Preface

Evaluation license

Specification: JSR 365: Contexts and Dependency Injection for Java 2.0

Version: 2.0 EDR1

Status: Early Draft

Specification Lead:  Red Hat, Inc.

Release: June 25, 2015

Copyright 2015 Red Hat, Inc.
100 East Davie Street, Raleigh, NC 27601, U.S.A.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

   http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

Foreword

Contexts and Dependency Injection 2.0 (JSR-365) is an update to Contexts and Dependency Injection 1.2 (JSR-346).

Starting with version 2.0 CDI targets Java SE and Java EE platforms. CDI in java SE and CDI in a Java EE container share the features defined in core CDI.

Organisation of this document

This document is organized in 4 parts:

Major changes

This Early draft of CDI 2.0 includes important changes for the platform.

The full changelog can be found here.

1. Architecture

This specification defines a powerful set of complementary services that help improve the structure of application code.

  • A well-defined lifecycle for stateful objects bound to lifecycle contexts, where the set of contexts is extensible

  • A sophisticated, typesafe dependency injection mechanism, including the ability to select dependencies at either development or deployment time, without verbose configuration

  • Support for Java EE modularity and the Java EE component architecture - the modular structure of a Java EE application is taken into account when resolving dependencies between Java EE components

  • Integration with the Unified Expression Language (EL), allowing any contextual object to be used directly within a JSF or JSP page

  • The ability to decorate injected objects

  • The ability to associate interceptors to objects via typesafe interceptor bindings

  • An event notification model

  • A web conversation context in addition to the three standard web contexts defined by the Java Servlets specification

  • An SPI allowing portable extensions to integrate cleanly with the container

The services defined by this specification allow objects to be bound to lifecycle contexts, to be injected, to be associated with interceptors and decorators, and to interact in a loosely coupled fashion by firing and observing events. Various kinds of objects are injectable, including EJB 3 session beans, managed beans and Java EE resources. We refer to these objects in general terms as beans and to instances of beans that belong to contexts as contextual instances. Contextual instances may be injected into other objects by the dependency injection service.

To take advantage of these facilities, the developer provides additional bean-level metadata in the form of Java annotations and application-level metadata in the form of an XML descriptor.

The use of these services significantly simplifies the task of creating Java EE applications by integrating the Java EE web tier with Java EE enterprise services. In particular, EJB components may be used as JSF managed beans, thus integrating the programming models of EJB and JSF.

It’s even possible to integrate with third-party frameworks. A portable extension may provide objects to be injected or obtain contextual instances using the dependency injection service. The framework may even raise and observe events using the event notification service.

An application that takes advantage of these services may be designed to execute in either the Java EE environment or the Java SE environment. If the application uses Java EE services such as transaction management and persistence in the Java SE environment, the services are usually restricted to, at most, the subset defined for embedded usage by the EJB specification.

1.1. Contracts

This specification defines the responsibilities of:

  • the application developer who uses these services, and

  • the vendor who implements the functionality defined by this specification and provides a runtime environment in which the application executes.

This runtime environment is called the container. For example, the container might be a Java EE container or an embeddable EJB container.

Concepts, Programming model, Inheritance and specialization, Interceptor bindings, Decorator beans and Observer methods define the programming model for Java EE components that take advantage of the services defined by this specification, the responsibilities of the component developer, and the annotations used by the component developer to specify metadata.

Dependency injection and lookup, Scopes and contexts, Lifecycle of contextual instances, Decorators, Events and Interceptor resolution define the semantics and behavior of the services, the responsibilities of the container implementation and the APIs used by the application to interact directly with the container.

Packaging and deployment defines how Java EE applications that use the services defined by this specification must be packaged into bean archives, and the responsibilities of the container implementation at application initialization time.

Portable extensions, The Contextual interface and The Context interface define an SPI that allows portable extensions to integrate with the container.

1.2. Relationship to other specifications

An application developer creates container-managed components such as JavaBeans, EJBs or servlets and then provides additional metadata that declares additional behavior defined by this specification. These components may take advantage of the services defined by this specification, together with the enterprise and presentational aspects defined by other Java EE platform technologies.

In addition, this specification defines an SPI that allows alternative, non-platform technologies to integrate with the container and the Java EE environment, for example, alternative web presentation technologies.

1.2.1. Relationship to the Java EE platform specification

In the Java EE environment, all component classes supporting injection, as defined by the Java EE platform specification, may inject beans via the dependency injection service.

The Java EE platform specification defines a facility for injecting resources that exist in the Java EE component environment. Resources are identified by string-based names. This specification bolsters that functionality, adding the ability to inject an open-ended set of object types, including, but not limited to, component environment resources, based upon typesafe qualifiers.

1.2.2. Relationship to EJB

EJB defines a programming model for application components that access transactional resources in a multi-user environment. EJB allows concerns such as role-based security, transaction demarcation, concurrency and scalability to be specified declaratively using annotations and XML deployment descriptors and enforced by the EJB container at runtime.

EJB components may be stateful, but are not by nature contextual. References to stateful component instances must be explicitly passed between clients and stateful instances must be explicitly destroyed by the application.

This specification enhances the EJB component model with contextual lifecycle management.

Any session bean instance obtained via the dependency injection service is a contextual instance. It is bound to a lifecycle context and is available to other objects that execute in that context. The container automatically creates the instance when it is needed by a client. When the context ends, the container automatically destroys the instance.

Message-driven and entity beans are by nature non-contextual objects and may not be injected into other objects.

The container performs dependency injection on all session and message-driven bean instances, even those which are not contextual instances.

1.2.3. Relationship to managed beans

The Managed Beans specification defines the basic programming model for application components managed by the Java EE container.

As defined by this specification, most Java classes, including all JavaBeans, are managed beans.

This specification defines contextual lifecycle management and dependency injection as generic services applicable to all managed beans.

Any managed bean instance obtained via the dependency injection service is a contextual instance. It is bound to a lifecycle context and is available to other objects that execute in that context. The container automatically creates the instance when it is needed by a client. When the context ends, the container automatically destroys the instance.

The container performs dependency injection on all managed bean instances, even those which are not contextual instances.

1.2.4. Relationship to Dependency Injection for Java

The Dependency Injection for Java specification defines a set of annotations for the declaring injected fields, methods and constructors of a bean. The dependency injection service makes use of these annotations.

1.2.5. Relationship to Java Interceptors

The Java Interceptors specification defines the basic programming model and semantics for interceptors. This specification enhances that model by providing the ability to associate interceptors with beans using typesafe interceptor bindings.

1.2.6. Relationship to JSF

JavaServer Faces is a web-tier presentation framework that provides a component model for graphical user interface components and an event-driven interaction model that binds user interface components to objects accessible via Unified EL.

This specification allows any bean to be assigned a name. Thus, a JSF application may take advantage of the sophisticated context and dependency injection model defined by this specification.

1.2.7. Relationship to Bean Validation

Bean Validation provides a unified way of declaring and defining constraints on an object model, defines a runtime engine to validate objects and provides method validation.

The Bean Validation specification defines beans for Bean Validation managed objects including Validator and ValidatorFactory. A number of Bean Validation managed instances, including ConstraintValidator s can take advantage of dependency injection. Bean Validation also provides support for method parameter validation on any bean.

1.3. Introductory examples

The following examples demonstrate the use of lifecycle contexts and dependency injection.

1.3.1. JSF example

The following JSF page defines a login prompt for a web application:

<f:view>
    <h:form>
        <h:panelGrid columns="2" rendered="#{!login.loggedIn}">
            <h:outputLabel for="username">Username:</h:outputLabel>
            <h:inputText id="username" value="#{credentials.username}"/>
            <h:outputLabel for="password">Password:</h:outputLabel>
            <h:inputText id="password" value="#{credentials.password}"/>
        </h:panelGrid>
        <h:commandButton value="Login" action="#{login.login}" rendered="#{!login.loggedIn}"/>
        <h:commandButton value="Logout" action="#{login.logout}" rendered="#{login.loggedIn}"/>
    </h:form>
</f:view>

The Unified EL expressions in this page refer to beans named credentials and login.

The Credentials bean has a lifecycle that is bound to the JSF request:

@Model
public class Credentials {

    private String username;
    private String password;

    public String getUsername() { return username; }
    public void setUsername(String username) { this.username = username; }

    public String getPassword() { return password; }
    public void setPassword(String password) { this.password = password; }

}

The @Model annotation defined in Built-in stereotypes is a stereotype that identifies the Credentials bean as a model object in an MVC architecture.

The Login bean has a lifecycle that is bound to the HTTP session:

@SessionScoped @Model
public class Login implements Serializable {

    @Inject Credentials credentials;
    @Inject @Users EntityManager userDatabase;

    private CriteriaQuery<User> query;
    private Parameter<String> usernameParam;
    private Parameter<String> passwordParam;

    private User user;

    @Inject
    void initQuery(@Users EntityManagerFactory emf) {
        CriteriaBuilder cb = emf.getCriteriaBuilder();
        usernameParam = cb.parameter(String.class);
        passwordParam = cb.parameter(String.class);
        query = cb.createQuery(User.class);
        Root<User> u = query.from(User.class);
        query.select(u);
        query.where( cb.equal(u.get(User_.username), usernameParam),
                     cb.equal(u.get(User_.password), passwordParam) );
    }

    public void login() {

        List<User> results = userDatabase.createQuery(query)
            .setParameter(usernameParam, credentials.getUsername())
            .setParameter(passwordParam, credentials.getPassword())
            .getResultList();

        if ( !results.isEmpty() ) {
            user = results.get(0);
        }

    }

    public void logout() {
        user = null;
    }

    public boolean isLoggedIn() {
        return user!=null;
    }

    @Produces @LoggedIn User getCurrentUser() {
        if (user==null) {
            throw new NotLoggedInException();
        }
        else {
            return user;
        }
    }

}

The @SessionScoped annotation defined in Built-in scope types is a scope type that specifies the lifecycle of instances of Login. Managed beans with this scope must be serializable.

The @Inject annotation defined by the Dependency Injection for Java specification identifies an injected field which is initialized by the container when the bean is instantiated, or an initializer method which is called by the container after the bean is instantiated, with injected parameters.

The @Users annotation is a qualifier type defined by the application:

@Qualifier
@Retention(RUNTIME)
@Target({METHOD, FIELD, PARAMETER, TYPE})
public @interface Users {}

The @LoggedIn annotation is another qualifier type defined by the application:

@Qualifier
@Retention(RUNTIME)
@Target({METHOD, FIELD, PARAMETER, TYPE})
public @interface LoggedIn {}

The @Produces annotation defined in Declaring a producer method identifies the method getCurrentUser() as a producer method, which will be called whenever another bean in the system needs the currently logged-in user, for example, whenever the user attribute of the DocumentEditor class is injected by the container:

@Model
public class DocumentEditor {

    @Inject Document document;
    @Inject @LoggedIn User currentUser;
    @Inject @Documents EntityManager docDatabase;

    public void save() {
        document.setCreatedBy(currentUser);
        em.persist(document);
    }

}

The @Documents annotation is another application-defined qualifier type. The use of distinct qualifier types enables the container to distinguish which JPA persistence unit is required.

