Spring 核心之 IOC 容器学习二-拓冰建站

基于 Annotation 的 IOC 初始化

Annotation 的前世今生

从 Spring2.0 以后的版本中，Spring 也引入了基于注解(Annotation)方式的配置，注解(Annotation)是 JDK1.5 中引入的一个新特性，用于简化 Bean 的配置，可以取代 XML 配置文件。开发人员对注解(Annotation)的态度也是萝卜青菜各有所爱，个人认为注解可以大大简化配置，提高开发速度，但也给后期维护增加了难度。目前来说 XML 方式发展的相对成熟，方便于统一管理。随着 Spring Boot 的兴起，基于注解的开发甚至实现了零配置。但作为个人的习惯而言，还是倾向于 XML 配置文件和注解(Annotation)相互配合使用。Spring IOC 容器对于类级别的注解和类内部的注解分以下两种处理策略：
1)、类级别的注解：如@Component、@Repository、@Controller、@Service 以及 JavaEE6 的@ManagedBean 和@Named 注解，都是添加在类上面的类级别注解，Spring 容器根据注解的过滤规则扫描读取注解 Bean 定义类，并将其注册到 Spring IOC 容器中。
2)、类内部的注解：如@Autowire、@Value、@Resource 以及 EJB 和 WebService 相关的注解等，都是添加在类内部的字段或者方法上的类内部注解，SpringIOC 容器通过 Bean 后置注解处理器解析Bean 内部的注解。下面将根据这两种处理策略，分别分析 Spring 处理注解相关的源码。

定位 Bean 扫描路径

在 Spring 中 管 理 注 解 Bean 定 义 的 容 器 有 两 个 ： AnnotationConfigApplicationContext 和AnnotationConfigWebApplicationContex。这两个类是专门处理 Spring 注解方式配置的容器，直接依赖于注解作为容器配置信息来源的 IOC 容器。AnnotationConfigWebApplicationContext 是AnnotationConfigApplicationContext 的 Web 版本，两者的用法以及对注解的处理方式几乎没有差别。现在我们以 AnnotationConfigApplicationContext 为例看看它的源码：

