你所不知道的Spring的@Autowired實(shí)現(xiàn)細(xì)節(jié)分析

更新時(shí)間：2020年08月26日 09:27:58 作者：夜勿語(yǔ)

這篇文章主要介紹了你所不知道的Spring的@Autowired實(shí)現(xiàn)細(xì)節(jié)分析，具有很好的參考價(jià)值，希望對(duì)大家有所幫助。一起跟隨小編過(guò)來(lái)看看吧

前言

相信很多Java開(kāi)發(fā)都遇到過(guò)一個(gè)面試題：Resource和Autowired的區(qū)別是什么？這個(gè)問(wèn)題的答案相信基本都清楚，但是這兩者在Spring中是如何實(shí)現(xiàn)的呢？這就要分析Spring源碼才能知道了。友情提示：本篇主要是講解Autowired的實(shí)現(xiàn)原理，不會(huì)分析Spring初始化的過(guò)程，不熟悉的讀者可以先閱讀筆者之前的一篇文章《這一次搞懂Spring的Bean實(shí)例化原理》。

正文

在Spring Bean的整個(gè)創(chuàng)建過(guò)程中會(huì)調(diào)用很多BeanPostProcessor接口的的實(shí)現(xiàn)類(lèi)：

上圖是我整理的Spring Bean的創(chuàng)建過(guò)程及整個(gè)過(guò)程中各個(gè)BeanPostProcessor和回調(diào)的調(diào)用，右邊相同顏色的連線代表是同一組調(diào)用，主要看到AutowiredAnnotationBeanPostProcessor和CommonAnnotationBeanPostProcessor，前者就是支持 @Autowired和@Value注解，后者則是支持@PostConstruct、@PreDestroy、@Resource注解。先了解這兩個(gè)Processor的作用，下面從頭分析。

從圖中可以看到，在createBeanInstance方法中會(huì)調(diào)用SmartInstantiationAwareBeanPostProcessor類(lèi)型的determineCandidateConstructors，這個(gè)方法是做什么的呢？看代碼：

protected BeanWrapper createBeanInstance(String beanName, RootBeanDefinition mbd, @Nullable Object[] args) {
 ......
 // Candidate constructors for autowiring?
 //尋找當(dāng)前正在實(shí)例化的bean中有@Autowired注解的構(gòu)造函數(shù)
 Constructor<?>[] ctors = determineConstructorsFromBeanPostProcessors(beanClass, beanName);
 if (ctors != null || mbd.getResolvedAutowireMode() == AUTOWIRE_CONSTRUCTOR ||
 mbd.hasConstructorArgumentValues() || !ObjectUtils.isEmpty(args)) {
 //如果ctors不為空，就說(shuō)明構(gòu)造函數(shù)上有@Autowired注解
 return autowireConstructor(beanName, mbd, ctors, args);
 } 
 ......
 
 // No special handling: simply use no-arg constructor.
 return instantiateBean(beanName, mbd);
 }

 protected Constructor<?>[] determineConstructorsFromBeanPostProcessors(@Nullable Class<?> beanClass, String beanName)
 throws BeansException {

 if (beanClass != null && hasInstantiationAwareBeanPostProcessors()) {
 for (BeanPostProcessor bp : getBeanPostProcessors()) {
 if (bp instanceof SmartInstantiationAwareBeanPostProcessor) {
  SmartInstantiationAwareBeanPostProcessor ibp = (SmartInstantiationAwareBeanPostProcessor) bp;
  Constructor<?>[] ctors = ibp.determineCandidateConstructors(beanClass, beanName);
  if (ctors != null) {
  return ctors;
  }
 }
 }
 }
 return null;
 }

createBeanInstance方法是去實(shí)例化Bean，而調(diào)用AutowiredAnnotationBeanPostProcessor.determineCandidateConstructors的目的就是先去找到帶有@Autowired注解的構(gòu)造方法（自動(dòng)注入有三種模式：屬性、構(gòu)造方法、普通方法），也就是通過(guò)構(gòu)造方法注入，如果沒(méi)有找到則通過(guò)反射調(diào)用無(wú)參構(gòu)造實(shí)例化。平時(shí)我們基本上都是使用的屬性注入，所以一般都不會(huì)進(jìn)入determineCandidateConstructors方法，所以這里也不詳細(xì)闡述，感興趣的讀者可自行看看。