When the login form is submitted, JSF assigns the entered username and password to an instance of the Credentials bean that is automatically instantiated by the container. Next, JSF calls the login() method of an instance of Login that is automatically instantiated by the container. This instance continues to exist for and be available to other requests in the same HTTP session, and provides the User object representing the current user to any other bean that requires it (for example, DocumentEditor). If the producer method is called before the login() method initializes the user object, it throws a NotLoggedInException.

1.3.2. EJB example

Alternatively, we could write our Login bean to take advantage of the functionality defined by EJB:

@Stateful @SessionScoped @Model
public class Login {

    @Inject Credentials credentials;
    @Inject @Users EntityManager userDatabase;

    ...

    private User user;

    @Inject
    void initQuery(@Users EntityManagerFactory emf) {
       ...
    }

    @TransactionAttribute(REQUIRES_NEW)
    @RolesAllowed("guest")
    public void login() {
        ...
    }

    public void logout() {
        user = null;
    }

    public boolean isLoggedIn() {
        return user!=null;
    }

    @RolesAllowed("user")
    @Produces @LoggedIn User getCurrentUser() {
        ...
    }

}

The EJB @Stateful annotation specifies that this bean is an EJB stateful session bean. The EJB @TransactionAttribute and @RolesAllowed annotations declare the EJB transaction demarcation and security attributes of the annotated methods.

1.3.3. Java EE component environment example

In the previous examples, we injected container-managed persistence contexts using qualifier types. We need to tell the container what persistence context is being referred to by which qualifier type. We can declare references to persistence contexts and other resources in the Java EE component environment in Java code.

public class Databases {

    @Produces @PersistenceContext(unitName="UserData")
    @Users EntityManager userDatabaseEntityManager;

    @Produces @PersistenceUnit(unitName="UserData")
    @Users EntityManagerFactory userDatabaseEntityManagerFactory;

    @Produces @PersistenceContext(unitName="DocumentData")
    @Documents EntityManager docDatabaseEntityManager;

}

The JPA @PersistenceContext and @PersistenceUnit annotations identify the JPA persistence unit.

1.3.4. Event example

Beans may raise events. For example, our Login class could raise events when a user logs in or out.

@SessionScoped @Model
public class Login implements Serializable {

    @Inject Credentials credentials;
    @Inject @Users EntityManager userDatabase;

    @Inject @LoggedIn Event<User> userLoggedInEvent;
    @Inject @LoggedOut Event<User> userLoggedOutEvent;

    ...

    private User user;

    @Inject
    void initQuery(@Users EntityManagerFactory emf) {
       ...
    }

    public void login() {

        List<User> results = ... ;

        if ( !results.isEmpty() ) {
            user = results.get(0);
            userLoggedInEvent.fire(user);
        }

    }

    public void logout() {
        userLoggedOutEvent.fire(user);
        user = null;
    }

    public boolean isLoggedIn() {
        return user!=null;
    }

    @Produces @LoggedIn User getCurrentUser() {
        ...
    }

}

The method fire() of the built-in bean of type Event defined in The Event interface allows the application to fire events. Events consist of an event object - in this case the User - and event qualifiers. Event qualifier - such as @LoggedIn and @LoggedOut - allow event consumers to specify which events of a certain type they are interested in.

Other beans may observe these events and use them to synchronize their internal state, with no coupling to the bean producing the events:

@SessionScoped
public class Permissions implements Serializable {

    @Produces
    private Set<Permission> permissions = new HashSet<Permission>();

    @Inject @Users EntityManager userDatabase;
    Parameter<String> usernameParam;
    CriteriaQuery<Permission> query;

    @Inject
    void initQuery(@Users EntityManagerFactory emf) {
        CriteriaBuilder cb = emf.getCriteriaBuilder();
        usernameParam = cb.parameter(String.class);
        query = cb.createQuery(Permission.class);
        Root<Permission> p = query.from(Permission.class);
        query.select(p);
        query.where( cb.equal(p.get(Permission_.user).get(User_.username),
                     usernameParam) );
    }

    void onLogin(@Observes @LoggedIn User user) {
        permissions = new HashSet<Permission>( userDatabase.createQuery(query)
            .setParameter(usernameParam, user.getUsername())
            .getResultList() );
    }

    void onLogout(@Observes @LoggedOut User user {
        permissions.clear();
    }


}

The @Produces annotation applied to a field identifies the field as a producer field, as defined in Producer fields, a kind of shortcut version of a producer method. This producer field allows the permissions of the current user to be injected to an injection point of type Set<Permission>.

The @Observes annotation defined in Declaring an observer method identifies the method with the annotated parameter as an observer method that is called by the container whenever an event matching the type and qualifiers of the annotated parameter is fired.

1.3.5. Injection point metadata example

It is possible to implement generic beans that introspect the injection point to which they belong. This makes it possible to implement injection for Logger s, for example.

class Loggers {

    @Produces Logger getLogger(InjectionPoint injectionPoint) {
        return Logger.getLogger( injectionPoint.getMember().getDeclaringClass().getSimpleName() );
    }

}

The InjectionPoint interface defined in Injection point metadata, provides metadata about the injection point to the object being injected into it.

Then this class will have a Logger named "Permissions" injected:

@SessionScoped
public class Permissions implements Serializable {

    @Inject Logger log;

    ...

}

1.3.6. Interceptor example

Interceptors allow common, cross-cutting concerns to be applied to beans via custom annotations. Interceptor types may be individually enabled or disabled at deployment time.

The AuthorizationInterceptor class defines a custom authorization check:

@Secure @Interceptor
public class AuthorizationInterceptor {

    @Inject @LoggedIn User user;
    @Inject Logger log;

    @AroundInvoke
    public Object authorize(InvocationContext ic) throws Exception {
        try {
            if ( !user.isBanned() ) {
                log.fine("Authorized");
                return ic.proceed();
            }
            else {
                log.fine("Not authorized");
                throw new NotAuthorizedException();
            }
        }
        catch (NotAuthenticatedException nae) {
            log.fine("Not authenticated");
            throw nae;
        }
    }

}

The @Interceptor annotation, defined in Declaring the interceptor bindings of an interceptor, identifies the AuthorizationInterceptor class as an interceptor. The @Secure annotation is a custom interceptor binding type, as defined in Interceptor binding types.

@Inherited
@InterceptorBinding
@Target({TYPE, METHOD})
@Retention(RUNTIME)
public @interface Secure {}

The @Secure annotation is used to apply the interceptor to a bean:

@Model
public class DocumentEditor {

    @Inject Document document;
    @Inject @LoggedIn User user;
    @Inject @Documents EntityManager em;

    @Secure
    public void save() {
        document.setCreatedBy(currentUser);
        em.persist(document);
    }

}

When the save() method is invoked, the authorize() method of the interceptor will be called. The invocation will proceed to the DocumentEditor class only if the authorization check is successful.

1.3.7. Decorator example

Decorators are similar to interceptors, but apply only to beans of a particular Java interface. Like interceptors, decorators may be easily enabled or disabled at deployment time. Unlike interceptors, decorators are aware of the semantics of the intercepted method.

For example, the DataAccess interface might be implemented by many beans:

public interface DataAccess<T, V> {

    public V getId(T object);
    public T load(V id);
    public void save(T object);
    public void delete(T object);

    public Class<T> getDataType();

}

The DataAccessAuthorizationDecorator class defines the authorization checks:

@Decorator
public abstract class DataAccessAuthorizationDecorator<T, V> implements DataAccess<T, V> {

    @Inject @Delegate DataAccess<T, V> delegate;

    @Inject Logger log;
    @Inject Set<Permission> permissions;

    public void save(T object) {
        authorize(SecureAction.SAVE, object);
        delegate.save(object);
    }

    public void delete(T object) {
        authorize(SecureAction.DELETE, object);
        delegate.delete(object);
    }

    private void authorize(SecureAction action, T object) {
        V id = delegate.getId(object);
        Class<T> type = delegate.getDataType();
        if ( permissions.contains( new Permission(action, type, id) ) ) {
            log.fine("Authorized for " + action);
        }
        else {
            log.fine("Not authorized for " + action);
            throw new NotAuthorizedException(action);
        }
    }

}

The @Decorator annotation defined in Declaring a decorator identifies the DataAccessAuthorizationDecorator class as a decorator. The @Delegate annotation defined in Decorator delegate injection points identifies the delegate, which the decorator uses to delegate method calls to the container. The decorator applies to any bean that implements DataAccess.

The decorator intercepts invocations just like an interceptor. However, unlike an interceptor, the decorator contains functionality that is specific to the semantics of the method being called.

Decorators may be declared abstract, relieving the developer of the responsibility of implementing all methods of the decorated interface. If a decorator does not implement a method of a decorated interface, the decorator will simply not be called when that method is invoked upon the decorated bean.

Part I - Core CDI

2. Concepts

A bean is a source of contextual objects which define application state and/or logic. These objects are called contextual instances of the bean. The container creates and destroys these instances and associates them with the appropriate context. Contextual instances of a bean may be injected into other objects (including other bean instances) that execute in the same context. A bean may bear metadata defining its lifecycle and interactions with other beans.

A bean comprises the following attributes:

  • A (nonempty) set of bean types

  • A (nonempty) set of qualifiers

  • A scope

  • Optionally, a bean name

  • A set of interceptor bindings

  • A bean implementation

Furthermore, a bean may or may not be an alternative.

A bean developer provides the bean implementation by writing business logic in Java code. The developer then defines the remaining attributes by explicitly annotating the bean class, or by allowing them to be defaulted by the container, as specified in Programming model.

The bean types and qualifiers of a bean determine where its instances will be injected by the container, as defined in Dependency injection and lookup.

The bean developer may also create interceptors and/or decorators or reuse existing interceptors and/or decorators. The interceptor bindings of a bean determine which interceptors will be applied at runtime. The bean types and qualifiers of a bean determine which decorators will be applied at runtime. Interceptors are defined by Java interceptors specification, and interceptor bindings are specified in Interceptor bindings. Decorators are defined in Decorators.

2.1. Functionality provided by the container to the bean

A bean is provided by the container with the following capabilities:

2.2. Bean types

A bean type defines a client-visible type of the bean. A bean may have multiple bean types. For example, the following bean has four bean types:

public class BookShop
        extends Business
        implements Shop<Book> {
    ...
}

The bean types are BookShop, Business, Shop<Book> and Object.

The rules for determining the (unrestricted) set of bean types for a bean are defined in Bean types of a managed bean, Bean types of a producer method and Bean types of a producer field.

All beans have the bean type java.lang.Object.

The bean types of a bean are used by the rules of typesafe resolution defined in Typesafe resolution.

Almost any Java type may be a bean type of a bean:

  • A bean type may be an interface, a concrete class or an abstract class, and may be declared final or have final methods.

  • A bean type may be a parameterized type with actual type parameters and type variables.

