平时看spring源码中经常用到Conditional这个注解,今天来看一下他是怎么生效的
- 其实我是发现Conditional注解是在ConditionEvaluator类的shouldSkip方法中调用了
org.springframework.context.annotation.ConditionEvaluator#shouldSkip
shouldSkip方法有个枚举类型的参数ConfigurationPhase,PARSE_CONFIGURATION
是标示配置解析阶段;REGISTER_BEAN
标示配置注册阶段
/**
* The various configuration phases where the condition could be evaluated.
*/
enum ConfigurationPhase {
/**
* The {@link Condition} should be evaluated as a {@code @Configuration}
* class is being parsed.
* <p>If the condition does not match at this point, the {@code @Configuration}
* class will not be added.
*/
PARSE_CONFIGURATION,
/**
* The {@link Condition} should be evaluated when adding a regular
* (non {@code @Configuration}) bean. The condition will not prevent
* {@code @Configuration} classes from being added.
* <p>At the time that the condition is evaluated, all {@code @Configuration}s
* will have been parsed.
*/
REGISTER_BEAN
}
- 接着发现shouldSkip方法是在ConfigurationClassParser类的processConfigurationClass方法中进行了调用
org.springframework.context.annotation.ConfigurationClassParser#processConfigurationClass
回到ConfigurationClassParser
这个类就比较好理解,说明还是在解析配置阶段根据Conditional
注解进行处理的
ConfigurationClassParser 解析配置流程,点击查看