JPA static metamodel with #MappedSuperclass (with a bit of Generation Gap Pattern) - jpa

I tried to achieve Generation Gap Pattern with JPA entities.
Here is the solution we choose ( <-- are inheritance)
BaseEntity <-- EntityGenerated <-- Entity
The EntityGenerated type is abstract and mapped with #MappedSuperclass, all field are generated with correct mapping annotation, relation point to the concrete subclass, not the Generated one.
The Entity is a concrete type, generated only if the class doesn't exist, initially there is just the class declaration annotated with #Entity. Other mapping attributes such as the #Table, etc are in a generated orm.xml.
Now, when we generate the jpa static metamodel (using hibernate or openjpa metamodel generator), the generated classes look like :
public class BaseEntity_ {
public static volatile SingularAttribute<PersistentDomainObject,Long> id;
public static volatile SingularAttribute<PersistentDomainObject,Long> timeStamp;
}
public class UserGenerated_ extends BaseEntity_ {
public static volatile SetAttribute<UserGenerated,Group> groups;
}
public class User_ extends UserGenerated_ {
}
If I want to use User_ in a jpa criteria query, I'll do something like :
CriteriaQuery<User> query = criteriaBuilder.createQuery(User.class);
Root<User> root = query.from(User.class);
query.where(root.get(User_.groups).in(paramGroups));
But It won't compile.... User_.groups is of type SetAttribute and the jpa path api for the get method is :
<E, C extends java.util.Collection<E>> Expression<C> get(PluralAttribute<X, C, E> collection);
(In comparaison, the get method for the singular attribute is
<Y> Path<Y> get(SingularAttribute<? super X, Y> attribute)
witch work better)
So, now, the questions are :
Why the metamodel generators generate the class for MappedSuperclass as there is no way to query it directly ?, attribute and relations for superclass should be defined in each subclass (where X is of subclass type)
Why the jpa criteria Path api doesn't define the get method for plural attribut as
get(PluralAttribute<? super X, C, E> collection)
?
How can I achieve the Generation Gap Pattern on JPA entity without giving up criteria query ?
Thanks

Related

JPA Hibernate 5: OneToOne in nested Embeddable causes metamodel issue

I have an entity:
#Entity
public class Test {
#Embedded
Content content;
// getters setters..
}
This contains an embedded class as you can see:
#Embeddable
public class Content {
#OneToOne
Person person;
#Embedded
Language language;
// getters setters..
}
This contains again an embeddable. 2 times nested embeddable
#Embeddable
public class Language {
String format;
#OneToOne
IdentifierCode identifierCode;
// getters setters..
}
When using the automatic schema generation feature of JPA all columns are generated in the correct way.
I use the #Data annotation on each #Entity and #Embeddable to generate getters, setters, constructors, etc..
When starting the application server (EAP 7), I notice this warning in the logs:
HHH015011: Unable to locate static metamodel field :
org.package.Language_#identifierCode; this may or may not indicate a
problem with the static metamodel
Indeed, when opening the metamodel class Language_; no identifierCode column reference is present:
#Generated(value = "org.hibernate.jpamodelgen.JPAMetaModelEntityProcessor")
#StaticMetamodel(Language.class)
public abstract class Language_ {
public static volatile SingularAttribute<Language, String> format;
}
I don't see what I'm doing wroing. Is it not possible to use #OneToOne in a nested #Embeddable? The metamodel Content_ correctly generates the singular attribute for person!
It seems when using multiple nested embeddables, something goes wrong. When using only one level of embeddables, it works.
I tried other stuff:
Adding Access.Field on the class. Nothing happens.
Instantiation the #Embedded class, like #Embedded Language language = new Language(). Nothing happens.
Replaced the #OneToOne with #ManyToOne. Nothing happens.
This sounds like a bug in your JPA provider, which you should report to them.
The JPA provider I use (DataNucleus) creates a
public static volatile SingularAttribute<Language, mydomain.model.IdentifierCode> identifierCode;
One option you have is to just use the datanucleus-jpa-query.jar in your CLASSPATH to generate the static metamodel and use those generated classes with your existing provider, alternatively use it for persistence too.