  • A bean type may be an array type. Two array types are considered identical only if the element type is identical.

  • A bean type may be a primitive type. Primitive types are considered to be identical to their corresponding wrapper types in java.lang.

  • A bean type may be a raw type.

However, some Java types are not legal bean types :

  • A type variable is not a legal bean type.

  • A parameterized type that contains a wildcard type parameter is not a legal bean type.

  • An array type whose component type is not a legal bean type.

Note that certain additional restrictions are specified in Unproxyable bean types for beans with a normal scope, as defined in Normal scopes and pseudo-scopes.

2.2.2. Restricting the bean types of a bean

The bean types of a bean may be restricted by annotating the bean class or producer method or field with the annotation @javax.enterprise.inject.Typed.

@Typed(Shop.class)
public class BookShop
        extends Business
        implements Shop<Book> {
    ...
}

When a @Typed annotation is explicitly specified, only the types whose classes are explicitly listed using the value member, together with java.lang.Object, are bean types of the bean.

In the example, the bean has a two bean types: Shop<Book> and Object.

If a bean class or producer method or field specifies a @Typed annotation, and the value member specifies a class which does not correspond to a type in the unrestricted set of bean types of a bean, the container automatically detects the problem and treats it as a definition error.

2.2.3. Typecasting between bean types

A client of a bean may typecast its contextual reference to a bean to any bean type of the bean which is a Java interface. However, the client may not in general typecast its contextual reference to an arbitrary concrete bean type of the bean. For example, if our managed bean was injected to the following field:

@Inject Business biz;

Then the following typecast is legal:

Shop<Book> bookShop = (Shop<Book>) biz;

However, the following typecast is not legal and might result in an exception at runtime:

BookShop bookShop = (BookShop) biz;

2.3. Qualifiers

For a given bean type, there may be multiple beans which implement the type. For example, an application may have two implementations of the interface PaymentProcessor:

class SynchronousPaymentProcessor
        implements PaymentProcessor {
    ...
}
class AsynchronousPaymentProcessor
        implements PaymentProcessor {
    ...
}

A client that needs a PaymentProcessor that processes payments synchronously needs some way to distinguish between the two different implementations. One approach would be for the client to explicitly specify the class that implements the PaymentProcessor interface. However, this approach creates a hard dependence between client and implementation - exactly what use of the interface was designed to avoid!

A qualifier type represents some client-visible semantic associated with a type that is satisfied by some implementations of the type (and not by others). For example, we could introduce qualifier types representing synchronicity and asynchronicity. In Java code, qualifier types are represented by annotations.

@Synchronous
class SynchronousPaymentProcessor
        implements PaymentProcessor {
    ...
}
@Asynchronous
class AsynchronousPaymentProcessor
        implements PaymentProcessor {
    ...
}

Finally, qualifier types are applied to injection points to distinguish which implementation is required by the client. For example, when the container encounters the following injected field, an instance of SynchronousPaymentProcessor will be injected:

@Inject @Synchronous PaymentProcessor paymentProcessor;

But in this case, an instance of AsynchronousPaymentProcessor will be injected:

@Inject @Asynchronous PaymentProcessor paymentProcessor;

The container inspects the qualifier annotations and type of the injected attribute to determine the bean instance to be injected, according to the rules of typesafe resolution defined in Typesafe resolution.

An injection point may even specify multiple qualifiers.

Qualifier types are also used as event selectors by event consumers, as defined in Events, and to bind decorators to beans, as specified in Decorators.

2.3.1. Built-in qualifier types

Three standard qualifier types are defined in the package javax.enterprise.inject. In addition, the built-in qualifier type @Named is defined by the package javax.inject.

Every bean has the built-in qualifier @Any, even if it does not explicitly declare this qualifier, except for the special @New qualified beans defined in @New qualified beans.

If a bean does not explicitly declare a qualifier other than @Named, the bean has exactly one additional qualifier, of type @Default. This is called the default qualifier.

The following declarations are equivalent:

@Default
public class Order { ... }
public class Order { ... }

Both declarations result in a bean with two qualifiers: @Any and @Default.

The following declaration results in a bean with three qualifiers: @Any, @Default and @Named("ord").

@Named("ord")
public class Order { ... }

The default qualifier is also assumed for any injection point that does not explicitly declare a qualifier, as defined in The default qualifier at injection points. The following declarations, in which the use of the @Inject annotation identifies the constructor parameter as an injection point, are equivalent:

public class Order {
    @Inject
    public Order(@Default OrderProcessor processor) { ... }
}
public class Order {
    @Inject
    public Order(OrderProcessor processor) { ... }
}

2.3.2. Defining new qualifier types

A qualifier type is a Java annotation defined as @Retention(RUNTIME). Typically a qualifier type is defined as @Target({METHOD, FIELD, PARAMETER, TYPE}).

A qualifier type may be declared by specifying the @javax.inject.Qualifier meta-annotation.

@Qualifier
@Retention(RUNTIME)
@Target({METHOD, FIELD, PARAMETER, TYPE})
public @interface Synchronous {}
@Qualifier
@Retention(RUNTIME)
@Target({METHOD, FIELD, PARAMETER, TYPE})
public @interface Asynchronous {}

A qualifier type may define annotation members.

@Qualifier
@Retention(RUNTIME)
@Target({METHOD, FIELD, PARAMETER, TYPE})
public @interface PayBy {
    PaymentMethod value();
}

2.3.3. Declaring the qualifiers of a bean

The qualifiers of a bean are declared by annotating the bean class or producer method or field with the qualifier types.

@LDAP
class LdapAuthenticator
        implements Authenticator {
    ...
}
public class Shop {

   @Produces @All
   public List<Product> getAllProducts() { ... }

   @Produces @WishList
   public List<Product> getWishList() { ... }

}

Any bean may declare multiple qualifier types.

@Synchronous @Reliable
class SynchronousReliablePaymentProcessor
        implements PaymentProcessor {
    ...
}

2.3.4. Specifying qualifiers of an injected field

Qualifier types may be applied to injected fields (see Injected fields) to determine the bean that is injected, according to the rules of typesafe resolution defined in Typesafe resolution.

@Inject @LDAP Authenticator authenticator;

A bean may only be injected to an injection point if it has all the qualifiers of the injection point.

@Inject @Synchronous @Reliable PaymentProcessor paymentProcessor;
@Inject @All List<Product> catalog;
@Inject @WishList List<Product> wishList;

2.3.5. Specifying qualifiers of a method or constructor parameter

Qualifier types may be applied to parameters of producer methods, initializer methods, disposer methods, observer methods or bean constructors (see Programming model) to determine the bean instance that is passed when the method is called by the container. The container uses the rules of typesafe resolution defined in Typesafe resolution to determine values for these parameters.

For example, when the container encounters the following producer method, an instance of SynchronousPaymentProcessor will be passed to the first parameter and an instance of AsynchronousPaymentProcessor will be passed to the second parameter:

@Produces
PaymentProcessor getPaymentProcessor(@Synchronous PaymentProcessor sync,
                                     @Asynchronous PaymentProcessor async) {
    return isSynchronous() ? sync : async;
}

2.4. Scopes

Scoped objects, exist in a well-defined lifecycle context:

  • they may be automatically created when needed and then automatically destroyed when the context in which they were created ends, and

  • their state is automatically shared by clients that execute in the same context.

All beans have a scope. The scope of a bean determines the lifecycle of its instances, and which instances of the bean are visible to instances of other beans, as defined in Scopes and contexts. A scope type is represented by an annotation type.

For example, an object that represents the current user is represented by a session scoped object:

@Produces @SessionScoped User getCurrentUser() { ... }

An object that represents an order is represented by a conversation scoped object:

@ConversationScoped
public class Order { ... }

A list that contains the results of a search screen might be represented by a request scoped object:

@Produces @RequestScoped @Named("orders")
List<Order> getOrderSearchResults() { ... }

The set of scope types is extensible.

2.4.1. Built-in scope types

There are five standard scope types defined by this specification, all defined in the package javax.enterprise.context.

If an interceptor or decorator has any scope other than @Dependent, non-portable behavior results.

2.4.2. Defining new scope types

A scope type is a Java annotation defined as @Retention(RUNTIME). Typically a scope type is defined as @Target({TYPE, METHOD, FIELD}). All scope types must also specify the @javax.inject.Scope or @javax.enterprise.context.NormalScope meta-annotation.

A scope type must not have any attributes. If a scope type has attributes non-portable behavior results.

For example, the following annotation declares a "business process scope":

@Inherited
@NormalScope
@Target({TYPE, METHOD, FIELD})
@Retention(RUNTIME)
public @interface BusinessProcessScoped {}

Custom scopes are normally defined by portable extensions, which must also provide a context object, as defined in The Context interface, that implements the custom scope.

2.4.3. Declaring the bean scope

The scope of a bean is defined by annotating the bean class or producer method or field with a scope type.

A bean class or producer method or field may specify at most one scope type annotation. If a bean class or producer method or field specifies multiple scope type annotations, the container automatically detects the problem and treats it as a definition error.

public class Shop {

   @Produces @ApplicationScoped @All
   public List<Product> getAllProducts() { ... }

   @Produces @SessionScoped @WishList
   public List<Product> getWishList() { ..... }

}

Likewise, a bean with the custom business process scope may be declared by annotating it with the @BusinessProcessScoped annotation:

@BusinessProcessScoped
public class Order { ... }

Alternatively, a scope type may be specified using a stereotype annotation, as defined in Declaring the stereotypes for a bean.

2.4.4. Default scope

When no scope is explicitly declared by annotating the bean class or producer method or field the scope of a bean is defaulted.

The default scope for a bean which does not explicitly declare a scope depends upon its declared stereotypes:

  • If the bean does not declare any stereotype with a declared default scope, the default scope for the bean is @Dependent.

  • If all stereotypes declared by the bean that have some declared default scope have the same default scope, then that scope is the default scope for the bean.

  • If there are two different stereotypes declared by the bean that declare different default scopes, then there is no default scope and the bean must explicitly declare a scope. If it does not explicitly declare a scope, the container automatically detects the problem and treats it as a definition error.

If a bean explicitly declares a scope, any default scopes declared by stereotypes are ignored.

2.5. Default bean discovery mode

The default bean discovery mode for a bean archive is annotated, and such a bean archive is said to be an implicit bean archive as defined in Bean archives.

If the bean discovery mode is annotated then:

  • bean classes that don’t have bean defining annotation (as defined in Bean defining annotations) are not discovered, and

  • producer methods (as defined in Producer methods) whose bean class does not have a bean defining annotation are not discovered, and

  • producer fields (as defined in Producer fields) whose bean class does not have a bean defining annotation are not discovered, and

  • disposer methods (as defined in Disposer methods) whose bean class does not have a bean defining annotation are not discovered, and

  • observer methods (as defined in Declaring an observer method) whose bean class does not have a bean defining annotation are not discovered.

2.5.1. Bean defining annotations

A bean class may have a bean defining annotation, allowing it to be placed anywhere in an application, as defined in Bean archives. A bean class with a bean defining annotation is said to be an implicit bean.