接著回到doCreateBean方法中，就可以看到調(diào)用了applyMergedBeanDefinitionPostProcessors方法：

 protected void applyMergedBeanDefinitionPostProcessors(RootBeanDefinition mbd, Class<?> beanType, String beanName) {
 for (BeanPostProcessor bp : getBeanPostProcessors()) {
 if (bp instanceof MergedBeanDefinitionPostProcessor) {
 /**
  * CommonAnnotationBeanPostProcessor 支持了@PostConstruct，@PreDestroy,@Resource注解
  * AutowiredAnnotationBeanPostProcessor 支持 @Autowired,@Value注解
  */
 MergedBeanDefinitionPostProcessor bdp = (MergedBeanDefinitionPostProcessor) bp;
 bdp.postProcessMergedBeanDefinition(mbd, beanType, beanName);
 }
 }
 }

這個(gè)方法本質(zhì)上就是調(diào)用MergedBeanDefinitionPostProcessor類(lèi)型的postProcessMergedBeanDefinition方法，通過(guò)這個(gè)方法去收集@Autowired、@Resource等注解，這里主要分析AutowiredAnnotationBeanPostProcessor的實(shí)現(xiàn)：

 public void postProcessMergedBeanDefinition(RootBeanDefinition beanDefinition, Class<?> beanType, String beanName) {
 InjectionMetadata metadata = findAutowiringMetadata(beanName, beanType, null);
 metadata.checkConfigMembers(beanDefinition);
 }

 private InjectionMetadata findAutowiringMetadata(String beanName, Class<?> clazz, @Nullable PropertyValues pvs) {
 // Fall back to class name as cache key, for backwards compatibility with custom callers.
 String cacheKey = (StringUtils.hasLength(beanName) ? beanName : clazz.getName());
 // Quick check on the concurrent map first, with minimal locking.
 InjectionMetadata metadata = this.injectionMetadataCache.get(cacheKey);
 if (InjectionMetadata.needsRefresh(metadata, clazz)) {
 synchronized (this.injectionMetadataCache) {
 metadata = this.injectionMetadataCache.get(cacheKey);
 if (InjectionMetadata.needsRefresh(metadata, clazz)) {
  if (metadata != null) {
  metadata.clear(pvs);
  }
  //主要看這個(gè)方法
  metadata = buildAutowiringMetadata(clazz);
  this.injectionMetadataCache.put(cacheKey, metadata);
 }
 }
 }
 return metadata;
 }

 private InjectionMetadata buildAutowiringMetadata(final Class<?> clazz) {
 List<InjectionMetadata.InjectedElement> elements = new ArrayList<>();
 Class<?> targetClass = clazz;

 do {
 final List<InjectionMetadata.InjectedElement> currElements = new ArrayList<>();

 // 找到帶有@Autowired注解的屬性并封裝為AnnotationAttributes
 ReflectionUtils.doWithLocalFields(targetClass, field -> {
 AnnotationAttributes ann = findAutowiredAnnotation(field);
 if (ann != null) {
  if (Modifier.isStatic(field.getModifiers())) {
  if (logger.isInfoEnabled()) {
  logger.info("Autowired annotation is not supported on static fields: " + field);
  }
  return;
  }
  boolean required = determineRequiredStatus(ann);
  currElements.add(new AutowiredFieldElement(field, required));
 }
 });

 // 找到帶有@Autowired注解的方法并封裝為AnnotationAttributes
 ReflectionUtils.doWithLocalMethods(targetClass, method -> {
 Method bridgedMethod = BridgeMethodResolver.findBridgedMethod(method);
 if (!BridgeMethodResolver.isVisibilityBridgeMethodPair(method, bridgedMethod)) {
  return;
 }
 AnnotationAttributes ann = findAutowiredAnnotation(bridgedMethod);
 if (ann != null && method.equals(ClassUtils.getMostSpecificMethod(method, clazz))) {
  if (Modifier.isStatic(method.getModifiers())) {
  if (logger.isInfoEnabled()) {
  logger.info("Autowired annotation is not supported on static methods: " + method);
  }
  return;
  }
  if (method.getParameterCount() == 0) {
  if (logger.isInfoEnabled()) {
  logger.info("Autowired annotation should only be used on methods with parameters: " +
   method);
  }
  }
  boolean required = determineRequiredStatus(ann);
  PropertyDescriptor pd = BeanUtils.findPropertyForMethod(bridgedMethod, clazz);
  currElements.add(new AutowiredMethodElement(method, required, pd));
 }
 });

