Spring数据存储库方法实现

Question

I'm working in a personal web project and I would like to use Spring Data and JPA. I found this nice blog entry in springsource.com and I've spent several hours trying to understand how this work and why they do it this way. I need help with two questions mainly:

1. Why is necesary to use Service and ServiceImpl classes. That seems like a typical JPA 2 implementation but using Spring Data underneath with an Autowired repository. This really confuse me. Why not use the repository directly?
2. The writer defines a pair of new methods in CustomerRepository but there isn't any implementation. He continues implementing CustomerService that it's not related with the repository. I've seen this in many sites but I don't understand how it works.

As I said I've been looking for a good article to find answers to my questions but most of them seem that are following these one. I've also checked github code of the article's project but it's uncomplete. I hope you can help me.

Answer 1

The point of the blog post is to talk about spring-data repositories. So it doesn't elaborate on why services are necessary.

They are because a typical business method will deal with several DAOs/Repositories, in a single transaction. For example, transferring money might include calling the AccountRepository, the AudiRepository and the CustomerRepository. The service's main goal is to provide transactional business methods to the presentation layer.