The set of bean defining annotations contains:

  • @ApplicationScoped, @SessionScoped, @ConversationScoped and @RequestScoped annotations,

  • all other normal scope types,

  • @Interceptor and @Decorator annotations,

  • all stereotype annotations (i.e. annotations annotated with @Stereotype),

  • and the @Dependent scope annotation.

If one of these annotations is declared on a bean class, then the bean class is said to have a bean defining annotation. For example, this dependent scoped bean has a bean defining annotation:

@Dependent
public class BookShop
        extends Business
        implements Shop<Book> {
    ...
}

whilst this dependent scoped bean does not have a bean defining annotation:

public class CoffeeShop
        extends Business
        implements Shop<Coffee> {
    ...
}

Note that to ensure compatibility with other JSR-330 implementations, all pseudo-scope annotations except @Dependent are not bean defining annotations. However, a stereotype annotation including a pseudo-scope annotation is a bean defining annotation.

2.6. Bean names

A bean may have a bean name. A bean with a name may be referred to by its name when used in a non typesafe environment (like the Unified Expression Language). A valid bean name is a period-separated list of valid EL identifiers.

The following strings are valid bean names:

com.acme.settings
orderManager

Subject to the restrictions defined in Ambiguous names, multiple beans may share the same bean name.

Bean names are used by the rules of bean name resolution defined in Name resolution.

2.6.1. Declaring the bean name

To specify the name of a bean, the qualifier @javax.inject.Named is applied to the bean class or producer method or field. This bean is named currentOrder:

@Named("currentOrder")
public class Order { ... }

2.6.2. Default bean names

In the following circumstances, a default name must be assigned by the container:

  • A bean class or producer method or field of a bean declares a @Named annotation and no bean name is explicitly specified by the value member.

  • A bean declares a stereotype that declares an empty @Named annotation, and the bean does not explicitly specify a bean name.

The default name for a bean depends upon the kind of the bean. The rules for determining the default name for a bean are defined in Default bean name for a managed bean, Default bean name for a producer method and Default bean name for a producer field.

2.6.3. Beans with no name

If @Named is not declared by the bean, nor by its stereotypes, a bean has no name.

If an interceptor or decorator has a name, non-portable behavior results.

2.7. Alternatives

An alternative is a bean that must be explicitly selected if it should be available for lookup, injection or name resolution.

2.7.1. Declaring an alternative

An alternative may be declared by annotating the bean class or producer method or field with the @Alternative annotation.

@Alternative
public class MockOrder extends Order { ... }

Alternatively, an alternative may be declared by annotating a bean, producer method or producer field with a stereotype that declares an @Alternative annotation.

If an interceptor or decorator is an alternative, non-portable behavior results.

2.8. Stereotypes

In many systems, use of architectural patterns produces a set of recurring bean roles. A stereotype allows a framework developer to identify such a role and declare some common metadata for beans with that role in a central place.

A stereotype encapsulates any combination of:

  • a default scope, and

  • a set of interceptor bindings.

A stereotype may also specify that:

  • all beans with the stereotype have defaulted bean names, or that

  • all beans with the stereotype are alternatives.

A bean may declare zero, one or multiple stereotypes.

2.8.1. Defining new stereotypes

A bean stereotype is a Java annotation defined as @Retention(RUNTIME). Typically a bean stereotype is defined as @Target({TYPE, METHOD, FIELD}), @Target(TYPE), @Target(METHOD), @Target(FIELD) or @Target({METHOD, FIELD}).

A stereotype may be declared by specifying the @javax.enterprise.inject.Stereotype meta-annotation.

@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface Action {}
Declaring the default scope for a stereotype

The default scope of a stereotype is defined by annotating the stereotype with a scope type. A stereotype may declare at most one scope. If a stereotype declares more than one scope, the container automatically detects the problem and treats it as a definition error.

For example, the following stereotype might be used to identify action classes in a web application:

@RequestScoped
@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface Action {}

Then actions would have scope @RequestScoped unless the scope is explicitly specified by the bean.

Specifying interceptor bindings for a stereotype

The interceptor bindings of a stereotype are defined by annotating the stereotype with the interceptor binding types. A stereotype may declare zero, one or multiple interceptor bindings, as defined in Interceptor bindings for stereotypes.

We may specify interceptor bindings that apply to all actions:

@RequestScoped
@Secure
@Transactional
@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface Action {}
Declaring a @Named stereotype

A stereotype may declare an empty @Named annotation, which specifies that every bean with the stereotype has a defaulted name when a name is not explicitly specified by the bean. A @Named qualifier declared by a stereotype is not added to the qualifiers of a bean with the stereotype.

If a stereotype declares a non-empty @Named annotation, the container automatically detects the problem and treats it as a definition error.

We may specify that all actions have bean names:

@RequestScoped
@Secure
@Transactional
@Named
@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface Action {}

A stereotype should not declare any qualifier annotation other than @Named. If a stereotype declares any other qualifier annotation, non-portable behavior results.

A stereotype should not be annotated @Typed. If a stereotype is annotated @Typed, non-portable behavior results.

Declaring an @Alternative stereotype

A stereotype may declare an @Alternative annotation, which specifies that every bean with the stereotype is an alternative.

We may specify that all mock objects are alternatives:

@Alternative
@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface Mock {}
Stereotypes with additional stereotypes

A stereotype may declare other stereotypes.

@Auditable
@Action
@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface AuditableAction {}

Stereotype declarations are transitive - a stereotype declared by a second stereotype is inherited by all beans and other stereotypes that declare the second stereotype.

Stereotypes declared @Target(TYPE) may not be applied to stereotypes declared @Target({TYPE, METHOD, FIELD}), @Target(METHOD), @Target(FIELD) or @Target({METHOD, FIELD}).

2.8.2. Declaring the stereotypes for a bean

Stereotype annotations may be applied to a bean class or producer method or field.

@Action
public class LoginAction { ... }

The default scope declared by the stereotype may be overridden by the bean:

@Mock @ApplicationScoped @Action
public class MockLoginAction extends LoginAction { ... }

Multiple stereotypes may be applied to the same bean:

@Dao @Action
public class LoginAction { ... }

2.8.3. Built-in stereotypes

The built-in stereotype @javax.enterprise.inject.Model is intended for use with beans that define the model layer of an MVC web application architecture such as JSF:

@Named
@RequestScoped
@Stereotype
@Target({TYPE, METHOD, FIELD})
@Retention(RUNTIME)
public @interface Model {}

In addition, the special-purpose @Interceptor and @Decorator stereotypes are defined in Declaring the interceptor bindings of an interceptor and Declaring a decorator.

2.9. Problems detected automatically by the container

When the application violates a rule defined by this specification, the container automatically detects the problem. There are three kinds of problem:

  • Definition errors - occur when a single bean definition violates the rules of this specification. If a definition error exists, the container must throw a subclass of javax.enterprise.inject.spi.DefinitionException.

  • Deployment problems - occur when there are problems resolving dependencies, or inconsistent specialization, in a particular deployment. If a deployment problem occurs, the container must throw a subclass of javax.enterprise.inject.spi.DeploymentException.

  • Exceptions - occur at runtime

Definition errors are developer errors. They may be detected by tooling at development time, and are also detected by the container at initialization time. If a definition error exists in a deployment, initialization will be aborted by the container.

Deployment problems are detected by the container at initialization time. If a deployment problem exists in a deployment, initialization will be aborted by the container.

The container is permitted to define a non-portable mode, for use at development time, in which some definition errors and deployment problems do not cause application initialization to abort.

Exceptions represent problems that may not be detected until they actually occur at runtime. All exceptions defined by this specification are unchecked exceptions. All exceptions defined by this specification may be safely caught and handled by the application.

3. Programming model

The container provides built-in support for injection and contextual lifecycle management of the following kinds of bean:

  • Managed beans

  • Producer methods and fields

All containers must support managed beans, producer methods and producer fields.

A portable extension may provide other kinds of beans by implementing the interface Bean defined in The Bean interface.

3.1. Managed beans

A managed bean is a bean that is implemented by a Java class. This class is called the bean class of the managed bean. The basic lifecycle and semantics of managed beans are defined by the Managed Beans specification.

If the bean class of a managed bean is annotated with both @Interceptor and @Decorator, the container automatically detects the problem and treats it as a definition error.

If a managed bean has a non-static public field, it must have scope @Dependent. If a managed bean with a non-static public field declares any scope other than @Dependent, the container automatically detects the problem and treats it as a definition error.

If the managed bean class is a generic type, it must have scope @Dependent. If a managed bean with a parameterized bean class declares any scope other than @Dependent, the container automatically detects the problem and treats it as a definition error.

3.1.1. Which Java classes are managed beans?

A top-level Java class is a managed bean if it meets all of the following conditions:

  • It is not a non-static inner class.

  • It is a concrete class, or is annotated @Decorator.

  • It does not implement javax.enterprise.inject.spi.Extension.

  • It is not annotated @Vetoed or in a package annotated @Vetoed.

  • It has an appropriate constructor - either:

    • the class has a constructor with no parameters, or

    • the class declares a constructor annotated @Inject.

All Java classes that meet these conditions are managed beans and thus no special declaration is required to define a managed bean.

If packages annotated @Vetoed are split across classpath entries, non-portable behavior results. An application can prevent packages being split across jars by sealing the package as defined by the Extension Mechanism Architecture.

3.1.2. Bean types of a managed bean

The unrestricted set of bean types for a managed bean contains the bean class, every superclass and all interfaces it implements directly or indirectly.

Note the additional restrictions upon bean types of beans with normal scopes defined in Unproxyable bean types.

3.1.3. Declaring a managed bean

A managed bean with a constructor that takes no parameters does not require any special annotations. The following classes are beans:

public class Shop { .. }
class PaymentProcessorImpl implements PaymentProcessor { ... }

If the managed bean does not have a constructor that takes no parameters, it must have a constructor annotated @Inject. No additional special annotations are required.

A bean class may specify a scope, bean name, stereotypes and/or qualifiers:

@ConversationScoped @Default
public class ShoppingCart { ... }

A managed bean may extend another managed bean:

@Named("loginAction")
public class LoginAction { ... }
@Mock
@Named("loginAction")
public class MockLoginAction extends LoginAction { ... }

The second bean is a "mock object" that overrides the implementation of LoginAction when running in an embedded CDI based integration testing environment.

3.1.4. Specializing a managed bean

If a bean class of a managed bean X is annotated @Specializes, then the bean class of X must directly extend the bean class of another managed bean Y. Then X directly specializes Y, as defined in Specialization.

If the bean class of X does not directly extend the bean class of another managed bean, the container automatically detects the problem and treats it as a definition error.

For example, MockLoginAction directly specializes LoginAction:

public class LoginAction { ... }
@Mock @Specializes
public class MockLoginAction extends LoginAction { ... }

3.1.5. Default bean name for a managed bean

The default name for a managed bean is the unqualified class name of the bean class, after converting the first character to lower case.

For example, if the bean class is named ProductList, the default bean name is productList.