 elements.addAll(0, currElements);
 targetClass = targetClass.getSuperclass();
 }
 while (targetClass != null && targetClass != Object.class);

 return new InjectionMetadata(clazz, elements);
 }

收集的邏輯主要在findAutowiringMetadata方法中，層層調(diào)用后可以看到是通過(guò)findAutowiredAnnotation這個(gè)方法去找到帶有@Autowired和@Value注解的屬性和方法：

 private final Set<Class<? extends Annotation>> autowiredAnnotationTypes = new LinkedHashSet<>(4);

 public AutowiredAnnotationBeanPostProcessor() {
 this.autowiredAnnotationTypes.add(Autowired.class);
 this.autowiredAnnotationTypes.add(Value.class);
 try {
 this.autowiredAnnotationTypes.add((Class<? extends Annotation>)
  ClassUtils.forName("javax.inject.Inject", AutowiredAnnotationBeanPostProcessor.class.getClassLoader()));
 logger.trace("JSR-330 'javax.inject.Inject' annotation found and supported for autowiring");
 }
 catch (ClassNotFoundException ex) {
 // JSR-330 API not available - simply skip.
 }
 }


 private AnnotationAttributes findAutowiredAnnotation(AccessibleObject ao) {
 if (ao.getAnnotations().length > 0) { // autowiring annotations have to be local
 for (Class<? extends Annotation> type : this.autowiredAnnotationTypes) {
 AnnotationAttributes attributes = AnnotatedElementUtils.getMergedAnnotationAttributes(ao, type);
 if (attributes != null) {
  return attributes;
 }
 }
 }
 return null;
 }

最后將其封裝為AutowiredFieldElement和AutowiredMethodElement對(duì)象的list并連同Class一起封裝成InjectionMetadata返回，這就完成了相關(guān)注解的收集。

收集完成后在哪里使用呢？對(duì)Bean生命周期熟悉的讀者都知道，之后就會(huì)進(jìn)行依賴(lài)注入，自然相關(guān)的調(diào)用就在populateBean這個(gè)方法里：

protected void populateBean(String beanName, RootBeanDefinition mbd, @Nullable BeanWrapper bw) {
 
 ......

 PropertyValues pvs = (mbd.hasPropertyValues() ? mbd.getPropertyValues() : null);

 if (mbd.getResolvedAutowireMode() == AUTOWIRE_BY_NAME || mbd.getResolvedAutowireMode() == AUTOWIRE_BY_TYPE) {
 MutablePropertyValues newPvs = new MutablePropertyValues(pvs);
 // Add property values based on autowire by name if applicable.
 if (mbd.getResolvedAutowireMode() == AUTOWIRE_BY_NAME) {
 autowireByName(beanName, mbd, bw, newPvs);
 }
 // Add property values based on autowire by type if applicable.
 if (mbd.getResolvedAutowireMode() == AUTOWIRE_BY_TYPE) {
 autowireByType(beanName, mbd, bw, newPvs);
 }
 pvs = newPvs;
 }

 boolean hasInstAwareBpps = hasInstantiationAwareBeanPostProcessors();
 boolean needsDepCheck = (mbd.getDependencyCheck() != AbstractBeanDefinition.DEPENDENCY_CHECK_NONE);

 PropertyDescriptor[] filteredPds = null;