public class AnnotationConfigApplicationContext extends GenericApplicationContext implements AnnotationConfigRegistry {//保存一个读取注解的Bean定义读取器，并将其设置到容器中private final AnnotatedBeanDefinitionReader reader;//保存一个扫描指定类路径中注解Bean定义的扫描器，并将其设置到容器中private final ClassPathBeanDefinitionScanner scanner;/*** Create a new AnnotationConfigApplicationContext that needs to be populated* through {@link #register} calls and then manually {@linkplain #refresh refreshed}.*///默认构造函数，初始化一个空容器，容器不包含任何 Bean 信息，需要在稍后通过调用其register()//方法注册配置类，并调用refresh()方法刷新容器，触发容器对注解Bean的载入、解析和注册过程public AnnotationConfigApplicationContext() {this.reader = new AnnotatedBeanDefinitionReader(this);this.scanner = new ClassPathBeanDefinitionScanner(this);}/*** Create a new AnnotationConfigApplicationContext with the given DefaultListableBeanFactory.* @param beanFactory the DefaultListableBeanFactory instance to use for this context*/public AnnotationConfigApplicationContext(DefaultListableBeanFactory beanFactory) {super(beanFactory);this.reader = new AnnotatedBeanDefinitionReader(this);this.scanner = new ClassPathBeanDefinitionScanner(this);}/*** Create a new AnnotationConfigApplicationContext, deriving bean definitions* from the given annotated classes and automatically refreshing the context.* @param annotatedClasses one or more annotated classes,* e.g. {@link Configuration @Configuration} classes*///最常用的构造函数，通过将涉及到的配置类传递给该构造函数，以实现将相应配置类中的Bean自动注册到容器中public AnnotationConfigApplicationContext(Class<?>... annotatedClasses) {this();register(annotatedClasses);refresh();}/*** Create a new AnnotationConfigApplicationContext, scanning for bean definitions* in the given packages and automatically refreshing the context.* @param basePackages the packages to check for annotated classes*///该构造函数会自动扫描以给定的包及其子包下的所有类，并自动识别所有的Spring Bean，将其注册到容器中public AnnotationConfigApplicationContext(String... basePackages) {this();scan(basePackages);refresh();}/*** {@inheritDoc}* <p>Delegates given environment to underlying {@link AnnotatedBeanDefinitionReader}* and {@link ClassPathBeanDefinitionScanner} members.*/@Overridepublic void setEnvironment(ConfigurableEnvironment environment) {super.setEnvironment(environment);this.reader.setEnvironment(environment);this.scanner.setEnvironment(environment);}/*** Provide a custom {@link BeanNameGenerator} for use with {@link AnnotatedBeanDefinitionReader}* and/or {@link ClassPathBeanDefinitionScanner}, if any.* <p>Default is {@link org.springframework.context.annotation.AnnotationBeanNameGenerator}.* <p>Any call to this method must occur prior to calls to {@link #register(Class...)}* and/or {@link #scan(String...)}.* @see AnnotatedBeanDefinitionReader#setBeanNameGenerator* @see ClassPathBeanDefinitionScanner#setBeanNameGenerator*///为容器的注解Bean读取器和注解Bean扫描器设置Bean名称产生器public void setBeanNameGenerator(BeanNameGenerator beanNameGenerator) {this.reader.setBeanNameGenerator(beanNameGenerator);this.scanner.setBeanNameGenerator(beanNameGenerator);getBeanFactory().registerSingleton(AnnotationConfigUtils.CONFIGURATION_BEAN_NAME_GENERATOR, beanNameGenerator);}/*** Set the {@link ScopeMetadataResolver} to use for detected bean classes.* <p>The default is an {@link AnnotationScopeMetadataResolver}.* <p>Any call to this method must occur prior to calls to {@link #register(Class...)}* and/or {@link #scan(String...)}.*///为容器的注解Bean读取器和注解Bean扫描器设置作用范围元信息解析器public void setScopeMetadataResolver(ScopeMetadataResolver scopeMetadataResolver) {this.reader.setScopeMetadataResolver(scopeMetadataResolver);this.scanner.setScopeMetadataResolver(scopeMetadataResolver);}//---------------------------------------------------------------------// Implementation of AnnotationConfigRegistry//---------------------------------------------------------------------/*** Register one or more annotated classes to be processed.* <p>Note that {@link #refresh()} must be called in order for the context* to fully process the new classes.* @param annotatedClasses one or more annotated classes,* e.g. {@link Configuration @Configuration} classes* @see #scan(String...)* @see #refresh()*///为容器注册一个要被处理的注解Bean，新注册的Bean，必须手动调用容器的//refresh()方法刷新容器，触发容器对新注册的Bean的处理public void register(Class<?>... annotatedClasses) {Assert.notEmpty(annotatedClasses, "At least one annotated class must be specified");this.reader.register(annotatedClasses);}/*** Perform a scan within the specified base packages.* <p>Note that {@link #refresh()} must be called in order for the context* to fully process the new classes.* @param basePackages the packages to check for annotated classes* @see #register(Class...)* @see #refresh()*///扫描指定包路径及其子包下的注解类，为了使新添加的类被处理，必须手动调用//refresh()方法刷新容器public void scan(String... basePackages) {Assert.notEmpty(basePackages, "At least one base package must be specified");this.scanner.scan(basePackages);}//---------------------------------------------------------------------// Convenient methods for registering individual beans//---------------------------------------------------------------------/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations, and optionally providing explicit constructor* arguments for consideration in the autowiring process.* <p>The bean name will be generated according to annotated component rules.* @param annotatedClass the class of the bean* @param constructorArguments argument values to be fed into Spring's* constructor resolution algorithm, resolving either all arguments or just* specific ones, with the rest to be resolved through regular autowiring* (may be {@code null} or empty)* @since 5.0*/public <T> void registerBean(Class<T> annotatedClass, Object... constructorArguments) {registerBean(null, annotatedClass, constructorArguments);}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations, and optionally providing explicit constructor* arguments for consideration in the autowiring process.* @param beanName the name of the bean (may be {@code null})* @param annotatedClass the class of the bean* @param constructorArguments argument values to be fed into Spring's* constructor resolution algorithm, resolving either all arguments or just* specific ones, with the rest to be resolved through regular autowiring* (may be {@code null} or empty)* @since 5.0*/public <T> void registerBean(@Nullable String beanName, Class<T> annotatedClass, Object... constructorArguments) {this.reader.doRegisterBean(annotatedClass, null, beanName, null,bd -> {for (Object arg : constructorArguments) {bd.getConstructorArgumentValues().addGenericArgumentValue(arg);}});}@Overridepublic <T> void registerBean(@Nullable String beanName, Class<T> beanClass, @Nullable Supplier<T> supplier,BeanDefinitionCustomizer... customizers) {this.reader.doRegisterBean(beanClass, supplier, beanName, null, customizers);}}

通过上面的源码分析，我们可以看啊到 Spring 对注解的处理分为两种方式：
1)、直接将注解 Bean 注册到容器中
可以在初始化容器时注册；也可以在容器创建之后手动调用注册方法向容器注册，然后通过手动刷新容器，使得容器对注册的注解 Bean 进行处理。
2)、通过扫描指定的包及其子包下的所有类
在初始化注解容器时指定要自动扫描的路径，如果容器创建以后向给定路径动态添加了注解 Bean，则需要手动调用容器扫描的方法，然后手动刷新容器，使得容器对所注册的 Bean 进行处理。
接下来，将会对两种处理方式详细分析其实现过程。

读取 Annotation 元数据

当创建注解处理容器时，如果传入的初始参数是具体的注解 Bean 定义类时，注解容器读取并注册。
1)、AnnotationConfigApplicationContext 通过调用注解 Bean 定义读取器
AnnotatedBeanDefinitionReader 的 register()方法向容器注册指定的注解 Bean，注解 Bean 定义读取器向容器注册注解 Bean 的源码如下：

