i have an aggregate that internally has two 1 to M relationship
@Getter
@Setter
@AllArgsConstructor
@Builder
@EqualsAndHashCode(of = {"id"})
@Table("TOOL")
public class AgrTool {
private @Id Long id;
private @Version Long version;
private String status;
//my flat parameters...
private LocalDate creationDate;
private LocalDate acquireDate;
private String serialNo;
................
................
//my two 1-to-M
private List<VoDefinitionCalibrationInternal> calibrationInternalDefinitions = new ArrayList<>();
private List<VoDefinitionCalibrationExternal> calibrationExternalDefinitions = new ArrayList<>();
}
based on our business logic we need to inquery just the TOOL table returing a list of object.
Pratically we don't want to fetch the two relationship (calibrationInternalDefinitions and calibrationExternalDefinitions) we need to show a table wiht the miminal set of attribute of the TOOL table
as defined in the documentation we apply the projections expecting, that the query is limited to the field exposed by the consrtuctor of the Projection Class:
(official documentation) --> If the store optimizes the query execution by limiting the fields to be loaded, the fields to be loaded are determined from the parameter names of the constructor that is exposed.
then we implemented the minimal value object without the two relationships
calibrationInternalDefinitions
calibrationExternalDefinitions
@Getter
@Setter
@AllArgsConstructor
@Builder
public class PrjTool{
private Long id;
private String status;
//my flat parameters...
private LocalDate creationDate;
private LocalDate acquireDate;
private String serialNo;
................
................
}
and in our repository we add two methods:
public interface RepositoryTool extends ListCrudRepository<AgrTool,Long> {
<T> Page<T> findAllByCustomerAndEnvironment(Long customer, Long environment, Pageable pageable , Class<T> type);
<T> Collection<T> findAllByCustomerAndEnvironment(Long customer, Long environment, Class<T> type);
analisyng the result of our Unit test where we call this two methods:
Page<PrjTool> result = null;
result = repository.findAllByCustomerAndEnvironment(-160L, -160L, pageRequest, PrjTool.class);
result.forEach(p -> log.info("element:" + p));
assertThat(result.getNumberOfElements()).isEqualTo(2);
Collection<PrjTool> result = null;
result = repository.findAllByCustomerAndEnvironment(-160L, -160L, PrjTool.class);
result.forEach(p -> log.info("element:" + p));
assertThat(result.isEmpty()).isEqualTo(false);
activating the trace of the query from the log we see that the query generate continue to do select in the relasionsip also if no define in the projection object,
mp911de
commented
2 weeks ago
i have an aggregate that internally has two 1 to M relationship
based on our business logic we need to inquery just the TOOL table returing a list of object. Pratically we don't want to fetch the two relationship (calibrationInternalDefinitions and calibrationExternalDefinitions) we need to show a table wiht the miminal set of attribute of the TOOL table
as defined in the documentation we apply the projections expecting, that the query is limited to the field exposed by the consrtuctor of the Projection Class: (official documentation) --> If the store optimizes the query execution by limiting the fields to be loaded, the fields to be loaded are determined from the parameter names of the constructor that is exposed.
then we implemented the minimal value object without the two relationships
and in our repository we add two methods:
analisyng the result of our Unit test where we call this two methods:
activating the trace of the query from the log we see that the query generate continue to do select in the relasionsip also if no define in the projection object,
Expectation only the table TOOL table should be analyzed if required I can create a test to show the behavior