I am trying to understand the difference here. I see that a class has been annotated with both of them with same package example :
@Configuration
@EntityScan("some.known.persistence")
@ComponentScan({ "some.known.persistence"})
public class ApiConfig {
}
I understand the differences from API docs but want to understand in detail. Also does it mean that anything that is scanned by @ComponentScan is having wider visibility wrt Spring context and @EntityScan does not. If so using some attribute with @ComponentScan should have sufficed the need to bind in a JPA context, isn't it?
The @ComponentScan annotation is used to automatically create beans for every class annotated with @Component, @Service, @Controller, @RestController, @Repository, ... and adds them to the Spring container (allowing them to be @Autowired).
The @EntityScan on the other hand does not create beans as far as I know. It only identifies which classes should be used by a specific persistence context. Since Spring boot 1.4 that includes JPA, MongoDB, neo4j, Cassandra and CouchBase.
Why they're not merged? Well, I'm not on the Spring team, but since they have different meanings, why should they be merged? The @EntityScan should mainly be used to scan your entity packages, while the @ComponentScan should scan all packages that contain Spring beans, so the following is very likely:
@ComponentScan("org.example.base")
@EntityScan("org.example.base.entities")
public class MyConfig {
}