/*** Register one or more annotated classes to be processed.* <p>Calls to {@code register} are idempotent; adding the same* annotated class more than once has no additional effect.* @param annotatedClasses one or more annotated classes,* e.g. {@link Configuration @Configuration} classes*///注册多个注解Bean定义类public void register(Class<?>... annotatedClasses) {for (Class<?> annotatedClass : annotatedClasses) {registerBean(annotatedClass);}}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations.* @param annotatedClass the class of the bean*///注册一个注解Bean定义类public void registerBean(Class<?> annotatedClass) {doRegisterBean(annotatedClass, null, null, null);}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations, using the given supplier for obtaining a new* instance (possibly declared as a lambda expression or method reference).* @param annotatedClass the class of the bean* @param instanceSupplier a callback for creating an instance of the bean* (may be {@code null})* @since 5.0*/public <T> void registerBean(Class<T> annotatedClass, @Nullable Supplier<T> instanceSupplier) {doRegisterBean(annotatedClass, instanceSupplier, null, null);}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations, using the given supplier for obtaining a new* instance (possibly declared as a lambda expression or method reference).* @param annotatedClass the class of the bean* @param name an explicit name for the bean* @param instanceSupplier a callback for creating an instance of the bean* (may be {@code null})* @since 5.0*/public <T> void registerBean(Class<T> annotatedClass, String name, @Nullable Supplier<T> instanceSupplier) {doRegisterBean(annotatedClass, instanceSupplier, name, null);}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations.* @param annotatedClass the class of the bean* @param qualifiers specific qualifier annotations to consider,* in addition to qualifiers at the bean class level*///Bean定义读取器注册注解Bean定义的入口方法@SuppressWarnings("unchecked")public void registerBean(Class<?> annotatedClass, Class<? extends Annotation>... qualifiers) {doRegisterBean(annotatedClass, null, null, qualifiers);}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations.* @param annotatedClass the class of the bean* @param name an explicit name for the bean* @param qualifiers specific qualifier annotations to consider,* in addition to qualifiers at the bean class level*///Bean定义读取器向容器注册注解Bean定义类@SuppressWarnings("unchecked")public void registerBean(Class<?> annotatedClass, String name, Class<? extends Annotation>... qualifiers) {doRegisterBean(annotatedClass, null, name, qualifiers);}/*** Register a bean from the given bean class, deriving its metadata from* class-declared annotations.* @param annotatedClass the class of the bean* @param instanceSupplier a callback for creating an instance of the bean* (may be {@code null})* @param name an explicit name for the bean* @param qualifiers specific qualifier annotations to consider, if any,* in addition to qualifiers at the bean class level* @param definitionCustomizers one or more callbacks for customizing the* factory's {@link BeanDefinition}, e.g. setting a lazy-init or primary flag* @since 5.0*///Bean定义读取器向容器注册注解Bean定义类<T> void doRegisterBean(Class<T> annotatedClass, @Nullable Supplier<T> instanceSupplier, @Nullable String name,@Nullable Class<? extends Annotation>[] qualifiers, BeanDefinitionCustomizer... definitionCustomizers) {//根据指定的注解Bean定义类，创建Spring容器中对注解Bean的封装的数据结构AnnotatedGenericBeanDefinition abd = new AnnotatedGenericBeanDefinition(annotatedClass);if (this.conditionEvaluator.shouldSkip(abd.getMetadata())) {return;}abd.setInstanceSupplier(instanceSupplier);//解析注解Bean定义的作用域，若@Scope("prototype")，则Bean为原型类型；//若@Scope("singleton")，则Bean为单态类型ScopeMetadata scopeMetadata = this.scopeMetadataResolver.resolveScopeMetadata(abd);//为注解Bean定义设置作用域abd.setScope(scopeMetadata.getScopeName());//为注解Bean定义生成Bean名称String beanName = (name != null ? name : this.beanNameGenerator.generateBeanName(abd, this.registry));//处理注解Bean定义中的通用注解AnnotationConfigUtils.processCommonDefinitionAnnotations(abd);//如果在向容器注册注解Bean定义时，使用了额外的限定符注解，则解析限定符注解。//主要是配置的关于autowiring自动依赖注入装配的限定条件，即@Qualifier注解//Spring自动依赖注入装配默认是按类型装配，如果使用@Qualifier则按名称if (qualifiers != null) {for (Class<? extends Annotation> qualifier : qualifiers) {//如果配置了@Primary注解，设置该Bean为autowiring自动依赖注入装//配时的首选if (Primary.class == qualifier) {abd.setPrimary(true);}//如果配置了@Lazy注解，则设置该Bean为非延迟初始化，如果没有配置，//则该Bean为预实例化else if (Lazy.class == qualifier) {abd.setLazyInit(true);}//如果使用了除@Primary和@Lazy以外的其他注解，则为该Bean添加一//个autowiring自动依赖注入装配限定符，该Bean在进autowiring//自动依赖注入装配时，根据名称装配限定符指定的Beanelse {abd.addQualifier(new AutowireCandidateQualifier(qualifier));}}}for (BeanDefinitionCustomizer customizer : definitionCustomizers) {customizer.customize(abd);}//创建一个指定Bean名称的Bean定义对象，封装注解Bean定义类数据BeanDefinitionHolder definitionHolder = new BeanDefinitionHolder(abd, beanName);//根据注解Bean定义类中配置的作用域，创建相应的代理对象definitionHolder = AnnotationConfigUtils.applyScopedProxyMode(scopeMetadata, definitionHolder, this.registry);//向IOC容器注册注解Bean类定义对象BeanDefinitionReaderUtils.registerBeanDefinition(definitionHolder, this.registry);}

从上面的源码我们可以看出，注册注解 Bean 定义类的基本步骤：
a、需要使用注解元数据解析器解析注解 Bean 中关于作用域的配置
b、使用 AnnotationConfigUtils 的 processCommonDefinitionAnnotations()方法处理注解 Bean 定义类中通用的注解。
c、使用 AnnotationConfigUtils 的 applyScopedProxyMode()方法创建对于作用域的代理对象。
d、通过 BeanDefinitionReaderUtils 向容器注册 Bean。
2)、AnnotationScopeMetadataResolver 解析作用域元数据
AnnotationScopeMetadataResolver 通过 resolveScopeMetadata()方法解析注解 Bean 定义类的作用域元信息，即判断注册的 Bean 是原生类型(prototype)还是单态(singleton)类型，其源码如下：