3.2. Producer methods

A producer method acts as a source of objects to be injected, where:

  • the objects to be injected are not required to be instances of beans, or

  • the concrete type of the objects to be injected may vary at runtime, or

  • the objects require some custom initialization that is not performed by the bean constructor.

A producer method must be a default-access, public, protected or private, non-abstract method of a managed bean class. A producer method may be either static or non-static.

If a producer method sometimes returns a null value, then the producer method must have scope @Dependent. If a producer method returns a null value at runtime, and the producer method declares any other scope, an IllegalProductException is thrown by the container. This restriction allows the container to use a client proxy, as defined in Client proxies.

If the producer method return type is a parameterized type, it must specify an actual type parameter or type variable for each type parameter.

If a producer method return type contains a wildcard type parameter or is an array type whose component type contains a wildcard type parameter, the container automatically detects the problem and treats it as a definition error.

If the producer method return type is a parameterized type with a type variable, it must have scope @Dependent. If a producer method with a parameterized return type with a type variable declares any scope other than @Dependent, the container automatically detects the problem and treats it as a definition error.

If a producer method return type is a type variable or an array type whose component type is a type variable the container automatically detects the problem and treats it as a definition error.

The application may call producer methods directly. However, if the application calls a producer method directly, no parameters will be passed to the producer method by the container; the returned object is not bound to any context; and its lifecycle is not managed by the container.

A bean may declare multiple producer methods.

3.2.1. Bean types of a producer method

The bean types of a producer method depend upon the method return type:

  • If the return type is an interface, the unrestricted set of bean types contains the return type, all interfaces it extends directly or indirectly and java.lang.Object.

  • If a return type is primitive or is a Java array type, the unrestricted set of bean types contains exactly two types: the method return type and java.lang.Object.

  • If the return type is a class, the unrestricted set of bean types contains the return type, every superclass and all interfaces it implements directly or indirectly.

Note the additional restrictions upon bean types of beans with normal scopes defined in Unproxyable bean types.

3.2.2. Declaring a producer method

A producer method may be declared by annotating a method with the @javax.enterprise.inject.Produces annotation.

public class Shop {
   @Produces PaymentProcessor getPaymentProcessor() { ... }
   @Produces List<Product> getProducts() { ... }
}

A producer method may also specify scope, bean name, stereotypes and/or qualifiers.

public class Shop {
   @Produces @ApplicationScoped @Catalog @Named("catalog")
   List<Product> getProducts() { ... }
}

If a producer method is annotated @Inject, has a parameter annotated @Disposes, or has a parameter annotated @Observes, the container automatically detects the problem and treats it as a definition error.

Interceptors and decorators may not declare producer methods. If an interceptor or decorator has a method annotated @Produces, the container automatically detects the problem and treats it as a definition error.

A producer method may have any number of parameters. All producer method parameters are injection points.

public class OrderFactory {

   @Produces @ConversationScoped
   public Order createCurrentOrder(Shop shop, @Selected Product product) {
       Order order = new Order(product, shop);
       return order;
   }

}

3.2.3. Specializing a producer method

If a producer method X is annotated @Specializes, then it must be non-static and directly override another producer method Y. Then X directly specializes Y, as defined in Specialization.

If the method is static or does not directly override another producer method, the container automatically detects the problem and treats it as a definition error.

@Mock
public class MockShop extends Shop {

   @Override @Specializes
   @Produces
   PaymentProcessor getPaymentProcessor() {
      return new MockPaymentProcessor();
   }

   @Override @Specializes
   @Produces
   List<Product> getProducts() {
      return PRODUCTS;
   }

   ...

}

3.2.4. Default bean name for a producer method

The default name for a producer method is the method name, unless the method follows the JavaBeans property getter naming convention, in which case the default name is the JavaBeans property name.

For example, this producer method is named products:

@Produces @Named
public List<Product> getProducts() { ... }

This producer method is named paymentProcessor:

@Produces @Named
public PaymentProcessor paymentProcessor() { ... }

3.3. Producer fields

A producer field is a slightly simpler alternative to a producer method.

A producer field must be a default-access, public, protected or private, field of a managed bean class. A producer field may be either static or non-static.

If a producer field sometimes contains a null value when accessed, then the producer field must have scope @Dependent. If a producer field contains a null value at runtime, and the producer field declares any other scope, an IllegalProductException is thrown by the container. This restriction allows the container to use a client proxy, as defined in Client proxies.

If the producer field type is a parameterized type, it must specify an actual type parameter or type variable for each type parameter.

If a producer field type contains a wildcard type parameter or is an array type whose component type contains a wildcard parameter, the container automatically detects the problem and treats it as a definition error.

If the producer field type is a parameterized type with a type variable, it must have scope @Dependent. If a producer field with a parameterized type with a type variable declares any scope other than @Dependent, the container automatically detects the problem and treats it as a definition error.

If a producer field type is a type variable or is an array type whose component type is a type variable the container automatically detects the problem and treats it as a definition error.

The application may access producer fields directly. However, if the application accesses a producer field directly, the returned object is not bound to any context; and its lifecycle is not managed by the container.

A bean may declare multiple producer fields.

3.3.1. Bean types of a producer field

The bean types of a producer field depend upon the field type:

  • If the field type is an interface, the unrestricted set of bean types contains the field type, all interfaces it extends directly or indirectly and java.lang.Object.

  • If a field type is primitive or is a Java array type, the unrestricted set of bean types contains exactly two types: the field type and java.lang.Object.

  • If the field type is a class, the unrestricted set of bean types contains the field type, every superclass and all interfaces it implements directly or indirectly.

Note the additional restrictions upon bean types of beans with normal scopes defined in Unproxyable bean types.

3.3.2. Declaring a producer field

A producer field may be declared by annotating a field with the @javax.enterprise.inject.Produces annotation.

public class Shop {
   @Produces PaymentProcessor paymentProcessor = ....;
   @Produces List<Product> products = ....;
}

A producer field may also specify scope, bean name, stereotypes and/or qualifiers.

public class Shop {
   @Produces @ApplicationScoped @Catalog @Named("catalog")
   List<Product> products = ....;
}

If a producer field is annotated @Inject, the container automatically detects the problem and treats it as a definition error.

Interceptors and decorators may not declare producer fields. If an interceptor or decorator has a field annotated @Produces, the container automatically detects the problem and treats it as a definition error.

3.3.3. Default bean name for a producer field

The default name for a producer field is the field name.

For example, this producer field is named products:

@Produces @Named
public List<Product> products = ...;

3.4. Disposer methods

A disposer method allows the application to perform customized cleanup of an object returned by a producer method or producer field.

A disposer method must be a default-access, public, protected or private, non-abstract method of a managed bean class. A disposer method may be either static or non-static.

A bean may declare multiple disposer methods.

3.4.1. Disposed parameter of a disposer method

Each disposer method must have exactly one disposed parameter, of the same type as the corresponding producer method return type or producer field type. When searching for disposer methods for a producer method or producer field the container considers the type and qualifiers of the disposed parameter. If a producer method or producer field declared by the same bean class is assignable to the disposed parameter, according to the rules of typesafe resolution defined in Typesafe resolution, the container must call this method when destroying any instance returned by that producer method or producer field.

A disposer method may resolve to multiple producer methods or producer fields declared by the bean class, in which case the container must call it when destroying any instance returned by any of these producer methods or producer fields.

3.4.2. Declaring a disposer method

A disposer method may be declared by annotating a parameter @javax.enterprise.inject.Disposes. That parameter is the disposed parameter. Qualifiers may be declared by annotating the disposed parameter:

public class UserDatabaseEntityManager {

    @Produces @ConversationScoped @UserDatabase
    public EntityManager create(EntityManagerFactory emf) {
        return emf.createEntityManager();
    }

    public void close(@Disposes @UserDatabase EntityManager em) {
        em.close();
    }

}
public class Resources {

    @PersistenceContext
    @Produces @UserDatabase
    private EntityManager em;

    public void close(@Disposes @UserDatabase EntityManager em) {
        em.close();
    }

}

If a method has more than one parameter annotated @Disposes, the container automatically detects the problem and treats it as a definition error.

If a disposer method is annotated @Produces or @Inject or has a parameter annotated @Observes, the container automatically detects the problem and treats it as a definition error.

Interceptors and decorators may not declare disposer methods. If an interceptor or decorator has a method annotated @Disposes, the container automatically detects the problem and treats it as a definition error.

In addition to the disposed parameter, a disposer method may declare additional parameters, which may also specify qualifiers. These additional parameters are injection points.

public void close(@Disposes @UserDatabase EntityManager em, Logger log) { ... }

3.4.3. Disposer method resolution

A disposer method is bound to a producer method or producer field if:

If there are multiple disposer methods for a single producer method or producer field, the container automatically detects the problem and treats it as a definition error.

If there is no producer method or producer field declared by the bean class that is assignable to the disposed parameter of a disposer method, the container automatically detects the problem and treats it as a definition error.

3.5. Bean constructors

When the container instantiates a bean class, it calls the bean constructor. The bean constructor is a default-access, public, protected or private constructor of the bean class.

The application may call bean constructors directly. However, if the application directly instantiates the bean, no parameters are passed to the constructor by the container; the returned object is not bound to any context; no dependencies are injected by the container; and the lifecycle of the new instance is not managed by the container.

3.5.1. Declaring a bean constructor

The bean constructor may be identified by annotating the constructor @Inject.

@SessionScoped
public class ShoppingCart implements Serializable {

   private User customer;

   @Inject
   public ShoppingCart(User customer) {
       this.customer = customer;
   }

   public ShoppingCart(ShoppingCart original) {
       this.customer = original.customer;
   }

   ShoppingCart() {}

   ...

}
@ConversationScoped
public class Order {

   private Product product;
   private User customer;

   @Inject
   public Order(@Selected Product product, User customer) {
       this.product = product;
       this.customer = customer;
   }

   public Order(Order original) {
       this.product = original.product;
       this.customer = original.customer;
   }

   Order() {}

   ...

}

If a bean class does not explicitly declare a constructor using @Inject, the constructor that accepts no parameters is the bean constructor.

If a bean class has more than one constructor annotated @Inject, the container automatically detects the problem and treats it as a definition error.

If a bean constructor has a parameter annotated @Disposes, or @Observes, the container automatically detects the problem and treats it as a definition error.

A bean constructor may have any number of parameters. All parameters of a bean constructor are injection points.

3.6. Injected fields

An injected field is a non-static, non-final field of a bean class or of any other classes supporting injection.

3.6.1. Declaring an injected field

An injected field may be declared by annotating the field @javax.inject.Inject.

@ConversationScoped
public class Order {

   @Inject @Selected Product product;
   @Inject User customer;

}

If an injected field is annotated @Produces, the container automatically detects the problem and treats it as a definition error.

3.7. Initializer methods

An initializer method is a default-access, public, protected or private, non-abstract, non-static, non-generic method of a bean class or of any other classes supporting injection.

A bean class may declare multiple (or zero) initializer methods.