 //重點(diǎn)看這個(gè)if代碼塊，重要程度 5
 if (hasInstAwareBpps) {
 if (pvs == null) {
 pvs = mbd.getPropertyValues();
 }
 for (BeanPostProcessor bp : getBeanPostProcessors()) {
 if (bp instanceof InstantiationAwareBeanPostProcessor) {
  InstantiationAwareBeanPostProcessor ibp = (InstantiationAwareBeanPostProcessor) bp;
  // 依賴(lài)注入過(guò)程，@Autowired的支持
  PropertyValues pvsToUse = ibp.postProcessProperties(pvs, bw.getWrappedInstance(), beanName);
  if (pvsToUse == null) {
  if (filteredPds == null) {
  filteredPds = filterPropertyDescriptorsForDependencyCheck(bw, mbd.allowCaching);
  }

  // 老版本用這個(gè)完成依賴(lài)注入過(guò)程，@Autowired的支持
  pvsToUse = ibp.postProcessPropertyValues(pvs, filteredPds, bw.getWrappedInstance(), beanName);
  if (pvsToUse == null) {
  return;
  }
  }
  pvs = pvsToUse;
 }
 }
 }
 if (needsDepCheck) {
 if (filteredPds == null) {
 filteredPds = filterPropertyDescriptorsForDependencyCheck(bw, mbd.allowCaching);
 }
 checkDependencies(beanName, mbd, filteredPds, pvs);
 }

 // xml中<property>標(biāo)簽的依賴(lài)注入
 if (pvs != null) {
 applyPropertyValues(beanName, mbd, bw, pvs);
 }
 }

mbd.getResolvedAutowireMode() == AUTOWIRE_BY_NAME || mbd.getResolvedAutowireMode() == AUTOWIRE_BY_TYPE

很多讀者包括網(wǎng)上很多文章在看到這個(gè)判斷時(shí)，都認(rèn)為自動(dòng)注入的邏輯就是這里，但實(shí)際上并不是，這里是自動(dòng)注入沒(méi)錯(cuò)，但卻是針對(duì)以前xml配置，如下：

<?xml version="1.0" encoding="UTF-8"?> 
<beans xmlns="http://www.springframework.org/schema/beans" 
  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" 
  xsi:schemaLocation=" 
   http://www.springframework.org/schema/beans 
   http://www.springframework.org/schema/beans/spring-beans-2.5.xsd"
  default-autowire="byName">
 <bean id="a" class="com.A" autowire="byName"/>
 <bean id="b" class="com.B" autowire="byType"/>
</beans>

頭文件中的default-autowire屬性就是開(kāi)啟全局自動(dòng)注入，而bean標(biāo)簽上的autowire則是特定針對(duì)當(dāng)前bean的，會(huì)覆蓋全局配置。這樣我們配置后，bean標(biāo)簽就無(wú)需配置prototype子標(biāo)簽，也能自動(dòng)注入對(duì)應(yīng)的對(duì)象。這些屬性包含了以下幾個(gè)值：

no：默認(rèn)值。表示不進(jìn)行自動(dòng)注入。對(duì)應(yīng)BeanDefinition中autowireMode的值為 0。

byName：根據(jù)名稱(chēng)進(jìn)行自動(dòng)注入。對(duì)應(yīng)BeanDefinition中autowireMode的值為1。

byType：根據(jù)類(lèi)型進(jìn)行自動(dòng)注入，如果容器中找到兩個(gè)及以上符合該類(lèi)型的Bean就將拋出異常。對(duì)應(yīng)BeanDefinition中autowireMode的值為2。

constructor：等同于byType，只是當(dāng)指定autowire=”constructor”時(shí)表示將通過(guò)構(gòu)造方法根據(jù)類(lèi)型進(jìn)行自動(dòng)注入。對(duì)應(yīng)BeanDefinition中autowireMode的值為3。

這就是xml配置中的自動(dòng)注入，而我們使用@Autowired注解時(shí)，BeanDefinition中autowireMode的值為 0，即表示不進(jìn)行自動(dòng)注入。插一句題外話，網(wǎng)上很多人在爭(zhēng)論@Autowired是自動(dòng)注入還是手動(dòng)注入，我個(gè)人認(rèn)為都算自動(dòng)注入，不能說(shuō)它沒(méi)有進(jìn)入這段邏輯就不能叫自動(dòng)注入，只是它以另外一種方式實(shí)現(xiàn)了，至少也沒(méi)有讓我們自己手動(dòng)new并設(shè)置屬性。

