Will hibernate java programs have no sql code?

I haven't worked with hibernate. I have little bit of experience in java. I was going through source of a beast of an java application created by Oracle(Retail Price Management). I was expecting a lot of sql code embed开发者_如何转开发ded in there as the application makes heavy use of database. But to my surprise, NO embedded SQL code! so far. I found that it was using what is called as "Hibernate" from the lot of .hbm.xml files. Is it a trademark for java programs using hibernate or maybe I haven't seen the complete codebase?. Could someone enlighten me how this is possible?. Thanks.

Hibernate, as all ORM tools, indeed lessens or eliminates the need to use raw SQL in Java code, due to the following:

• many associations between various entities are recorded in the Hibernate mapping, so these are fetched automatically by Hibernate - i.e. if you have an aggregation relationshiop between two classes on the Java side, this may be mapped as a foreign key relationship in the DB, and Hibernate, whenever an instance of class A is loaded, can automatically load the associated instances of class B too,
• many queries can be done in Hibernate's own HQL query language, or using its Criteria API.

Under the hood Hibernate does generate SQL to communicate with the DB, but this is not visible on the Java side. It can be seen in the logs though, if it is enabled.

Due to this, programs using Hibernate very rarely need to use JDBC or SQL directly. The exceptions are typically ralted to "tricky" legacy DB schemas which can't be fully handled by Hibernate.

Because that's the whole purpose of using Hibernate or any other object-relational mapping framework.

Hibernate solves object-relational impedance mismatch problems by replacing direct persistence-related database accesses with high-level object handling functions.

Hibernate generates SQL for all its standard database operations. It understands different SQL dialects, and the mapping files (.hbm.xml) tell it about the database structure so it knows how to construct its queries. There is a showSql setting you can turn on if you want to see it outputting its generated SQL as it runs.

Hibernate is an Object-Relational Mapper (ORM). ORMs are used to hide the ugly details of SQL incompatibility[sic] between databases from your program -- you define your tables and map them to an object hierarchy (the .hbm.xml files) and then Hibernate does the rest. Thus most programs that use Hibernate won't see a single phrase of SQL, unless there's a specific reason to execute a complicated query.

Hibernate is a tool, or technology that takes care of the interaction between the database and application for you. You have to tell the structure of the application and the database to it, this is what is in the .hbm.xml files.

The SQL is generated by Hibernate at runtime (kind of)

Say you have an Fruit class, and objects of this is persisted into a T_FRUIT table.

You say this to hibernate, via the .hbm.xml files. That there is a table T_FRUIT, this table is represented by the Fruit class, and which fields in the Fruit class correspond to which columns in th T_FRUIT table.

And then it knows whenever you are trying to save a fruit, it should insert/update to the T_FRUIT table.

When you want to create an Apple, you create an object of fruit corresponding to apple and save "save this fruit". Hibernate takes care of persisting it.

You can have relationships defined between tables, and Hibernate is intelligent enough to persist in multiple tables.

When you fetch a fruit, hibernate fetches the details of the fruit and its children also(data from referencing tables). And you can say whether you want fetch all the children at once, or as and when required.

And so on. Aim is to make your life easier, and code maintainable, easy to read, portable,...

With this info, let me redirect you.