Method interceptors are never called when the container calls an initializer method.

The application may call initializer methods directly, but then no parameters will be passed to the method by the container.

3.7.1. Declaring an initializer method

An initializer method may be declared by annotating the method @javax.inject.Inject.

@ConversationScoped
public class Order {

   private Product product;
   private User customer;

   @Inject
   void setProduct(@Selected Product product) {
       this.product = product;
   }

   @Inject
   public void setCustomer(User customer) {
       this.customer = customer;
   }

}

If a generic method of a bean is annotated @Inject, the container automatically detects the problem and treats it as a definition error.

If an initializer method is annotated @Produces, has a parameter annotated @Disposes, or has a parameter annotated @Observes, the container automatically detects the problem and treats it as a definition error.

An initializer method may have any number of parameters. All initializer method parameters are injection points.

3.8. The default qualifier at injection points

If an injection point declares no qualifier, the injection point has exactly one qualifier, the default qualifier @Default.

The following are equivalent:

@ConversationScoped
public class Order {

   private Product product;
   private User customer;

   @Inject
   public void init(@Selected Product product, User customer) {
       this.product = product;
       this.customer = customer;
   }

}
@ConversationScoped
public class Order {

   private Product product;
   private User customer;

   @Inject
   public void init(@Selected Product product, @Default User customer) {
       this.product = product;
       this.customer = customer;
   }

}

The following definitions are equivalent:

public class Payment {

   public Payment(BigDecimal amount) { ... }

   @Inject Payment(Order order) {
      this(order.getAmount();
   }

}
public class Payment {

   public Payment(BigDecimal amount) { ... }

   @Inject Payment(@Default Order order) {
      this(order.getAmount();
   }

}

Finally, the following are equivalent:

@Inject Order order;
@Inject @Default Order order;

3.9. The qualifier @Named at injection points

The use of @Named as an injection point qualifier is not recommended, except in the case of integration with legacy code that uses string-based names to identify beans.

If an injected field declares a @Named annotation that does not specify the value member, the name of the field is assumed. For example, the following field has the qualifier @Named("paymentService"):

@Inject @Named PaymentService paymentService;

If any other injection point declares a @Named annotation that does not specify the value member, the container automatically detects the problem and treats it as a definition error.

3.10. @New qualified beans

The @New qualifier was deprecated in CDI 1.1. CDI applications are encouraged to inject @Dependent scoped beans instead.

For each managed bean, a second bean exists which:

  • has the same bean class,

  • has the same bean types,

  • has the same bean constructor, initializer methods and injected fields, and

  • has the same interceptor bindings.

However, this second bean:

  • has scope @Dependent,

  • has exactly one qualifier: @javax.enterprise.inject.New(X.class) where X is the bean class,

  • has no bean name,

  • has no stereotypes,

  • has no observer methods, producer methods or fields or disposer methods, and

  • is not an alternative, and

  • is enabled, in the sense of Enabled and disabled beans, if and only if some other enabled bean has an injection point with the qualifier @New(X.class) where X is the bean class.

This bean is called the @New qualified bean for the class X.

Note that this second bean exists - and may be enabled and available for injection - even if the first bean is disabled, as defined by Enabled and disabled beans, or if the bean class is deployed outside of a bean archive, as defined in Bean archives, and is therefore not discovered during the bean discovery process defined in Packaging and deployment. The container discovers @New qualified beans by inspecting injection points of other enabled beans.

This allows the application to obtain a new instance of a bean which is not bound to the declared scope, but has had dependency injection performed.

@Produces @ConversationScoped
@Special Order getSpecialOrder(@New(Order.class) Order order) {
    ...
    return order;
}

When the qualifier @New is specified at an injection point and no value member is explicitly specified, the container defaults the value to the declared type of the injection point. So the following injection point has qualifier @New(Order.class):

@Produces @ConversationScoped
@Special Order getSpecialOrder(@New Order order) { ... }

3.11. Unproxyable bean types

The container uses proxies to provide certain functionality. Certain legal bean types cannot be proxied by the container:

  • classes which don’t have a non-private constructor with no parameters,

  • classes which are declared final,

  • classes which have non-static, final methods with public, protected or default visibility,

  • primitive types,

  • and array types.

A bean type must be proxyable if an injection point resolves to a bean:

  • that requires a client proxy, or

  • that has an associated decorator, or

  • that has a bound interceptor.

Otherwise, the container automatically detects the problem, and treats it as a deployment problem.

4. Inheritance and specialization

A bean may inherit type-level metadata and members from its superclasses.

Inheritance of type-level metadata by beans from their superclasses is controlled via use of the Java @Inherited meta-annotation. Type-level metadata is never inherited from interfaces implemented by a bean.

Member-level metadata is not inherited. However, injected fields, initializer methods, lifecycle callback methods and non-static observer methods are inherited by beans from their superclasses.

The implementation of a bean may be extended by the implementation of a second bean. This specification recognizes two distinct scenarios in which this situation occurs:

  • The second bean specializes the first bean in certain deployment scenarios. In these deployments, the second bean completely replaces the first, fulfilling the same role in the system.

  • The second bean is simply reusing the Java implementation, and otherwise bears no relation to the first bean. The first bean may not even have been designed for use as a contextual object.

The two cases are quite dissimilar.

By default, Java implementation reuse is assumed. In this case, the two beans have different roles in the system, and may both be available in a particular deployment.

The bean developer may explicitly specify that the second bean specializes the first. Then the second bean inherits, and may not override, the qualifiers and bean name of the first bean. The second bean is able to serve the same role in the system as the first. In a particular deployment, only one of the two beans may fulfill that role.

4.1. Inheritance of type-level metadata

Suppose a class X is extended directly or indirectly by the bean class of a managed bean Y.

  • If X is annotated with a qualifier type, stereotype or interceptor binding type Z then Y inherits the annotation if and only if Z declares the @Inherited meta-annotation and neither Y nor any intermediate class that is a subclass of X and a superclass of Y declares an annotation of type Z. (This behavior is defined by the Java Language Specification.)

  • If X is annotated with a scope type Z then Y inherits the annotation if and only if Z declares the @Inherited meta-annotation and neither Y nor any intermediate class that is a subclass of X and a superclass of Y declares a scope type. (This behavior is different to what is defined in the Java Language Specification.)

A scope type explicitly declared by X and inherited by Y from X takes precedence over default scopes of stereotypes declared or inherited by Y.

For annotations defined by the application or third-party extensions, it is recommended that:

  • scope types should be declared @Inherited,

  • qualifier types should not be declared @Inherited,

  • interceptor binding types should be declared @Inherited, and

  • stereotypes may be declared @Inherited, depending upon the semantics of the stereotype.

All scope types, qualifier types, and interceptor binding types defined by this specification adhere to these recommendations.

The stereotypes defined by this specification are not declared @Inherited.

However, in special circumstances, these recommendations may be ignored.

Note that the @Named annotation is not declared @Inherited and bean names are not inherited unless specialization is used.

4.2. Inheritance of member-level metadata

Suppose a class X is extended directly or indirectly by the bean class of a managed bean Y.

  • If X declares an injected field x then Y inherits x. (This behavior is defined by the Common Annotations for the Java Platform specification.)

  • If X declares an initializer, non-static observer, @PostConstruct or @PreDestroy method x() then Y inherits x() if and only if neither Y nor any intermediate class that is a subclass of X and a superclass of Y overrides the method x(). (This behavior is defined by the Common Annotations for the Java Platform specification.)

  • If X declares a non-static method x() annotated with an interceptor binding type Z then Y inherits the binding if and only if neither Y nor any intermediate class that is a subclass of X and a superclass of Y overrides the method x(). (This behavior is defined by the Common Annotations for the Java Platform specification.)

  • If X declares a non-static producer or disposer method x() then Y does not inherit this method. (This behavior is different to what is defined in the Common Annotations for the Java Platform specification.)

  • If X declares a non-static producer field x then Y does not inherit this field. (This behavior is different to what is defined in the Common Annotations for the Java Platform specification.)

If X is a generic type, and an injection point or observer method declared by X is inherited by Y, and the declared type of the injection point or event parameter contains type variables declared by X, the type of the injection point or event parameter inherited in Y is the declared type, after substitution of actual type arguments declared by Y or any intermediate class that is a subclass of X and a superclass of Y.

For example, the bean DaoClient has an injection point of type Dao<T>.

public class DaoClient<T> {

    @Inject Dao<T> dao;
    ...

}

This injection point is inherited by UserDaoClient, but the type of the inherited injection point is Dao<User>.

public class UserDaoClient
        extends DaoClient<User> { ... }

4.3. Specialization

If two beans both support a certain bean type, and share at least one qualifier, then they are both eligible for injection to any injection point with that declared type and qualifier.

Consider the following beans:

@Default @Asynchronous
public class AsynchronousService implements Service {
    ...
}
@Alternative
public class MockAsynchronousService extends AsynchronousService {
    ...
}

Suppose that the MockAsynchronousService alternative is selected, as defined in Modularity:

@Alternative @Priority(APPLICATION+100)
public class MockAsynchronousService extends AsynchronousService {
    ...
}

Then, according to the rules of Unsatisfied and ambiguous dependencies, the following ambiguous dependency is resolvable, and so the attribute will receive an instance of MockAsynchronousService:

@Inject Service service;

However, the following attribute will receive an instance of AsynchronousService, even though MockAsynchronousService is a selected alternative, because MockAsynchronousService does not have the qualifier @Asynchronous:

@Inject @Asynchronous Service service;

This is a useful behavior in some circumstances, however, it is not always what is intended by the developer.

The only way one bean can completely override a second bean at all injection points is if it implements all the bean types and declares all the qualifiers of the second bean. However, if the second bean declares a producer method or observer method, then even this is not enough to ensure that the second bean is never called!

To help prevent developer error, the first bean may:

  • directly extend the bean class of the second bean, or

  • directly override the producer method, in the case that the second bean is a producer method, and then

explicitly declare that it specializes the second bean.

@Alternative @Specializes
public class MockAsynchronousService extends AsynchronousService {
    ...
}

When an enabled bean, as defined in Enabled and disabled beans, specializes a second bean, we can be certain that the second bean is never instantiated or called by the container. Even if the second bean defines a producer or observer method, the method will never be called.

4.3.1. Direct and indirect specialization

The annotation @javax.enterprise.inject.Specializes is used to indicate that one bean directly specializes another bean, as defined in Specializing a managed bean and Specializing a producer method.

Formally, a bean X is said to specialize another bean Y if either:

  • X directly specializes Y, or

  • a bean Z exists, such that X directly specializes Z and Z specializes Y.

Then X will inherit the qualifiers and bean name of Y:

  • the qualifiers of X include all qualifiers of Y, together with all qualifiers declared explicitly by X, and

  • if Y has a bean name, the bean name of X is the same as the bean name of Y.

Furthermore, X must have all the bean types of Y. If X does not have some bean type of Y, the container automatically detects the problem and treats it as a definition error.

If Y has a bean name and X declares a bean name explicitly the container automatically detects the problem and treats it as a definition error.

For example, the following bean would have the inherited qualifiers @Default and @Asynchronous:

@Mock @Specializes
public class MockAsynchronousService extends AsynchronousService {
    ...
}

If AsynchronousService declared a bean name:

@Default @Asynchronous @Named("asyncService")
public class AsynchronousService implements Service{
    ...
}

Then the bean name would also automatically be inherited by MockAsynchronousService.

If an interceptor or decorator is annotated @Specializes, non-portable behavior results.

5. Dependency injection and lookup

The container injects references to contextual instances to the following kinds of injection point:

  • Any injected field of a bean class

  • Any parameter of a bean constructor, bean initializer method, producer method or disposer method

  • Any parameter of an observer method, except for the event parameter

References to contextual instances may also be obtained by programmatic lookup.

In general, a bean type or bean name does not uniquely identify a bean. When resolving a bean at an injection point, the container considers bean type, qualifiers and selected alternatives. This allows bean developers to decouple type from implementation.

The container is required to support circularities in the bean dependency graph where at least one bean participating in every circular chain of dependencies has a normal scope, as defined in Normal scopes and pseudo-scopes. The container is not required to support circular chains of dependencies where every bean participating in the chain has a pseudo-scope.

5.1. Modularity

Beans and their clients may be deployed in modules in a module architecture. In a module architecture, certain modules are considered bean archives. The library is a bean archive if it contains a beans.xml file, as defined in Bean archives.

A bean packaged in a certain module is available for injection, lookup and name resolution to classes packaged in some other module if and only if the bean class of the bean is required to be accessible to the other module by the class accessibility requirements of the module architecture.

An alternative is not available for injection, lookup or name resolution to classes in a module unless the module is a bean archive and the alternative is explicitly selected for the bean archive or the application.

5.1.1. Declaring selected alternatives

This specification defines two methods of selecting alternatives. From Contexts and Dependency Injection 1.1 onwards the @Priority annotation allows an alternative to be selected for an entire application. Contexts and Dependency Injection 1.0 allowed only for an alternative to be selected for a bean archive.

Declaring selected alternatives for an application

An alternative may be given a priority for the application:

  • by placing the @Priority annotation on the bean class of a managed bean, or

  • by placing the @Priority annotation on the bean class that declares the producer method, field or resource.

Declaring selected alternatives for a bean archive

An alternative may be explicitly declared using the <alternatives> element of the beans.xml file of the bean archive. The <alternatives> element contains a list of bean classes and stereotypes. An alternative is selected for the bean archive if either:

  • the alternative is a managed bean and the bean class of the bean is listed,

  • the alternative is a producer method, field or resource, and the bean class that declares the method or field is listed, or

  • any @Alternative stereotype of the alternative is listed.

<beans xmlns="http://xmlns.jcp.org/xml/ns/javaee"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://xmlns.jcp.org/xml/ns/javaee http://xmlns.jcp.org/xml/ns/javaee/beans_1_1.xsd">
   <alternatives>
      <class>com.acme.myfwk.InMemoryDatabase</class>
      <stereotype>com.acme.myfwk.Mock</stereotype>
      <stereotype>com.acme.site.Australian</stereotype>
   </alternatives>
</beans>

Each child <class> element must specify the name of a bean class of an alternative bean. If there is no bean whose bean class has the specified name, or if no bean whose bean class has the specified name is an alternative, the container automatically detects the problem and treats it as a deployment problem.

Each child <stereotype> element must specify the name of an @Alternative stereotype annotation. If there is no annotation with the specified name, or the annotation is not an @Alternative stereotype, the container automatically detects the problem and treats it as a deployment problem.

If the same type is listed twice under the <alternatives> element, the container automatically detects the problem and treats it as a deployment problem.

For a custom implementation of the Bean interface defined in The Bean interface, the container calls isAlternative() to determine whether the bean is an alternative, and getBeanClass() and getStereotypes() to determine whether an alternative is selected in a certain bean archive.

5.1.2. Enabled and disabled beans

A bean is said to be enabled if:

  • it is deployed in a bean archive, and

  • it is not a producer method or field of a disabled bean, and

  • it is not specialized by any other enabled bean, as defined in Specialization, and either

  • it is not an alternative, or it is a selected alternative of at least one bean archive or the application.

Otherwise, the bean is said to be disabled.

Note that @New qualified beans defines a special rule that determines whether a @New qualified bean is enabled or disabled. This rule applies as only to @New qualified beans, as an exception to the normal rule defined here.

5.1.3. Inconsistent specialization

Suppose an enabled bean X specializes a second bean Y. If there is another enabled bean that specializes Y we say that inconsistent specialization exists. The container automatically detects inconsistent specialization and treats it as a deployment problem.

5.1.4. Inter-module injection

A bean is available for injection in a certain module if:

  • the bean is not an interceptor or decorator,

  • the bean is enabled,

  • the bean is either not an alternative, or the module is a bean archive and the bean is a selected alternative of the bean archive, or the bean is a selected alternative of the application, and

  • the bean class is required to be accessible to classes in the module, according to the class accessibility requirements of the module architecture.

For a custom implementation of the Bean interface defined in The Bean interface, the container calls getBeanClass() to determine the bean class of the bean and InjectionPoint.getMember() and then Member.getDeclaringClass() to determine the class that declares an injection point.

5.2. Typesafe resolution

The process of matching a bean to an injection point is called typesafe resolution. Typesafe resolution usually occurs at application initialization time, allowing the container to warn the user if any enabled beans have unsatisfied or unresolvable ambiguous dependencies.

5.2.1. Performing typesafe resolution

The container considers bean type and qualifiers when resolving a bean to be injected to an injection point. The type and qualifiers of the injection point are called the required type and required qualifiers.

A bean is assignable to a given injection point if:

  • The bean has a bean type that matches the required type. For this purpose, primitive types are considered to match their corresponding wrapper types in java.lang and array types are considered to match only if their element types are identical. Parameterized and raw types are considered to match if they are identical or if the bean type is assignable to the required type, as defined in Assignability of raw and parameterized types or Assignability of raw and parameterized types for delegate injection points.

  • The bean has all the required qualifiers. If no required qualifiers were explicitly specified, the container assumes the required qualifier @Default. A bean has a required qualifier if it has a qualifier with (a) the same type and (b) the same annotation member value for each member which is not annotated @javax.enterprise.util.Nonbinding.

A bean is eligible for injection to a certain injection point if:

For a custom implementation of the Bean interface defined in The Bean interface, the container calls getTypes() and getQualifiers() to determine the bean types and qualifiers.

5.2.2. Unsatisfied and ambiguous dependencies

An unsatisfied dependency exists at an injection point when no bean is eligible for injection to the injection point. An ambiguous dependency exists at an injection point when multiple beans are eligible for injection to the injection point.

Note that an unsatisfied or ambiguous dependency cannot exist for a decorator delegate injection point, defined in Decorator delegate injection points.

When an ambiguous dependency exists, the container attempts to resolve the ambiguity. The container eliminates all eligible beans that are not alternatives, except for producer methods and fields of beans that are alternatives. If:

  • there is exactly one bean remaining, the container will select this bean, and the ambiguous dependency is called resolvable.

  • all the beans left are alternatives with a priority, or producer methods or fields of beans that are alternatives with a priority, then the container will determine the highest priority value, and eliminate all beans, except for alternatives with the highest priority and producer methods and fields of alternatives with the highest priority value. If there is exactly one bean remaining, the container will select this bean, and the ambiguous dependency is called resolvable.

The container must validate all injection points of all enabled beans, all observer methods and all disposer methods when the application is initialized to ensure that there are no unsatisfied or unresolvable ambiguous dependencies. If an unsatisfied or unresolvable ambiguous dependency exists, the container automatically detects the problem and treats it as a deployment problem.

For a custom implementation of the Bean interface defined in The Bean interface, the container calls getInjectionPoints() to determine the set of injection points.

Any legal bean type, as defined in Legal bean types may be the required type of an injection point. Furthermore, the required type of an injection point may contain a wildcard type parameter. However, a type variable is not a legal injection point type.

If an injection point type is a type variable, the container automatically detects the problem and treats it as a definition error.

5.2.4. Assignability of raw and parameterized types

A parameterized bean type is considered assignable to a raw required type if the raw types are identical and all type parameters of the bean type are either unbounded type variables or java.lang.Object.

A parameterized bean type is considered assignable to a parameterized required type if they have identical raw type and for each parameter:

  • the required type parameter and the bean type parameter are actual types with identical raw type, and, if the type is parameterized, the bean type parameter is assignable to the required type parameter according to these rules, or

  • the required type parameter is a wildcard, the bean type parameter is an actual type and the actual type is assignable to the upper bound, if any, of the wildcard and assignable from the lower bound, if any, of the wildcard, or

  • the required type parameter is a wildcard, the bean type parameter is a type variable and the upper bound of the type variable is assignable to or assignable from the upper bound, if any, of the wildcard and assignable from the lower bound, if any, of the wildcard, or

  • the required type parameter is an actual type, the bean type parameter is a type variable and the actual type is assignable to the upper bound, if any, of the type variable, or

  • the required type parameter and the bean type parameter are both type variables and the upper bound of the required type parameter is assignable to the upper bound, if any, of the bean type parameter.

For example, Dao is eligible for injection to any injection point of type @Default Dao<Order>, @Default Dao<User>, @Default Dao<?>, @Default Dao<? extends Persistent> or @Default Dao<X extends Persistent> where X is a type variable.

public class Dao<T extends Persistent> { ... }

Furthermore, UserDao is eligible for injection to any injection point of type @Default Dao<User>, @Default Dao<?>, @Default Dao<? extends Persistent> or @Default Dao<? extends User>.

public class UserDao extends Dao<User> { ... }

A raw bean type is considered assignable to a parameterized required type if the raw types are identical and all type parameters of the required type are either unbounded type variables or java.lang.Object.

Note that a special set of rules, defined in Assignability of raw and parameterized types for delegate injection points, apply if and only if the injection point is a decorator delegate injection point.

5.2.5. Primitive types and null values

For the purposes of typesafe resolution and dependency injection, primitive types and their corresponding wrapper types in the package java.lang are considered identical and assignable. If necessary, the container performs boxing or unboxing when it injects a value to a field or parameter of primitive or wrapper type.

If an injection point of primitive type resolves to a producer method or producer field that returns a null value at runtime, the container must inject the primitive type’s default value as defined by the Java Language Specification.

5.2.6. Qualifier annotations with members

Qualifier types may have annotation members.

@PayBy(CHEQUE) class ChequePaymentProcessor implements PaymentProcessor { ... }
@PayBy(CREDIT_CARD) class CreditCardPaymentProcessor implements PaymentProcessor { ... }

Then only ChequePaymentProcessor is a candidate for injection to the following attribute:

@Inject @PayBy(CHEQUE) PaymentProcessor paymentProcessor;

On the other hand, only CreditCardPaymentProcessor is a candidate for injection to this attribute:

@Inject @PayBy(CREDIT_CARD) PaymentProcessor paymentProcessor;

The container calls the equals() method of the annotation member value to compare values.

