentbeandependson. String[]dependsOn=mbd.getDependsOn(); if(dependsOn!=null){ for(StringdependsOnBean:dependsOn){ if(isDependent(beanName,dependsOnBean)){ thrownewBeanCreationException(mbd.getResourceDescription(),beanName, "Circulardepends-onrelationshipbetween'"+beanName+"'and'"+dependsOnBean+"'"); } registerDependentBean(dependsOnBean,beanName); getBean(dependsOnBean); } } //Createbeaninstance. if(mbd.isSingleton()){ //关注点2 sharedInstance=getSingleton(beanName,newObjectFactory
这个方法很长我们一点点说。先看我们的关注点1ObjectsharedInstance=getSingleton(beanName)根据名称从单例的集合中获取单例对象，我们看下这个方法,他最终是org.springframework.beans.factory.support.DefaultSingletonBeanRegistry#getSingleton(java.lang.String,boolean)
protectedObjectgetSingleton(StringbeanName,booleanallowEarlyReference){
ObjectsingletonObject=this.singletonObjects.get(beanName);
if(singletonObject==null&&isSingletonCurrentlyInCreation(beanName)){
synchronized(this.singletonObjects){
singletonObject=this.earlySingletonObjects.get(beanName);
if(singletonObject==null&&allowEarlyReference){
ObjectFactorysingletonFactory=this.singletonFactories.get(beanName);
if(singletonFactory!=null){
singletonObject=singletonFactory.getObject();
this.earlySingletonObjects.put(beanName,singletonObject);
this.singletonFactories.remove(beanName);
}
}
}
}
return(singletonObject!=NULL_OBJECT?singletonObject:null);
}
大家一定要注意这个方法，很关键，我们开篇提到了三级缓存，使用点之一就是这里。到底是哪三级缓存呢，第一级缓存singletonObjects里面放置的是实例化好的单例对象。第二级earlySingletonObjects里面存放的是提前曝光的单例对象（没有完全装配好）。第三级singletonFactories里面存放的是要被实例化的对象的对象工厂。解释好了三级缓存，我们再看看逻辑。第一次进来this.singletonObjects.get(beanName)返回的肯定是null。然后isSingletonCurrentlyInCreation决定了能否进入二级缓存中获取数据。
publicbooleanisSingletonCurrentlyInCreation(StringbeanName){
returnthis.singletonsCurrentlyInCreation.contains(beanName);
}
从singletonsCurrentlyInCreation这个Set中有没有包含传入的BeanName，前面没有地方设置，所以肯定不包含，所以这个方法返回false,后面的流程就不走了。getSingleton这个方法返回的是null。
下面我们看下关注点2.也是一个getSingleton只不过他是真实的创建Bean的过程，我们可以看到传入了一个匿名的ObjectFactory的对象，他的getObject方法中调用的是createBean这个真正的创建Bean的方法。当然我们可以先搁置一下，继续看我们的getSingleton方法
publicObjectgetSingleton(StringbeanName,ObjectFactorysingletonFactory){
Assert.notNull(beanName,"'beanName'mustnotbenull");
synchronized(this.singletonObjects){
ObjectsingletonObject=this.singletonObjects.get(beanName);
if(singletonObject==null){
if(this.singletonsCurrentlyInDestruction){
thrownewBeanCreationNotAllowedException(beanName,
"Singletonbeancreationnotallowedwhilethesingletonsofthisfactoryareindestruction"+
"(DonotrequestabeanfromaBeanFactoryinadestroymethodimplementation!)");
}
if(logger.isDebugEnabled()){
logger.debug("Creatingsharedinstanceofsingletonbean'"+beanName+"'");
}
beforeSingletonCreation(beanName);
booleannewSingleton=false;
booleanrecordSuppressedExceptions=(this.suppressedExceptions==null);
if(recordSuppressedExceptions){
this.suppressedExceptions=newLinkedHashSet();
}
try{
singletonObject=singletonFactory.getObject();
newSingleton=true;
}
catch(IllegalStateExceptionex){
//Hasthesingletonobjectimplicitlyappearedinthemeantime->
//ifyes,proceedwithitsincetheexceptionindicatesthatstate.
singletonObject=this.singletonObjects.get(beanName);
if(singletonObject==null){
throwex;
}
}
catch(BeanCreationExceptionex){
if(recordSuppressedExceptions){
for(ExceptionsuppressedException:this.suppressedExceptions){
ex.addRelatedCause(suppressedException);
}
}
throwex;
}
finally{
if(recordSuppressedExceptions){
this.suppressedExceptions=null;
}
afterSingletonCreation(beanName);
}
if(newSingleton){
addSingleton(beanName,singletonObject);
}
}
return(singletonObject!=NULL_OBJECT?singletonObject:null);
}
}
这个方法的第一句ObjectsingletonObject=this.singletonObjects.get(beanName)从一级缓存中取数据，肯定是null。随后就调用的beforeSingletonCreation方法。
protectedvoidbeforeSingletonCreation(StringbeanName){
if(!this.inCreationCheckExclusions.contains(beanName)&&!this.singletonsCurrentlyInCreation.add(beanName)){
thrownewBeanCurrentlyInCreationException(beanName);
}
}
其中就有往singletonsCurrentlyInCreation这个Set中添加beanName的过程，这个Set很重要，后面会用到。随后就是调用singletonFactory的getObject方法进行真正的创建过程，下面我们看下刚刚上文提到的真正的创建的过程createBean,它里面的核心逻辑是doCreateBean.
protectedObjectdoCreateBean(finalStringbeanName,finalRootBeanDefinitionmbd,finalObject[]args){
//Instantiatethebean.