那這另外一種方式是什么呢？就是下面這個(gè)代碼干的事：

if (hasInstAwareBpps) {
 if (pvs == null) {
 pvs = mbd.getPropertyValues();
 }
 for (BeanPostProcessor bp : getBeanPostProcessors()) {
 if (bp instanceof InstantiationAwareBeanPostProcessor) {
 InstantiationAwareBeanPostProcessor ibp = (InstantiationAwareBeanPostProcessor) bp;
 // 依賴(lài)注入過(guò)程，@Autowired的支持
 PropertyValues pvsToUse = ibp.postProcessProperties(pvs, bw.getWrappedInstance(), beanName);
 if (pvsToUse == null) {
 if (filteredPds == null) {
  filteredPds = filterPropertyDescriptorsForDependencyCheck(bw, mbd.allowCaching);
 }

 // 老版本用這個(gè)完成依賴(lài)注入過(guò)程，@Autowired的支持
 pvsToUse = ibp.postProcessPropertyValues(pvs, filteredPds, bw.getWrappedInstance(), beanName);
 if (pvsToUse == null) {
  return;
 }
 }
 pvs = pvsToUse;
 }
 }
}

可以看到這里又是調(diào)用了InstantiationAwareBeanPostProcessor類(lèi)型的postProcessProperties和postProcessPropertyValues方法，后者是老版本中的實(shí)現(xiàn)，已經(jīng)廢棄，所以直接看postProcessProperties，還是進(jìn)入到AutowiredAnnotationBeanPostProcessor類(lèi)中：

public PropertyValues postProcessProperties(PropertyValues pvs, Object bean, String beanName) {
 InjectionMetadata metadata = findAutowiringMetadata(beanName, bean.getClass(), pvs);
 try {
 metadata.inject(bean, beanName, pvs);
 }
 catch (BeanCreationException ex) {
 throw ex;
 }
 catch (Throwable ex) {
 throw new BeanCreationException(beanName, "Injection of autowired dependencies failed", ex);
 }
 return pvs;
 }

findAutowiringMetadata這個(gè)方法不陌生的，剛剛已經(jīng)分析了，是去收集對(duì)應(yīng)注解并封裝為InjectionMetadata對(duì)象放入到緩存，這里就是從緩存中拿到值，注入則是通過(guò)inject實(shí)現(xiàn)的：

public void inject(Object target, @Nullable String beanName, @Nullable PropertyValues pvs) throws Throwable {
 Collection<InjectedElement> checkedElements = this.checkedElements;
 Collection<InjectedElement> elementsToIterate =
 (checkedElements != null ? checkedElements : this.injectedElements);
 if (!elementsToIterate.isEmpty()) {
 for (InjectedElement element : elementsToIterate) {
 if (logger.isTraceEnabled()) {
 logger.trace("Processing injected element of bean '" + beanName + "': " + element);
 }
 if(element.isField) {
 Field field = (Field)element.member;
 System.out.println("==IOC/DI===beanName==" + beanName + "==field[" + field.getName() +"]-> getBean(" + field.getName() + ")");
 }
 element.inject(target, beanName, pvs);
 }
 }
}

最后就是調(diào)用element.inject實(shí)現(xiàn)注入，element我們剛剛也看到了，就是AutowiredFieldElement和AutowiredMethodElement，分別實(shí)現(xiàn)屬性注入和方法注入，這里我們看最常用的屬性注入就行了：