//解析注解Bean定义类中的作用域元信息@Overridepublic ScopeMetadata resolveScopeMetadata(BeanDefinition definition) {ScopeMetadata metadata = new ScopeMetadata();if (definition instanceof AnnotatedBeanDefinition) {AnnotatedBeanDefinition annDef = (AnnotatedBeanDefinition) definition;//从注解Bean定义类的属性中查找属性为”Scope”的值，即@Scope注解的值//annDef.getMetadata().getAnnotationAttributes()方法将Bean//中所有的注解和注解的值存放在一个map集合中AnnotationAttributes attributes = AnnotationConfigUtils.attributesFor(annDef.getMetadata(), this.scopeAnnotationType);//将获取到的@Scope注解的值设置到要返回的对象中if (attributes != null) {metadata.setScopeName(attributes.getString("value"));//获取@Scope注解中的proxyMode属性值，在创建代理对象时会用到ScopedProxyMode proxyMode = attributes.getEnum("proxyMode");//如果@Scope的proxyMode属性为DEFAULT或者NOif (proxyMode == ScopedProxyMode.DEFAULT) {//设置proxyMode为NOproxyMode = this.defaultProxyMode;}//为返回的元数据设置proxyModemetadata.setScopedProxyMode(proxyMode);}}//返回解析的作用域元信息对象return metadata;}

上述代码中的 annDef.getMetadata().getAnnotationAttributes()方法就是获取对象中指定类型的注解的值。
3)、AnnotationConfigUtils 处理注解 Bean 定义类中的通用注解
AnnotationConfigUtils 类的 processCommonDefinitionAnnotations()在向容器注册 Bean 之前，首先对注解 Bean 定义类中的通用 Spring 注解进行处理，源码如下：

//处理Bean定义中通用注解static void processCommonDefinitionAnnotations(AnnotatedBeanDefinition abd, AnnotatedTypeMetadata metadata) {AnnotationAttributes lazy = attributesFor(metadata, Lazy.class);//如果Bean定义中有@Lazy注解，则将该Bean预实例化属性设置为@lazy注解的值if (lazy != null) {abd.setLazyInit(lazy.getBoolean("value"));}else if (abd.getMetadata() != metadata) {lazy = attributesFor(abd.getMetadata(), Lazy.class);if (lazy != null) {abd.setLazyInit(lazy.getBoolean("value"));}}//如果Bean定义中有@Primary注解，则为该Bean设置为autowiring自动依赖注入装配的首选对象if (metadata.isAnnotated(Primary.class.getName())) {abd.setPrimary(true);}//如果Bean定义中有@ DependsOn注解，则为该Bean设置所依赖的Bean名称，//容器将确保在实例化该Bean之前首先实例化所依赖的BeanAnnotationAttributes dependsOn = attributesFor(metadata, DependsOn.class);if (dependsOn != null) {abd.setDependsOn(dependsOn.getStringArray("value"));}if (abd instanceof AbstractBeanDefinition) {AbstractBeanDefinition absBd = (AbstractBeanDefinition) abd;AnnotationAttributes role = attributesFor(metadata, Role.class);if (role != null) {absBd.setRole(role.getNumber("value").intValue());}AnnotationAttributes description = attributesFor(metadata, Description.class);if (description != null) {absBd.setDescription(description.getString("value"));}}}

4)、AnnotationConfigUtils 根据注解 Bean 定义类中配置的作用域为其应用相应的代理策略
AnnotationConfigUtils 类的 applyScopedProxyMode()方法根据注解 Bean 定义类中配置的作用域@Scope 注解的值，为 Bean 定义应用相应的代理模式，主要是在 Spring 面向切面编程(AOP)中使用。源码如下：

//根据作用域为Bean应用引用的代码模式static BeanDefinitionHolder applyScopedProxyMode(ScopeMetadata metadata, BeanDefinitionHolder definition, BeanDefinitionRegistry registry) {//获取注解Bean定义类中@Scope注解的proxyMode属性值ScopedProxyMode scopedProxyMode = metadata.getScopedProxyMode();//如果配置的@Scope注解的proxyMode属性值为NO，则不应用代理模式if (scopedProxyMode.equals(ScopedProxyMode.NO)) {return definition;}//获取配置的@Scope注解的proxyMode属性值，如果为TARGET_CLASS//则返回true，如果为INTERFACES，则返回falseboolean proxyTargetClass = scopedProxyMode.equals(ScopedProxyMode.TARGET_CLASS);//为注册的Bean创建相应模式的代理对象return ScopedProxyCreator.createScopedProxy(definition, registry, proxyTargetClass);}

这段为 Bean 引用创建相应模式的代理，这里不做深入的分析。
5)、BeanDefinitionReaderUtils 向容器注册 Bean
BeanDefinitionReaderUtils 主要是校验 BeanDefinition 信息，然后将 Bean 添加到容器中一个管理BeanDefinition 的 HashMap 中。

