We now allow the configuration whether null values of entity objects should be inserted as tombstones using INSERT statements. By default, this option is disable. Enabling tombstones allows upsert usage through INSERT statements as null values get inserted in Cassandra. This option is useful if an object should be persisted in its whole state. Inserting nulls causes tombstones to be created and can have a negative impact on performance.
This change allows SimpleRepository and SimpleCassandraRepository to use CassandraTemplate directly allowing emission of lifecycle events.
We now provide Kotlin extensions for imperative, asynchronous, and reactive Template API interfaces.
CqlOperations and CassandraOperations expose Kotlin-friendly methods accepting KClasses and leveraging reified generics. We also provide Kotlin-friendly method renames for methods such as in that do not require quoting in Kotlin code.
operations.query(Person::class).inTable("my_table").asType<User>().all()
operations.select<Person>(query(where("firstname").isEqualTo("Walter") and where("lastname").isEqualTo("White")))
We now consider user-type dependencies while dropping user-defined types. User types can utilize other user-defined types and dropping a type that is in use causes the drop operation to fail. We're introspecting user types reported by Cassandra and construct a dependency graph that is used to determine the proper drop order.
We now support Cassandra time columns via LocalTime types (JSR-310, Joda and ThreeTenBackport) in domain classes, queries and updates.
@Table
class Schedule {
@Id String id;
LocalTime scheduledAt;
}
We now support mapped tuple values via @Tuple and @Element(1) annotations. Mapped tuple values can be embedded within table entities and user-defined. Mapped tuples can contain user-defined types, collection types and simple property types and are supported with schema generation.
@Tuple
class Address {
@Element(0) String city;
@Element(1) String street;
@Element(2) int sortOrder;
}
We now export composable repositories through our CDI extension. Repositories can now be customized either by a single custom implementation (as it was before) and by providing fragment interfaces along their fragment implementation.
This change aligns CDI support with the existing RepositoryFactory support we provide within a Spring application context.
Add AfterConvertEvent, introduce base class for AbstractDeleteEvent. Turn table name in mapping events to non-nullable. Replace guessTableName(…) with getTableName(…) and table name extraction from statements. Pass table name to converter mapper function for event propagation. Refactor tests to base class and test operations accessor.
Introduce lifecycle events and ProjectionFactory to AsyncCassandraTemplate.
Extend JavaDoc, add author and since tags. Reduce copyright year to inception year of new classes. Extend reference documentation.
Original pull request: #123.
We now support persistence lifecycle callbacks via Spring's ApplicationEvents. Events are fired upon select, insert, update, delete, and truncate statements. The following events are available:
* BeforeSaveEvent: Before inserting/updating a row in the database, via insert(…) and update(…).
* AfterSaveEvent: After inserting/updating a row in the database, via insert(…) and update(…).
* BeforeDeleteEvent: Before deleting row from the database, via delete(…) and truncate(…).
* AfterDeleteEvent: After deleting row from the database, via delete(…) and truncate(…).
* AfterLoadEvent: After retrieving a row from the database, via select(…), slice(…), and stream(…).
* AfterConvertEvent: After converting a row from the database to a POJO, via select(…), slice(…), and stream(…).
Original pull request: #123.