protected void inject(Object bean, @Nullable String beanName, @Nullable PropertyValues pvs) throws Throwable {
 Field field = (Field) this.member;
 Object value;
 if (this.cached) {
 value = resolvedCachedArgument(beanName, this.cachedFieldValue);
 }
 else {
 DependencyDescriptor desc = new DependencyDescriptor(field, this.required);
 desc.setContainingClass(bean.getClass());
 Set<String> autowiredBeanNames = new LinkedHashSet<>(1);
 Assert.state(beanFactory != null, "No BeanFactory available");
 TypeConverter typeConverter = beanFactory.getTypeConverter();
 try {
 // 找到依賴(lài)對(duì)象
 value = beanFactory.resolveDependency(desc, beanName, autowiredBeanNames, typeConverter);
 }
 catch (BeansException ex) {
 throw new UnsatisfiedDependencyException(null, beanName, new InjectionPoint(field), ex);
 }
 synchronized (this) {
 if (!this.cached) {
 if (value != null || this.required) {
  this.cachedFieldValue = desc;
  registerDependentBeans(beanName, autowiredBeanNames);
  if (autowiredBeanNames.size() == 1) {
  String autowiredBeanName = autowiredBeanNames.iterator().next();
  if (beanFactory.containsBean(autowiredBeanName) &&
  beanFactory.isTypeMatch(autowiredBeanName, field.getType())) {
  this.cachedFieldValue = new ShortcutDependencyDescriptor(
   desc, autowiredBeanName, field.getType());
  }
  }
 }
 else {
  this.cachedFieldValue = null;
 }
 this.cached = true;
 }
 }
 }

 if (value != null) {
 ReflectionUtils.makeAccessible(field);
 field.set(bean, value);
 }
}

這段代碼整體邏輯比較清晰，首先根據(jù)field創(chuàng)建一個(gè)依賴(lài)對(duì)象的抽象DependencyDescriptor對(duì)象，然后通過(guò)beanFactory.resolveDependency解析拿到對(duì)應(yīng)的實(shí)例，最后通過(guò)反射注入即可。

因此我們主要看resolveDependency方法中做了什么：

public Object resolveDependency(DependencyDescriptor descriptor, @Nullable String requestingBeanName,
 @Nullable Set<String> autowiredBeanNames, @Nullable TypeConverter typeConverter) throws BeansException {

 descriptor.initParameterNameDiscovery(getParameterNameDiscoverer());
 if (Optional.class == descriptor.getDependencyType()) {
 return createOptionalDependency(descriptor, requestingBeanName);
 }
 else if (ObjectFactory.class == descriptor.getDependencyType() ||
 ObjectProvider.class == descriptor.getDependencyType()) {
 return new DependencyObjectProvider(descriptor, requestingBeanName);
 }
 else if (javaxInjectProviderClass == descriptor.getDependencyType()) {
 return new Jsr330Factory().createDependencyProvider(descriptor, requestingBeanName);
 }
 else {
 Object result = getAutowireCandidateResolver().getLazyResolutionProxyIfNecessary(
 descriptor, requestingBeanName);
 if (result == null) {
 result = doResolveDependency(descriptor, requestingBeanName, autowiredBeanNames, typeConverter);
 }
 return result;
 }
}

一般情況下，都是走的else分支并調(diào)用doResolveDependency方法：

public Object doResolveDependency(DependencyDescriptor descriptor, @Nullable String beanName,
 @Nullable Set<String> autowiredBeanNames, @Nullable TypeConverter typeConverter) throws BeansException {

 InjectionPoint previousInjectionPoint = ConstructorResolver.setCurrentInjectionPoint(descriptor);
 try {
 Object shortcut = descriptor.resolveShortcut(this);
 if (shortcut != null) {
 return shortcut;
 }

 Class<?> type = descriptor.getDependencyType();
 Object value = getAutowireCandidateResolver().getSuggestedValue(descriptor);
 if (value != null) {
 if (value instanceof String) {
 String strVal = resolveEmbeddedValue((String) value);
 BeanDefinition bd = (beanName != null && containsBean(beanName) ? getMergedBeanDefinition(beanName) : null);
 value = evaluateBeanDefinitionString(strVal, bd);
 }
 TypeConverter converter = (typeConverter != null ? typeConverter : getTypeConverter());
 return (descriptor.getField() != null ?
  converter.convertIfNecessary(value, type, descriptor.getField()) :
  converter.convertIfNecessary(value, type, descriptor.getMethodParameter()));
 }

 // 有多個(gè)實(shí)現(xiàn)類(lèi)需要注入，特指注入的是數(shù)組、集合或者M(jìn)ap
 Object multipleBeans = resolveMultipleBeans(descriptor, beanName, autowiredBeanNames, typeConverter);
 if (multipleBeans != null) {
 return multipleBeans;
 }