BeanWrapperinstanceWrapper=null;
if(mbd.isSingleton()){
instanceWrapper=this.factoryBeanInstanceCache.remove(beanName);
}
if(instanceWrapper==null){
instanceWrapper=createBeanInstance(beanName,mbd,args);
}
finalObjectbean=(instanceWrapper!=null?instanceWrapper.getWrappedInstance():null);
ClassbeanType=(instanceWrapper!=null?instanceWrapper.getWrappedClass():null);
//Allowpost-processorstomodifythemergedbeandefinition.
synchronized(mbd.postProcessingLock){
if(!mbd.postProcessed){
applyMergedBeanDefinitionPostProcessors(mbd,beanType,beanName);
mbd.postProcessed=true;
}
}
//Eagerlycachesingletonstobeabletoresolvecircularreferences
//evenwhentriggeredbylifecycleinterfaceslikeBeanFactoryAware.
//关注点3
booleanearlySingletonExposure=(mbd.isSingleton()&&this.allowCircularReferences&&
isSingletonCurrentlyInCreation(beanName));
if(earlySingletonExposure){
if(logger.isDebugEnabled()){
logger.debug("Eagerlycachingbean'"+beanName+
"'toallowforresolvingpotentialcircularreferences");
}
addSingletonFactory(beanName,newObjectFactory
createBeanInstance利用反射创建了对象，下面我们看看关注点3earlySingletonExposure属性值的判断，其中有一个判断点就是isSingletonCurrentlyInCreation(beanName)
publicbooleanisSingletonCurrentlyInCreation(StringbeanName){
returnthis.singletonsCurrentlyInCreation.contains(beanName);
}
发现使用的是singletonsCurrentlyInCreation这个Set,上文的步骤中已经将BeanName已经填充进去了，所以可以查到，所以earlySingletonExposure这个属性是结合其他的条件综合判断为true,进行下面的流程addSingletonFactory,这里是为这个Bean添加ObjectFactory,这个BeanName(A)对应的对象工厂，他的getObject方法的实现是通过getEarlyBeanReference这个方法实现的。首先我们看下addSingletonFactory的实现
protectedvoidaddSingletonFactory(StringbeanName,ObjectFactorysingletonFactory){
Assert.notNull(singletonFactory,"Singletonfactorymustnotbenull");
synchronized(this.singletonObjects){
if(!this.singletonObjects.containsKey(beanName)){
this.singletonFactories.put(beanName,singletonFactory);
this.earlySingletonObjects.remove(beanName);
this.registeredSingletons.add(beanName);
}
}
}
往第三级缓存singletonFactories存放数据，清除第二级缓存根据beanName的数据。这里有个很重要的点，是往三级缓存里面set了值，这是Spring处理循环依赖的核心点。getEarlyBeanReference这个方法是getObject的实现，可以简单认为是返回了一个为填充完毕的A的对象实例。设置完三级缓存后，就开始了填充A对象属性的过程。下面这段描述，没有源码提示，只是简单的介绍一下。
填充A的时候，发现需要B类型的Bean，于是继续调用getBean方法创建，记性的流程和上面A的完全一致，然后到了填充C类型的Bean的过程，同样的调用getBean(C)来执行，同样到了填充属性A的时候，调用了getBean(A),我们从这里继续说，调用了doGetBean中的ObjectsharedInstance=getSingleton(beanName),相同的代码，但是处理逻辑完全不一样了。
protectedObjectgetSingleton(StringbeanName,booleanallowEarlyReference){
ObjectsingletonObject=this.singletonObjects.get(beanName);
if(singletonObject==null&&isSingletonCurrentlyInCreation(beanName)){
synchronized(this.singletonObjects){
singletonObject=this.earlySingletonObjects.get(beanName);
if(singletonObject==null&&allowEarlyReference){
ObjectFactorysingletonFactory=this.singletonFactories.get(beanName);
if(singletonFactory!=null){
singletonObject=singletonFactory.getObject();
this.earlySingletonObjects.put(beanName,singletonObject);
this.singletonFactories.remove(beanName);
}
}
}
}
return(singletonObject!=NULL_OBJECT?singletonObject:null);
}
还是从singletonObjects获取对象获取不到，因为A是在singletonsCurrentlyInCreation这个Set中，所以进入了下面的逻辑，从二级缓存earlySingletonObjects中取，还是没有查到，然后从三级缓存singletonFactories找到对应的对象工厂调用getObject方法获取未完全填充完毕的A的实例对象，然后删除三级缓存的数据，填充二级缓存的数据，返回这个对象A。C依赖A的实例填充完毕了，虽然这个A是不完整的。不管怎么样C式填充完了，就可以将C放到一级缓存singletonObjects同时清理二级和三级缓存的数据。同样的流程B依赖的C填充好了，B也就填充好了，同理A依赖的B填充好了，A也就填充好了。Spring就是通过这种方式来解决循环引用的。
以上就是本文的全部内容，希望对大家的学习有所帮助，也希望大家多多支持毛票票。
声明：本文内容来源于网络，版权归原作者所有，内容由互联网用户自发贡献自行上传，本网站不拥有所有权，未作人工编辑处理，也不承担相关法律责任。如果您发现有涉嫌版权的内容，欢迎发送邮件至：czq8825#qq.com（发邮件时，请将#更换为@）进行举报，并提供相关证据，一经查实，本站将立刻删除涉嫌侵权内容。