Skip to content

Latest commit

 

History

History
105 lines (70 loc) · 5.5 KB

entity-persistence.adoc

File metadata and controls

105 lines (70 loc) · 5.5 KB

Persisting Entities

Saving an aggregate can be performed with the CrudRepository.save(…) method. If the aggregate is new, this results in an insert for the aggregate root, followed by insert statements for all directly or indirectly referenced entities.

If the aggregate root is not new, all referenced entities get deleted, the aggregate root gets updated, and all referenced entities get inserted again. Note that whether an instance is new is part of the instance’s state.

Note
 This approach has some obvious downsides. If only few of the referenced entities have been actually changed, the deletion and insertion is wasteful. While this process could and probably will be improved, there are certain limitations to what Spring Data JDBC can offer. It does not know the previous state of an aggregate. So any update process always has to take whatever it finds in the database and make sure it converts it to whatever is the state of the entity passed to the save method.

See also Entity State Detection for further details.

Loading Aggregates

Spring Data JDBC offers two ways how it can load aggregates:

  1. The traditional and before version 3.2 the only way is really simple: Each query loads the aggregate roots, independently if the query is based on a CrudRepository method, a derived query or a annotated query. If the aggregate root references other entities those are loaded with separate statements.

  2. Spring Data JDBC 3.2 allows the use of Single Query Loading. With this an arbitrary number of aggregates can be fully loaded with a single SQL query. This should be significantly more efficient, especially for complex aggregates, consisting of many entities.

    Currently, Single Query Loading is restricted in different ways:

    1. The aggregate must not have nested collections, this includes Map. The plan is to remove this constraint in the future.

    2. The aggregate must not use AggregateReference or embedded entities. The plan is to remove this constraint in the future.

    3. The database dialect must support it.Of the dialects provided by Spring Data JDBC all but H2 and HSQL support this. H2 and HSQL don’t support analytic functions (aka windowing functions).

    4. It only works for the find methods in CrudRepository, not for derived queries and not for annotated queries. The plan is to remove this constraint in the future.

    5. Single Query Loading needs to be enabled in the JdbcMappingContext, by calling setSingleQueryLoadingEnabled(true)

If any condition is not fulfilled Spring Data JDBC falls back to the default approach of loading aggregates.

Note
 Single Query Loading is to be considered experimental. We appreciate feedback on how it works for you.
Note
 While Single Query Loading can be abbreviated as SQL, but we highly discourage doing so since confusion with Structured Query Language is almost guaranteed.

Template API

As an alternative to repositories Spring Data JDBC offers the javadoc:org.springframework.data.jdbc.core.JdbcAggregateTemplate[] as a more direct means to load and persist entities in a relational database. To a large extent, repositories use JdbcAggregateTemplate to implement their features.

This section highlights only the most interesting parts of the JdbcAggregateTemplate. For a more complete overview, see the JavaDoc of JdbcAggregateTemplate.

Accessing the JdbcAggregateTemplate

JdbcAggregateTemplate is intended to be used as a Spring bean. If you have set up your application to include Spring Data JDBC, you can configure a dependency on JdbcAggregateTemplate in any Spring bean, and the Spring Framework injects a properly configured instance.

This includes fragments you use to implement custom methods for your Spring Data Repositories, letting you to use JdbcAggregateTemplate to customize and extend your repositories.

Persisting

JdbcAggregateTemplate offers three types of methods for persisting entities: save, insert, and update. Each comes in two flavors: Operating on single aggregates, named exactly as mentioned above, and with an All suffix operation on an Iterable.

save does the same as the method of same name in a repository.

insert and update skip the test if the entity is new and assume a new or existing aggregate as indicated by their name.

Querying

JdbcAggregateTemplate offers a considerable array of methods for querying aggregates and about collections of aggregates. There is one type of method that requires special attention. That’s the methods taking a Query as an argument. They allow the execution of programmatically constructed queries, as follows:

template.findOne(query(where("name").is("Gandalf")), Person.class);

The javadoc:org.springframework.data.relational.core.query.Query[] returned by the query method defines the list of columns to select, a where clause (through a CriteriaDefinition), and specification of limit and offset clauses. For details of the Query class, see its JavaDoc.

The javadoc:org.springframework.data.relational.core.query.Criteria[] class, of which where is a static member, provides implementations of org.springframework.data.relational.core.query.CriteriaDefinition[], which represent the where-clause of the query.

../commons/criteria-methods.adoc

Optimistic Locking