Spring Data JPA JPQL queries on parent interface

Say I have a #MappedSuperClass like this:
#MappedSuperclass
public abstract class Rating
{
#Id
private Long id;
#Column(name="USER_ID")
private Long userId;
private int rating;
...
With a concrete child entity like this
#Entity
#Table(name="ACTIVITY_RATING")
public class ActivityRating extends Rating
{
private Long activitySpecificData;
...
Then there is a Spring Data JPA repository like this:
#NoRepositoryBean
public interface RatingRepository<R extends Rating> extends JpaRepository<R, ID>
{
public List<R> findByUserId(Long userId);
...
and this:
public interface ActivityRatingRepository extends RatingRepository<ActivityRating>
{
}
This all works great and I can call findByUserId() on any of specific rating repositories that extend RatingRepository. I am now wanting to write some JPQL in the RatingRepository that all the child interfaces can inherit. I just don't know what (or if it's even possible) to put after the FROM in the query. For example:
#Query("SELECT NEW com.foo.RatingCountVo(e.rating, COUNT(e.rating)) FROM ??????? e GROUP BY e.rating")
public List<RatingCountVo> getRatingCounts();
I can add this method to each of the individual repositories that extend RatingRepository but everything would be exactly the same except for the specific entity name. If I want to change the query, I'd then have to go to all the child repositories and update them individually. I really want the query to live in the parent class and not be duplicated. Is there any way to accomplish this?
I'm currently using spring-data-jpa 1.7.2 and eclipselink 2.5.2. I'm not necessarily opposed to switching to newer versions if necessary.
Will it work if you will split query into 3 parts: start, entity and end of query? Than, if it'll work, in each interface you define constant like
String ENTITY = "ActivityRating";
And then you can use it like
#Query(RatingRepository.QUERY_START + ENTITY + RatingRepository.QUERY_END)
List<RatingCountVo> getRatingCounts();
BTW, there is no need to define public modifier in interface.
UPDATE: here is described another way:
#Query("SELECT NEW com.foo.RatingCountVo(e.rating, COUNT(e.rating)) FROM #{#entityName} e GROUP BY e.rating

Ebean inheritance "Abstract class with no readMethod" exception

I try to use inheritance with Ebean in Play! Framework 2.1.0. The inheritance strategy is "single table", as it is the only one supported by Ebean. I closely follow example from JPA Wikibook
#Entity
#Inheritance
#DiscriminatorColumn(name="price_type")
#Table(name="prices")
public abstract class Price {
#Id
public long id;
// Price value
#Column(precision=2, scale=18)
public BigDecimal value;
}
#Entity
#DiscriminatorValue("F")
public class FixedPrice extends Price {
// NO id field here
...
}
#Entity
#DiscriminatorValue("V")
public class VariablePrice extends Price {
// NO id field here
...
}
This code passes compilation, but I get
RuntimeException: Abstract class with no readMethod for models.Price.id
com.avaje.ebeaninternal.server.deploy.ReflectGetter.create(ReflectGetter.java:33)
com.avaje.ebeaninternal.server.deploy.BeanDescriptorManager.setBeanReflect(BeanDescriptorManager.java:1353)
com.avaje.ebeaninternal.server.deploy.BeanDescriptorManager.createByteCode(BeanDescriptorManager.java:1142)
com.avaje.ebeaninternal.server.deploy.BeanDescriptorManager.readDeployAssociations(BeanDescriptorManager.java:1058)
com.avaje.ebeaninternal.server.deploy.BeanDescriptorManager.readEntityDeploymentAssociations(BeanDescriptorManager.java:565)
com.avaje.ebeaninternal.server.deploy.BeanDescriptorManager.deploy(BeanDescriptorManager.java:252)
com.avaje.ebeaninternal.server.core.InternalConfiguration.<init>(InternalConfiguration.java:124)
com.avaje.ebeaninternal.server.core.DefaultServerFactory.createServer(DefaultServerFactory.java:210)
com.avaje.ebeaninternal.server.core.DefaultServerFactory.createServer(DefaultServerFactory.java:64)
com.avaje.ebean.EbeanServerFactory.create(EbeanServerFactory.java:59)
play.db.ebean.EbeanPlugin.onStart(EbeanPlugin.java:79)
Google search brings only one relevant link which is source code of ReflectGetter.java. The comment there says
For abstract classes that hold the id property we need to use reflection to get the id values some times.
This provides the BeanReflectGetter objects to do that.
If I drop abstract keyword from superclass declaration, exception disappears. I would really prefer not to make superclass concrete though.
Add getter/setter for your id field and it will go away.