 // 找到依賴(lài)對(duì)象的所有實(shí)現(xiàn)類(lèi)
 Map<String, Object> matchingBeans = findAutowireCandidates(beanName, type, descriptor);
 if (matchingBeans.isEmpty()) {
 if (isRequired(descriptor)) {
 raiseNoMatchingBeanFound(type, descriptor.getResolvableType(), descriptor);
 }
 return null;
 }

 String autowiredBeanName;
 Object instanceCandidate;

 // 依賴(lài)的對(duì)象有多個(gè)實(shí)例
 if (matchingBeans.size() > 1) {
 // 根據(jù)@Primary、@Priority和名稱(chēng)依次進(jìn)行匹配注入
 autowiredBeanName = determineAutowireCandidate(matchingBeans, descriptor);
 if (autowiredBeanName == null) {
 if (isRequired(descriptor) || !indicatesMultipleBeans(type)) {
  return descriptor.resolveNotUnique(descriptor.getResolvableType(), matchingBeans);
 }
 else {
  // In case of an optional Collection/Map, silently ignore a non-unique case:
  // possibly it was meant to be an empty collection of multiple regular beans
  // (before 4.3 in particular when we didn't even look for collection beans).
  return null;
 }
 }
 instanceCandidate = matchingBeans.get(autowiredBeanName);
 }
 else {
 // We have exactly one match.
 Map.Entry<String, Object> entry = matchingBeans.entrySet().iterator().next();
 autowiredBeanName = entry.getKey();
 instanceCandidate = entry.getValue();
 }

 if (autowiredBeanNames != null) {
 autowiredBeanNames.add(autowiredBeanName);
 }
 // 如果拿到的是class對(duì)象，通過(guò)getBean實(shí)例化返回
 if (instanceCandidate instanceof Class) {
 instanceCandidate = descriptor.resolveCandidate(autowiredBeanName, type, this);
 }
 Object result = instanceCandidate;
 if (result instanceof NullBean) {
 if (isRequired(descriptor)) {
 raiseNoMatchingBeanFound(type, descriptor.getResolvableType(), descriptor);
 }
 result = null;
 }
 if (!ClassUtils.isAssignableValue(type, result)) {
 throw new BeanNotOfRequiredTypeException(autowiredBeanName, type, instanceCandidate.getClass());
 }
 return result;
 }
 finally {
 ConstructorResolver.setCurrentInjectionPoint(previousInjectionPoint);
 }
}

這里面首先是通過(guò) getAutowireCandidateResolver().getSuggestedValue(descriptor)拿到@Value注解的值，然后通過(guò)TypeConverter進(jìn)行轉(zhuǎn)換，默認(rèn)可轉(zhuǎn)換的類(lèi)型是JDK和Spring內(nèi)置的一些類(lèi)型，自然不包含我們自定義的類(lèi)，所以如果不進(jìn)行擴(kuò)展在@Autowired注入我們自定義類(lèi)對(duì)象時(shí)同時(shí)使用@Value注解是會(huì)報(bào)錯(cuò)的。

接著是調(diào)用resolveMultipleBeans方法實(shí)現(xiàn)對(duì)Map、List、數(shù)組等屬性的注入，本質(zhì)上還是調(diào)用findAutowireCandidates方法找到所有的實(shí)現(xiàn)類(lèi)的對(duì)象裝入對(duì)應(yīng)的集合數(shù)組中，所以直接看findAutowireCandidates：

protected Map<String, Object> findAutowireCandidates(
 @Nullable String beanName, Class<?> requiredType, DependencyDescriptor descriptor) {

 String[] candidateNames = BeanFactoryUtils.beanNamesForTypeIncludingAncestors(
 this, requiredType, true, descriptor.isEager());
 Map<String, Object> result = new LinkedHashMap<>(candidateNames.length);
 
 ......
 
 for (String candidate : candidateNames) {
 if (!isSelfReference(beanName, candidate) && isAutowireCandidate(candidate, descriptor)) {
 addCandidateEntry(result, candidate, descriptor, requiredType);
 }
 }