扫描指定包并解析为 BeanDefinition

当创建注解处理容器时，如果传入的初始参数是注解 Bean 定义类所在的包时，注解容器将扫描给定的包及其子包，将扫描到的注解 Bean 定义载入并注册。
1)、ClassPathBeanDefinitionScanner 扫描给定的包及其子包
AnnotationConfigApplicationContext 通 过 调 用 类 路 径 Bean 定 义 扫 描 器ClassPathBeanDefinitionScanner 扫描给定包及其子包下的所有类，主要源码如下：

public class ClassPathBeanDefinitionScanner extends ClassPathScanningCandidateComponentProvider {/*** Create a new {@code ClassPathBeanDefinitionScanner} for the given bean factory.* @param registry the {@code BeanFactory} to load bean definitions into, in the form* of a {@code BeanDefinitionRegistry}*///创建一个类路径Bean定义扫描器public ClassPathBeanDefinitionScanner(BeanDefinitionRegistry registry) {this(registry, true);}
/*** Create a new {@code ClassPathBeanDefinitionScanner} for the given bean factory.* <p>If the passed-in bean factory does not only implement the* {@code BeanDefinitionRegistry} interface but also the {@code ResourceLoader}* interface, it will be used as default {@code ResourceLoader} as well. This will* usually be the case for {@link org.springframework.context.ApplicationContext}* implementations.* <p>If given a plain {@code BeanDefinitionRegistry}, the default {@code ResourceLoader}* will be a {@link org.springframework.core.io.support.PathMatchingResourcePatternResolver}.* <p>If the passed-in bean factory also implements {@link EnvironmentCapable} its* environment will be used by this reader.  Otherwise, the reader will initialize and* use a {@link org.springframework.core.env.StandardEnvironment}. All* {@code ApplicationContext} implementations are {@code EnvironmentCapable}, while* normal {@code BeanFactory} implementations are not.* @param registry the {@code BeanFactory} to load bean definitions into, in the form* of a {@code BeanDefinitionRegistry}* @param useDefaultFilters whether to include the default filters for the* {@link org.springframework.stereotype.Component @Component},* {@link org.springframework.stereotype.Repository @Repository},* {@link org.springframework.stereotype.Service @Service}, and* {@link org.springframework.stereotype.Controller @Controller} stereotype annotations* @see #setResourceLoader* @see #setEnvironment*///为容器创建一个类路径Bean定义扫描器，并指定是否使用默认的扫描过滤规则。//即Spring默认扫描配置：@Component、@Repository、@Service、@Controller//注解的Bean，同时也支持JavaEE6的@ManagedBean和JSR-330的@Named注解public ClassPathBeanDefinitionScanner(BeanDefinitionRegistry registry, boolean useDefaultFilters) {this(registry, useDefaultFilters, getOrCreateEnvironment(registry));}/*** Create a new {@code ClassPathBeanDefinitionScanner} for the given bean factory and* using the given {@link Environment} when evaluating bean definition profile metadata.* <p>If the passed-in bean factory does not only implement the {@code* BeanDefinitionRegistry} interface but also the {@link ResourceLoader} interface, it* will be used as default {@code ResourceLoader} as well. This will usually be the* case for {@link org.springframework.context.ApplicationContext} implementations.* <p>If given a plain {@code BeanDefinitionRegistry}, the default {@code ResourceLoader}* will be a {@link org.springframework.core.io.support.PathMatchingResourcePatternResolver}.* @param registry the {@code BeanFactory} to load bean definitions into, in the form* of a {@code BeanDefinitionRegistry}* @param useDefaultFilters whether to include the default filters for the* {@link org.springframework.stereotype.Component @Component},* {@link org.springframework.stereotype.Repository @Repository},* {@link org.springframework.stereotype.Service @Service}, and* {@link org.springframework.stereotype.Controller @Controller} stereotype annotations* @param environment the Spring {@link Environment} to use when evaluating bean* definition profile metadata* @since 3.1* @see #setResourceLoader*/public ClassPathBeanDefinitionScanner(BeanDefinitionRegistry registry, boolean useDefaultFilters,Environment environment) {this(registry, useDefaultFilters, environment,(registry instanceof ResourceLoader ? (ResourceLoader) registry : null));}/*** Create a new {@code ClassPathBeanDefinitionScanner} for the given bean factory and* using the given {@link Environment} when evaluating bean definition profile metadata.