Can I write a generic repository with the db first scenario to dynamically get me an entity by id?

I'd like to know how to create a method that will allow me to generically do this...
public class Repo<T> : IGenericRepo<T> where T : class
{
protected PteDotNetEntities db;
public T Get(int id)
{
//how do I dynamically get to the correct entity object and select it by
//id?
}
}
Yes you can. If you know that all your entities will have simple primary key property of type int and name Id you can do simply this:
public interface IEntity
{
int Id { get; }
}
All your entities must implement this interface. Next you can simply do:
public class Repo<T> : IGenericRepo<T> where T : class, IEntity
{
protected PteDotNetEntities db;
public T Get(int id)
{
return db.CreateObjectSet<T>().FirstOrDefault(e => e.Id == id);
}
}
This is the simplest possible solution. There are better solutions using GetObjectByKey but they are more complex. The difference between FirstOrDefault and GetObjectByKey is repeatable execution. FirstOrDefault always executes DB query whereas GetObjectByKey first checks if the entity with the same key was already loaded / attached to the context and returns it without querying the database. As reference for version using GetObjectByKey you can check similar questions:
Entity Framework Simple Generic GetByID but has differents PK Name
Generic GetById for complex PK
You can simplify those examples if you know the name of the key property upfront (forced by the interface).
In case of using code first / DbContext API you can also check this question:
Generic repository EF4 CTP5 getById

Generic repository implementation with EF

For a simple repository
public interface ISimpleRepository<T>
{
IApplicationState AppState { get; set; }
void Add(T instance);
void Delete(T instance);
void Delete(Guid rowGuid);
IQueryable<T> GetAll();
T Load(Guid rowGuid);
void SaveChanges();
void Update(T instance);
}
my implementation of the Load() method for specific repository for class Product might look like this:
public Product Load(Guid rowid)
{
return (from c in _ctx.Products where c.id == rowid select c).FirstOrDefault();
}
Now this is assumed when my repository implementation class looks like this:
public class EntityFrameworkProductsProvider : IRepository<Product> ...
What if I had like dozens or hundreds of this small and simple entities that would all use the same behaviour when doing CRUDs (use the same implementation of methods)? I certainly don't want to go and create a class to implement IRepository for each one of them..
I want something like this:
public class EntityFrameworkDefaultProvider<T> : IRepository<T> ...
but I don't know how to implement the LINQ Select expression then because of course I can't write from e in _ctx.T where e... or do I?
I haven't run into this scenario yet because so far I only had very specific entities with custom repository implementation.
Because you tagged your question with entity-framework and entity-framework-4 I assume you are using ObjectContext API. ObjectContext offers method CreateObjectSet<T> which is equivalent of Set<T> on DbContext.
This question is actually duplicate of either:
Generic GetById with DbContext
Generic GetById with ObjectContext
Instead of writing _ctx.Products, you can write _ctx.Set<T>. That takes care of half of the problem (you need to add a generic constraint where T: class to your repository)
Then, if rowid is the object's key, you can use _ctx.Set<T>.Find(rowid) instead of a LINQ query to retrieve by Id.
Alternatively, you can create a base interface IHaveId (or a BaseEntity class, whatever you like) which has the Id property, and then add that as an generic constraint on T, so you can use it in your queries.
If you're using EF 4.1, see the sample generic repository here:
http://www.asp.net/entity-framework/tutorials/implementing-the-repository-and-unit-of-work-patterns-in-an-asp-net-mvc-application
I know that this is possible in EF4.1 with the DbContext API, where you have a "Set" method on the context that gets you the entity set corresponding to the type T. this way, you could have your repository like this:
public class EntityFrameworkDefaultProvider<T> : IRepository<T> where T:class
{
public T Load(Guid rowId)
{
return _context.Set<T>().Find(rowId);
}
}
one more remark: I think you could use this syntax :
return _ctx.Products.FirstOrDefault(c=>c.id == rowid);
to get the entity you want instead of using the (from... in...). it's clearer (in my opinion) :)
Hope this helps