为什么JpaRepository.delete（）在删除实体之前先检索它？

Question

当我使用其ID删除实体时，会得到以下日志输出。

2020-09-19 17:07:35.536 DEBUG 14544 --- [nio-8080-exec-2] org.hibernate.SQL                        : select comment0_.id as id1_15_0_, comment0_.acceptedAsAnswer as accepted2_15_0_, comment0_.commentDate as commentD3_15_0_, comment0_.commenter_id as comment10_15_0_, comment0_.commenterIp as commente4_15_0_, comment0_.content as content5_15_0_, comment0_.dateDeleted as dateDele6_15_0_, comment0_.deleted as deleted7_15_0_, comment0_.deleter_id as deleter11_15_0_, comment0_.journal_entry_id as journal12_15_0_, comment0_.language_id as languag13_15_0_, comment0_.lastEditedOn as lastEdit8_15_0_, comment0_.likeTotal as likeTota9_15_0_, comment0_.owning_comment_id as owning_14_15_0_ from comments comment0_ where comment0_.id=?
2020-09-19 17:07:35.562 DEBUG 14544 --- [nio-8080-exec-2] org.hibernate.SQL                        : delete from comments where id=?

我正在构建一个使用Amazon的数据库服务的Web应用程序，因此该应用程序发出的每个请求都需要额外的钱。为什么Spring在删除对象之前先检索该对象，并且有什么方法可以阻止它执行此操作？

Answer 1

正如@Naim指出的那样，hibernate首先检索对象，因为可能存在拦截器。这是我的解决方案，可使用JpaRepository实现存储库直接对数据库进行删除查询。

@Modifying
@Transactional
@Query(nativeQuery = true, value = "DELETE FROM comments WHERE id = :id")
void deleteById(@Param("id") Long id); 

自定义方法deleteById不需要先获取实体。