* @param registry the {@code BeanFactory} to load bean definitions into, in the form* of a {@code BeanDefinitionRegistry}* @param useDefaultFilters whether to include the default filters for the* {@link org.springframework.stereotype.Component @Component},* {@link org.springframework.stereotype.Repository @Repository},* {@link org.springframework.stereotype.Service @Service}, and* {@link org.springframework.stereotype.Controller @Controller} stereotype annotations* @param environment the Spring {@link Environment} to use when evaluating bean* definition profile metadata* @param resourceLoader the {@link ResourceLoader} to use* @since 4.3.6*/public ClassPathBeanDefinitionScanner(BeanDefinitionRegistry registry, boolean useDefaultFilters,Environment environment, @Nullable ResourceLoader resourceLoader) {Assert.notNull(registry, "BeanDefinitionRegistry must not be null");//为容器设置加载Bean定义的注册器this.registry = registry;if (useDefaultFilters) {registerDefaultFilters();}setEnvironment(environment);//为容器设置资源加载器setResourceLoader(resourceLoader);}
/*** Perform a scan within the specified base packages.* @param basePackages the packages to check for annotated classes* @return number of beans registered*///调用类路径Bean定义扫描器入口方法public int scan(String... basePackages) {//获取容器中已经注册的Bean个数int beanCountAtScanStart = this.registry.getBeanDefinitionCount();//启动扫描器扫描给定包doScan(basePackages);// Register annotation config processors, if necessary.//注册注解配置(Annotation config)处理器if (this.includeAnnotationConfig) {AnnotationConfigUtils.registerAnnotationConfigProcessors(this.registry);}//返回注册的Bean个数return (this.registry.getBeanDefinitionCount() - beanCountAtScanStart);}/*** Perform a scan within the specified base packages,* returning the registered bean definitions.* <p>This method does <i>not</i> register an annotation config processor* but rather leaves this up to the caller.* @param basePackages the packages to check for annotated classes* @return set of beans registered if any for tooling registration purposes (never {@code null})*///类路径Bean定义扫描器扫描给定包及其子包protected Set<BeanDefinitionHolder> doScan(String... basePackages) {Assert.notEmpty(basePackages, "At least one base package must be specified");//创建一个集合，存放扫描到Bean定义的封装类Set<BeanDefinitionHolder> beanDefinitions = new LinkedHashSet<>();//遍历扫描所有给定的包for (String basePackage : basePackages) {//调用父类ClassPathScanningCandidateComponentProvider的方法//扫描给定类路径，获取符合条件的Bean定义Set<BeanDefinition> candidates = findCandidateComponents(basePackage);//遍历扫描到的Beanfor (BeanDefinition candidate : candidates) {//获取Bean定义类中@Scope注解的值，即获取Bean的作用域ScopeMetadata scopeMetadata = this.scopeMetadataResolver.resolveScopeMetadata(candidate);//为Bean设置注解配置的作用域candidate.setScope(scopeMetadata.getScopeName());//为Bean生成名称String beanName = this.beanNameGenerator.generateBeanName(candidate, this.registry);//如果扫描到的Bean不是Spring的注解Bean，则为Bean设置默认值，//设置Bean的自动依赖注入装配属性等if (candidate instanceof AbstractBeanDefinition) {postProcessBeanDefinition((AbstractBeanDefinition) candidate, beanName);}//如果扫描到的Bean是Spring的注解Bean，则处理其通用的Spring注解if (candidate instanceof AnnotatedBeanDefinition) {//处理注解Bean中通用的注解，在分析注解Bean定义类读取器时已经分析过AnnotationConfigUtils.processCommonDefinitionAnnotations((AnnotatedBeanDefinition) candidate);}//根据Bean名称检查指定的Bean是否需要在容器中注册，或者在容器中冲突if (checkCandidate(beanName, candidate)) {BeanDefinitionHolder definitionHolder = new BeanDefinitionHolder(candidate, beanName);//根据注解中配置的作用域，为Bean应用相应的代理模式definitionHolder =AnnotationConfigUtils.applyScopedProxyMode(scopeMetadata, definitionHolder, this.registry);beanDefinitions.add(definitionHolder);//向容器注册扫描到的BeanregisterBeanDefinition(definitionHolder, this.registry);}}}return beanDefinitions;}
}