 ......
 return result;
}

private void addCandidateEntry(Map<String, Object> candidates, String candidateName,
 DependencyDescriptor descriptor, Class<?> requiredType) {

 if (descriptor instanceof MultiElementDescriptor) {
 Object beanInstance = descriptor.resolveCandidate(candidateName, requiredType, this);
 if (!(beanInstance instanceof NullBean)) {
 candidates.put(candidateName, beanInstance);
 }
 }
 else if (containsSingleton(candidateName) || (descriptor instanceof StreamDependencyDescriptor &&
 ((StreamDependencyDescriptor) descriptor).isOrdered())) {
 Object beanInstance = descriptor.resolveCandidate(candidateName, requiredType, this);
 candidates.put(candidateName, (beanInstance instanceof NullBean ? null : beanInstance));
 }
 else {
 candidates.put(candidateName, getType(candidateName));
 }
}

首先通過(guò)beanNamesForTypeIncludingAncestors方法拿到依賴(lài)類(lèi)所有的實(shí)現(xiàn)類(lèi)的beanName，然后調(diào)用addCandidateEntry將beanName及對(duì)應(yīng)的實(shí)例或者Class對(duì)象放入到Map中。

接著回到doResolveDependency方法中：

if (matchingBeans.size() > 1) {
 // 根據(jù)@Primary、@Priority和名稱(chēng)依次進(jìn)行匹配注入
 autowiredBeanName = determineAutowireCandidate(matchingBeans, descriptor);
 if (autowiredBeanName == null) {
 if (isRequired(descriptor) || !indicatesMultipleBeans(type)) {
 return descriptor.resolveNotUnique(descriptor.getResolvableType(), matchingBeans);
 }
 else {
 // In case of an optional Collection/Map, silently ignore a non-unique case:
 // possibly it was meant to be an empty collection of multiple regular beans
 // (before 4.3 in particular when we didn't even look for collection beans).
 return null;
 }
 }
 instanceCandidate = matchingBeans.get(autowiredBeanName);
}
else {
 // We have exactly one match.
 Map.Entry<String, Object> entry = matchingBeans.entrySet().iterator().next();
 autowiredBeanName = entry.getKey();
 instanceCandidate = entry.getValue();
}

if (autowiredBeanNames != null) {
 autowiredBeanNames.add(autowiredBeanName);
}
// 如果拿到的是class對(duì)象，通過(guò)getBean實(shí)例化返回
if (instanceCandidate instanceof Class) {
 instanceCandidate = descriptor.resolveCandidate(autowiredBeanName, type, this);
}
Object result = instanceCandidate;

如果只有一個(gè)實(shí)例則直接返回該實(shí)例，如果實(shí)現(xiàn)類(lèi)有多個(gè)則調(diào)用determineAutowireCandidate進(jìn)行判斷該使用哪一個(gè)實(shí)例對(duì)象，判斷規(guī)則如下：

首先判斷實(shí)現(xiàn)類(lèi)上是否有@Primary注解，找到一個(gè)則返回當(dāng)前實(shí)例，找到多個(gè)則報(bào)錯(cuò)。

若沒(méi)有標(biāo)注@Primary注解，則判斷是否指定了優(yōu)先級(jí)，且只能是通過(guò)@Priority注解指定的，@Order不支持。

上述都沒(méi)有拿到合適的Bean則通過(guò)屬性名稱(chēng)獲取Bean。

拿到對(duì)應(yīng)的實(shí)例后，最后進(jìn)行反射注入即可。以上就是@Autowired的實(shí)現(xiàn)細(xì)節(jié)。

總結(jié)

本篇從源碼角度詳細(xì)分析了@Autowired的實(shí)現(xiàn)細(xì)節(jié)，只有真正閱讀了源碼才能了解更多的細(xì)節(jié)，在開(kāi)發(fā)中更加清楚如何注入多個(gè)實(shí)例以及如何指定注入的優(yōu)先級(jí)，同時(shí)在面試中也能更有理有據(jù)，而不是統(tǒng)一的大眾回答，先根據(jù)byType，再根據(jù)byName。另外對(duì)于方法注入和@Resource注解的處理本篇沒(méi)有涉及，但是相信看完本文讀者自己也能輕松分析這部分源碼。

您可能感興趣的文章: