简介

在Bean的创建过程中会有很多的后置处理器，例如实例化前、实例化后、初始化前、初始化后，属性填充等，这些都是通过BeanPostProcessor来实现的

那么既然每个Bean都有有这些生命周期，这些BeanPostProcessor肯定需要提前知道，并且实例化出来保存，而我们前面介绍来BeanDefinition的扫描，就可以把所有的Bean找出来，只要在实例化单例Bean之前把BeanPostProcessor实例化好就行了

在spring中就有这么一步注册BeanPostProcessor
就在AbstractApplicationContext#refresh方法中的一步registerBeanPostProcessors(beanFactory);

源码解析

public static void registerBeanPostProcessors(ConfigurableListableBeanFactory beanFactory, AbstractApplicationContext applicationContext) {// 通过getBeanNamesForType可以拿到所有的名字，因为这个时候还没有实例化，所以就拿到名字，里面使用isTypeMatch匹配String[] postProcessorNames = beanFactory.getBeanNamesForType(BeanPostProcessor.class, true, false);List<BeanPostProcessor> priorityOrderedPostProcessors = new ArrayList<>();    // 找出所有实现 PriorityOrdered的BeanPostProcessorList<BeanPostProcessor> internalPostProcessors = new ArrayList<>();    // 内置的BeanPostProcessor 如 MergedBeanDefinitionPostProcessorList<String> orderedPostProcessorNames = new ArrayList<>();    // 找出所有实现 Ordered的BeanPostProcessorList<String> nonOrderedPostProcessorNames = new ArrayList<>();    // 没有任何排序的BeanPostProcessorfor (String ppName : postProcessorNames) {if (beanFactory.isTypeMatch(ppName, PriorityOrdered.class)) {// 通过getBean获得BeanPostProcessor对象BeanPostProcessor pp = beanFactory.getBean(ppName, BeanPostProcessor.class);priorityOrderedPostProcessors.add(pp);if (pp instanceof MergedBeanDefinitionPostProcessor) {internalPostProcessors.add(pp);}}else if (beanFactory.isTypeMatch(ppName, Ordered.class)) {orderedPostProcessorNames.add(ppName);}else {nonOrderedPostProcessorNames.add(ppName);}}sortPostProcessors(priorityOrderedPostProcessors, beanFactory);    // 升序排序registerBeanPostProcessors(beanFactory, priorityOrderedPostProcessors); // 注册BeanPostProcessorsList<BeanPostProcessor> orderedPostProcessors = new ArrayList<>(orderedPostProcessorNames.size());for (String ppName : orderedPostProcessorNames) {BeanPostProcessor pp = beanFactory.getBean(ppName, BeanPostProcessor.class);         // 通过getBean获得BeanPostProcessor对象orderedPostProcessors.add(pp);if (pp instanceof MergedBeanDefinitionPostProcessor) {internalPostProcessors.add(pp);}}sortPostProcessors(orderedPostProcessors, beanFactory);    // 升序排序registerBeanPostProcessors(beanFactory, orderedPostProcessors); // 注册BeanPostProcessorsList<BeanPostProcessor> nonOrderedPostProcessors = new ArrayList<>(nonOrderedPostProcessorNames.size());for (String ppName : nonOrderedPostProcessorNames) {BeanPostProcessor pp = beanFactory.getBean(ppName, BeanPostProcessor.class);  // 通过getBean获得BeanPostProcessor对象nonOrderedPostProcessors.add(pp);if (pp instanceof MergedBeanDefinitionPostProcessor) {internalPostProcessors.add(pp);}}registerBeanPostProcessors(beanFactory, nonOrderedPostProcessors);sortPostProcessors(internalPostProcessors, beanFactory);     // 升序排序 MergedBeanDefinitionPostProcessor排在最后registerBeanPostProcessors(beanFactory, internalPostProcessors);beanFactory.addBeanPostProcessor(new ApplicationListenerDetector(applicationContext)); //注册 ApplicationListenerDetector 这个后续将
}

注册也会有个先后顺序，首先是实现了PriorityOrdered的排在前面，然后是实现Ordered接口的，再然后是没有现实排序接口的，最后是内置的例如MergedBeanDefinitionPostProcessor

这些BeanPostProcessor都会注册，通过beanFactory缓存起来

private final List<BeanPostProcessor> beanPostProcessors = new BeanPostProcessorCacheAwareList();public void addBeanPostProcessor(BeanPostProcessor beanPostProcessor) {Assert.notNull(beanPostProcessor, "BeanPostProcessor must not be null");this.beanPostProcessors.remove(beanPostProcessor);this.beanPostProcessors.add(beanPostProcessor);
}

总结

从上面的代码中可以看出，在其他的单例Bean实例化之前，会先把BeanPostProcessor实例化好，并且排序

后面讲实例化前的BeanPostProcessor

欢迎关注，学习不迷路！