类路径 Bean 定义扫描器 ClassPathBeanDefinitionScanner 主要通过 findCandidateComponents()方法调用其父类 ClassPathScanningCandidateComponentProvider 类来扫描获取给定包及其子包下的类。
2)、ClassPathScanningCandidateComponentProvider 扫描给定包及其子包的类
ClassPathScanningCandidateComponentProvider 类的 findCandidateComponents()方法具体实现扫描给定类路径包的功能，主要源码如下：

public class ClassPathScanningCandidateComponentProvider implements EnvironmentCapable, ResourceLoaderAware {//保存过滤规则要包含的注解，即Spring默认的@Component、@Repository、@Service、//@Controller注解的Bean，以及JavaEE6的@ManagedBean和JSR-330的@Named注解private final List<TypeFilter> includeFilters = new LinkedList<>();//保存过滤规则要排除的注解private final List<TypeFilter> excludeFilters = new LinkedList<>();/*** Create a ClassPathScanningCandidateComponentProvider with a {@link StandardEnvironment}.* @param useDefaultFilters whether to register the default filters for the* {@link Component @Component}, {@link Repository @Repository},* {@link Service @Service}, and {@link Controller @Controller}* stereotype annotations* @see #registerDefaultFilters()*///构造方法，该方法在子类ClassPathBeanDefinitionScanner的构造方法中被调用public ClassPathScanningCandidateComponentProvider(boolean useDefaultFilters) {this(useDefaultFilters, new StandardEnvironment());}/*** Create a ClassPathScanningCandidateComponentProvider with the given {@link Environment}.* @param useDefaultFilters whether to register the default filters for the* {@link Component @Component}, {@link Repository @Repository},* {@link Service @Service}, and {@link Controller @Controller}* stereotype annotations* @param environment the Environment to use* @see #registerDefaultFilters()*/public ClassPathScanningCandidateComponentProvider(boolean useDefaultFilters, Environment environment) {//如果使用Spring默认的过滤规则，则向容器注册过滤规则if (useDefaultFilters) {registerDefaultFilters();}setEnvironment(environment);setResourceLoader(null);}/*** Register the default filter for {@link Component @Component}.* <p>This will implicitly register all annotations that have the* {@link Component @Component} meta-annotation including the* {@link Repository @Repository}, {@link Service @Service}, and* {@link Controller @Controller} stereotype annotations.* <p>Also supports Java EE 6's {@link javax.annotation.ManagedBean} and* JSR-330's {@link javax.inject.Named} annotations, if available.**///向容器注册过滤规则@SuppressWarnings("unchecked")protected void registerDefaultFilters() {//向要包含的过滤规则中添加@Component注解类，注意Spring中@Repository//@Service和@Controller都是Component，因为这些注解都添加了@Component注解this.includeFilters.add(new AnnotationTypeFilter(Component.class));//获取当前类的类加载器ClassLoader cl = ClassPathScanningCandidateComponentProvider.class.getClassLoader();try {//向要包含的过滤规则添加JavaEE6的@ManagedBean注解this.includeFilters.add(new AnnotationTypeFilter(((Class<? extends Annotation>) ClassUtils.forName("javax.annotation.ManagedBean", cl)), false));logger.debug("JSR-250 'javax.annotation.ManagedBean' found and supported for component scanning");}catch (ClassNotFoundException ex) {// JSR-250 1.1 API (as included in Java EE 6) not available - simply skip.}try {//向要包含的过滤规则添加@Named注解this.includeFilters.add(new AnnotationTypeFilter(((Class<? extends Annotation>) ClassUtils.forName("javax.inject.Named", cl)), false));logger.debug("JSR-330 'javax.inject.Named' annotation found and supported for component scanning");}catch (ClassNotFoundException ex) {// JSR-330 API not available - simply skip.}}/*** Scan the class path for candidate components.* @param basePackage the package to check for annotated classes* @return a corresponding Set of autodetected bean definitions*///扫描给定类路径的包public Set<BeanDefinition> findCandidateComponents(String basePackage) {if (this.componentsIndex != null && indexSupportsIncludeFilters()) {return addCandidateComponentsFromIndex(this.componentsIndex, basePackage);}else {return scanCandidateComponents(basePackage);}}private Set<BeanDefinition> addCandidateComponentsFromIndex(CandidateComponentsIndex index, String basePackage) {//创建存储扫描到的类的集合Set<BeanDefinition> candidates = new LinkedHashSet<>();try {Set<String> types = new HashSet<>();for (TypeFilter filter : this.includeFilters) {String stereotype = extractStereotype(filter);if (stereotype == null) {throw new IllegalArgumentException("Failed to extract stereotype from "+ filter);}types.addAll(index.getCandidateTypes(basePackage, stereotype));}boolean traceEnabled = logger.isTraceEnabled();boolean debugEnabled = logger.isDebugEnabled();for (String type : types) {//为指定资源获取元数据读取器，元信息读取器通过汇编(ASM)读//取资源元信息MetadataReader metadataReader = getMetadataReaderFactory().getMetadataReader(type);//如果扫描到的类符合容器配置的过滤规则if (isCandidateComponent(metadataReader)) {//通过汇编(ASM)读取资源字节码中的Bean定义元信息AnnotatedGenericBeanDefinition sbd = new AnnotatedGenericBeanDefinition(metadataReader.getAnnotationMetadata());if (isCandidateComponent(sbd)) {if (debugEnabled) {logger.debug("Using candidate component class from index: " + type);}candidates.add(sbd);}else {if (debugEnabled) {logger.debug("Ignored because not a concrete top-level class: " + type);}}}else {if (traceEnabled) {logger.trace("Ignored because matching an exclude filter: " + type);}}}}catch (IOException ex) {throw new BeanDefinitionStoreException("I/O failure during classpath scanning", ex);}return candidates;}/*** Determine whether the given class does not match any exclude filter* and does match at least one include filter.* @param metadataReader the ASM ClassReader for the class* @return whether the class qualifies as a candidate component*///判断元信息读取器读取的类是否符合容器定义的注解过滤规则protected boolean isCandidateComponent(MetadataReader metadataReader) throws IOException {//如果读取的类的注解在排除注解过滤规则中，返回falsefor (TypeFilter tf : this.excludeFilters) {if (tf.match(metadataReader, getMetadataReaderFactory())) {return false;}}//如果读取的类的注解在包含的注解的过滤规则中，则返回turefor (TypeFilter tf : this.includeFilters) {if (tf.match(metadataReader, getMetadataReaderFactory())) {return isConditionMatch(metadataReader);}}//如果读取的类的注解既不在排除规则，也不在包含规则中，则返回falsereturn false;}
}

注册注解 BeanDefinition
AnnotationConfigWebApplicationContext 是 AnnotationConfigApplicationContext 的 Web 版，它们对于注解 Bean 的注册和扫描是基本相同的，但是 AnnotationConfigWebApplicationContext对注解 Bean 定义的载入稍有不同，AnnotationConfigWebApplicationContext 注入注解 Bean 定义源码如下：

/*** Register a {@link org.springframework.beans.factory.config.BeanDefinition} for* any classes specified by {@link #register(Class...)} and scan any packages* specified by {@link #scan(String...)}.* <p>For any values specified by {@link #setConfigLocation(String)} or* {@link #setConfigLocations(String[])}, attempt first to load each location as a* class, registering a {@code BeanDefinition} if class loading is successful,* and if class loading fails (i.e. a {@code ClassNotFoundException} is raised),* assume the value is a package and attempt to scan it for annotated classes.* <p>Enables the default set of annotation configuration post processors, such that* {@code @Autowired}, {@code @Required}, and associated annotations can be used.* <p>Configuration class bean definitions are registered with generated bean* definition names unless the {@code value} attribute is provided to the stereotype* annotation.* @param beanFactory the bean factory to load bean definitions into* @see #register(Class...)* @see #scan(String...)* @see #setConfigLocation(String)* @see #setConfigLocations(String[])* @see AnnotatedBeanDefinitionReader* @see ClassPathBeanDefinitionScanner*///载入注解Bean定义资源@Overrideprotected void loadBeanDefinitions(DefaultListableBeanFactory beanFactory) {//为容器设置注解Bean定义读取器AnnotatedBeanDefinitionReader reader = getAnnotatedBeanDefinitionReader(beanFactory);//为容器设置类路径Bean定义扫描器ClassPathBeanDefinitionScanner scanner = getClassPathBeanDefinitionScanner(beanFactory);//获取容器的Bean名称生成器BeanNameGenerator beanNameGenerator = getBeanNameGenerator();//为注解Bean定义读取器和类路径扫描器设置Bean名称生成器if (beanNameGenerator != null) {reader.setBeanNameGenerator(beanNameGenerator);scanner.setBeanNameGenerator(beanNameGenerator);beanFactory.registerSingleton(AnnotationConfigUtils.CONFIGURATION_BEAN_NAME_GENERATOR, beanNameGenerator);}//获取容器的作用域元信息解析器ScopeMetadataResolver scopeMetadataResolver = getScopeMetadataResolver();//为注解Bean定义读取器和类路径扫描器设置作用域元信息解析器if (scopeMetadataResolver != null) {reader.setScopeMetadataResolver(scopeMetadataResolver);scanner.setScopeMetadataResolver(scopeMetadataResolver);}if (!this.annotatedClasses.isEmpty()) {if (logger.isInfoEnabled()) {logger.info("Registering annotated classes: [" +StringUtils.collectionToCommaDelimitedString(this.annotatedClasses) + "]");}reader.register(this.annotatedClasses.toArray(new Class<?>[this.annotatedClasses.size()]));}if (!this.basePackages.isEmpty()) {if (logger.isInfoEnabled()) {logger.info("Scanning base packages: [" +StringUtils.collectionToCommaDelimitedString(this.basePackages) + "]");}scanner.scan(this.basePackages.toArray(new String[this.basePackages.size()]));}//获取容器定义的Bean定义资源路径String[] configLocations = getConfigLocations();//如果定位的Bean定义资源路径不为空if (configLocations != null) {for (String configLocation : configLocations) {try {//使用当前容器的类加载器加载定位路径的字节码类文件Class<?> clazz = ClassUtils.forName(configLocation, getClassLoader());if (logger.isInfoEnabled()) {logger.info("Successfully resolved class for [" + configLocation + "]");}reader.register(clazz);}catch (ClassNotFoundException ex) {if (logger.isDebugEnabled()) {logger.debug("Could not load class for config location [" + configLocation +"] - trying package scan. " + ex);}//如果容器类加载器加载定义路径的Bean定义资源失败//则启用容器类路径扫描器扫描给定路径包及其子包中的类int count = scanner.scan(configLocation);if (logger.isInfoEnabled()) {if (count == 0) {logger.info("No annotated classes found for specified class/package [" + configLocation + "]");}else {logger.info("Found " + count + " annotated classes in package [" + configLocation + "]");}}}}}}

以上就是解析和注入注解配置资源的全过程分析。

IOC 容器初始化小结

现在通过上面的代码，总结一下 IOC 容器初始化的基本步骤：
1、初始化的入口在容器实现中的 refresh()调用来完成。
2、对 Bean 定义载入 IOC 容器使用的方法是 loadBeanDefinition(), 其中的大致过程如下：通过 ResourceLoader 来完成资源文件位置的定位，DefaultResourceLoader是默认的实现，同时上下文本身就给出了 ResourceLoader 的实现，可以从类路径，文件系统,URL 等方式来定为资源位置。如果是 XmlBeanFactory 作为 IOC 容器，那么需要为它指定 Bean 定义的资源，也 就 是 说 Bean 定 义 文 件 时 通 过 抽 象 成 Resource 来 被 IOC 容 器 处 理 的 ， 容 器 通 过BeanDefinitionReader 来 完 成 定 义 信 息 的 解 析 和 Bean 信 息 的 注 册 , 往 往 使 用 的 是XmlBeanDefinitionReader 来 解 析 Bean 的 XML 定 义 文 件 - 实 际 的 处 理 过 程 是 委 托 给BeanDefinitionParserDelegate 来完成的，从而得到 bean 的定义信息，这些信息在 Spring 中使用BeanDefinition对象来表示-这个名字可以让我们想到loadBeanDefinition(),registerBeanDefinition()这些相关方法。它们都是为处理 BeanDefinitin 服务的，容器解析得到 BeanDefinition 以后，需要把它在 IOC 容器中注册，这由 IOC 实现 BeanDefinitionRegistry 接口来实现。注册过程就是在 IOC 容器内部维护的一个 HashMap 来保存得到的 BeanDefinition 的过程。这个 HashMap 是 IOC 容器持有Bean 信息的场所，以后对 Bean 的操作都是围绕这个 HashMap 来实现的。
然后我们就可以通过 BeanFactory 和 ApplicationContext 来享受到 Spring IOC 的服务了,在使用 IOC容器的时候，我们注意到除了少量粘合代码，绝大多数以正确 IOC 风格编写的应用程序代码完全不用关心如何到达工厂，因为容器将把这些对象与容器管理的其他对象钩在一起。基本的策略是把工厂放到已知的地方，最好是放在对预期使用的上下文有意义的地方，以及代码将实际需要访问工厂的地方。Spring本身提供了对声明式载入web应用程序用法的应用程序上下文,并将其存储在ServletContext中的框架实现。