An annotation member may be excluded from consideration using the @Nonbinding annotation.

@Qualifier
@Retention(RUNTIME)
@Target({METHOD, FIELD, PARAMETER, TYPE})
public @interface PayBy {
    PaymentMethod value();
    @Nonbinding String comment() default "";
}

Array-valued or annotation-valued members of a qualifier type should be annotated @Nonbinding in a portable application. If an array-valued or annotation-valued member of a qualifier type is not annotated @Nonbinding, non-portable behavior results.

5.2.7. Multiple qualifiers

A bean class or producer method or field may declare multiple qualifiers.

@Synchronous @PayBy(CHEQUE) class ChequePaymentProcessor implements PaymentProcessor { ... }

Then ChequePaymentProcessor would be considered a candidate for injection into any of the following attributes:

@Inject @PayBy(CHEQUE) PaymentProcessor paymentProcessor;
@Inject @Synchronous PaymentProcessor paymentProcessor;
@Inject @Synchronous @PayBy(CHEQUE) PaymentProcessor paymentProcessor;

A bean must declare all of the qualifiers that are specified at the injection point to be considered a candidate for injection.

5.3. Name resolution

The process of matching a bean to a name is called name resolution. Since there is no typing information available during name resolution, the container may consider only the bean name. Name resolution usually occurs at runtime.

A name resolves to a bean if:

  • the bean has the given bean name, and

  • the bean is available for injection in the module where the name resolution is requested.

For a custom implementation of the Bean interface defined in The Bean interface, the container calls getName() to determine the bean name.

5.3.1. Ambiguous names

An ambiguous name exists when a name resolves to multiple beans. When an ambiguous name exists, the container attempts to resolve the ambiguity. The container eliminates all eligible beans that are not alternatives selected for the bean archive or selected for the application, except for producer methods and fields of beans that are alternatives. If:

  • there is exactly one bean remaining, the container will select this bean, and the ambiguous dependency is called resolvable.

  • all the beans left are alternatives with a priority, then the container will determine the highest priority value, and eliminate all beans, except for producer methods and fields of beans that are alternatives with the highest priority value. If there is exactly one bean remaining, the container will select this bean, and the ambiguous dependency is called resolvable.

All unresolvable ambiguous names are detected by the container when the application is initialized. Suppose two beans are both available for injection in a certain module, and either:

  • the two beans have the same bean name and the name is not resolvable, or

  • the bean name of one bean is of the form x.y, where y is a valid bean name, and x is the bean name of the other bean,

the container automatically detects the problem and treats it as a deployment problem.

5.4. Client proxies

An injected reference, or reference obtained by programmatic lookup, is usually a contextual reference as defined by Contextual reference for a bean.

A contextual reference to a bean with a normal scope, as defined in Normal scopes and pseudo-scopes, is not a direct reference to a contextual instance of the bean (the object returned by Contextual.create()). Instead, the contextual reference is a client proxy object. A client proxy implements/extends some or all of the bean types of the bean and delegates all method calls to the current instance (as defined in Normal scopes and pseudo-scopes) of the bean.

There are a number of reasons for this indirection:

  • The container must guarantee that when any valid injected reference to a bean of normal scope is invoked, the invocation is always processed by the current instance of the injected bean. In certain scenarios, for example if a request scoped bean is injected into a session scoped bean, or into a servlet, this rule requires an indirect reference. (Note that the @Dependent pseudo-scope is not a normal scope.)

  • The container may use a client proxy when creating beans with circular dependencies. This is only necessary when the circular dependencies are initialized via a managed bean constructor or producer method parameter. (Beans with scope @Dependent never have circular dependencies.)

  • Finally, client proxies may be passivated, even when the bean itself may not be. Therefore the container must use a client proxy whenever a bean with normal scope is injected into a bean with a passivating scope, as defined in Passivation and passivating scopes. (On the other hand, beans with scope @Dependent must be serialized along with their client.)

Client proxies are never required for a bean whose scope is a pseudo-scope such as @Dependent.

Client proxies may be shared between multiple injection points. For example, a particular container might instantiate exactly one client proxy object per bean. (However, this strategy is not required by this specification.)

5.4.1. Client proxy invocation

Every time a method of the bean is invoked upon a client proxy, the client proxy must:

If the scope is not active, as specified in The active context object for a scope, the client proxy rethrows the ContextNotActiveException or IllegalStateException.

The behavior of all methods declared by java.lang.Object, except for toString(), is undefined for a client proxy. Portable applications should not invoke any method declared by java.lang.Object, except for toString(), on a client proxy.

5.5. Dependency injection

From time to time the container instantiates beans and other class supporting injection. The resulting instance may or may not be a contextual instance as defined by Contextual instance of a bean.

The container is required to perform dependency injection whenever it creates the following contextual objects:

  • contextual instances of managed beans.

The container is also required to perform dependency injection whenever it instantiates the following non-contextual objects:

  • non-contextual instances of managed beans.

The container interacts with instances of beans or objects supporting injection by calling methods and getting and setting field values.

The object injected by the container may not be a direct reference to a contextual instance of the bean. Instead, it is an injectable reference, as defined by Injectable references.

5.5.1. Injection using the bean constructor

When the container instantiates a managed bean with a constructor annotated @Inject, the container calls this constructor, passing an injectable reference to each parameter. If there is no constructor annotated @Inject, the container calls the constructor with no parameters.

5.5.2. Injection of fields and initializer methods

When the container creates a new instance of a managed bean, the container must:

  • Initialize the values of all injected fields. The container sets the value of each injected field to an injectable reference.

  • Call all initializer methods, passing an injectable reference to each parameter.

The container must ensure that:

  • Initializer methods declared by a class X in the type hierarchy of the bean are called after all injected fields declared by X or by superclasses of X have been initialized.

  • Any @PostConstruct callback declared by a class X in the type hierarchy of the bean is called after all initializer methods declared by X or by superclasses of X have been called, after all injected fields declared by X or by superclasses of X have been initialized.

5.5.3. Destruction of dependent objects

When the container destroys an instance of a bean, the container destroys all dependent objects, as defined in Destruction of objects with scope @Dependent, after the @PreDestroy callback completes.

5.5.4. Invocation of producer or disposer methods

When the container calls a producer or disposer method, the behavior depends upon whether the method is static or non-static:

  • If the method is static, the container must invoke the method.

  • Otherwise, if the method is non-static, the container must:

  • Obtain a contextual instance of the bean which declares the method, as defined by Contextual instance of a bean.

  • Invoke the method upon this instance, as a business method invocation, as defined in Container invocations and interception.

The container passes an injectable reference to each injected method parameter. The container is also responsible for destroying dependent objects created during this invocation, as defined in Destruction of objects with scope @Dependent.

5.5.5. Access to producer field values

When the container accesses the value of a producer field, the value depends upon whether the field is static or non-static:

  • If the producer field is static, the container must access the field value.

  • Otherwise, if the producer field is non-static, the container must:

  • Obtain an contextual instance of the bean which declares the producer field, as defined by Contextual instance of a bean.

  • Access the field value of this instance.

5.5.6. Invocation of observer methods

When the container calls an observer method (defined in Observer methods), the behavior depends upon whether the method is static or non-static:

  • If the observer method is static, the container must invoke the method.

  • Otherwise, if the observer method is non-static, the container must:

  • Obtain a contextual instance of the bean which declares the observer method according to Contextual instance of a bean. If this observer method is a conditional observer method, obtain the contextual instance that already exists, only if the scope of the bean that declares the observer method is currently active, without creating a new contextual instance.

  • Invoke the observer method on the resulting instance, if any, as a business method invocation, as defined in Container invocations and interception.

The container must pass the event object to the event parameter and an injectable instance to each injected method parameter. The container is also responsible for destroying dependent objects created during this invocation, as defined in Destruction of objects with scope @Dependent.

5.5.7. Injection point metadata

The interface javax.enterprise.inject.spi.InjectionPoint provides access to metadata about an injection point. An instance of InjectionPoint may represent:

  • an injected field or a parameter of a bean constructor, initializer method, producer method, disposer method or observer method, or

  • an instance obtained dynamically using Instance.get().

public interface InjectionPoint {
    public Type getType();
    public Set<Annotation> getQualifiers();
    public Bean<?> getBean();
    public Member getMember();
    public Annotated getAnnotated();
    public boolean isDelegate();
    public boolean isTransient();
}
  • The getBean() method returns the Bean object representing the bean that defines the injection point. If the injection point does not belong to a bean, getBean() returns a null value. If the injection point represents a dynamically obtained instance, the getBean() method should return the Bean object representing the bean that defines the Instance injection point.

  • The getType() and getQualifiers() methods return the required type and required qualifiers of the injection point. If the injection point represents a dynamically obtained instance, the getType() and getQualifiers() methods should return the required type (as defined by Instance.select()), and required qualifiers of the injection point including any additional required qualifiers (as defined by Instance.select()).

  • The getMember() method returns the Field object in the case of field injection, the Method object in the case of method parameter injection, or the Constructor object in the case of constructor parameter injection. If the injection point represents a dynamically obtained instance, the getMember() method returns the Field object representing the field that defines the Instance injection point in the case of field injection, the Method object representing the method that defines the Instance injection point in the case of method parameter injection, or the Constructor object representing the constructor that defines the Instance injection point in the case of constructor parameter injection.

  • The getAnnotated() method returns an instance of javax.enterprise.inject.spi.AnnotatedField or javax.enterprise.inject.spi.AnnotatedParameter, depending upon whether the injection point is an injected field or a constructor/method parameter. If the injection point represents a dynamically obtained instance, then the getAnnotated() method returns an instance of javax.enterprise.inject.spi.AnnotatedField or javax.enterprise.inject.spi.AnnotatedParameter representing the Instance injection point, depending upon whether the injection point is an injected field or a constructor/method parameter.

  • The isDelegate() method returns true if the injection point is a decorator delegate injection point, and false otherwise. If the injection point represents a dynamically obtained instance then isDelegate() returns false.

  • The isTransient() method returns true if the injection point is a transient field, and false otherwise. If the injection point represents a dynamically obtained instance then the isTransient() method returns true if the Instance injection point is a transient field, and false otherwise.

Occasionally, a bean with scope @Dependent needs to access metadata relating to the object into which it is injected. For example, the following producer method creates injectable Logger s. The log category of a Logger depends upon the class of the object into which it is injected:

@Produces Logger createLogger(InjectionPoint injectionPoint) {
    return Logger.getLogger( injectionPoint.getMember().getDeclaringClass().getName() );
}

The container must provide a bean with scope @Dependent, bean type InjectionPoint and qualifier @Default, allowing dependent objects, as defined in Dependent objects, to obtain information about the injection point to which they belong. The built-in implementation must be a passivation capable dependency, as defined in Passivation capable dependencies.

If a bean that declares any scope other than @Dependent has an injection point of type InjectionPoint and qualifier @Default, the container automatically detects the problem and treats it as a definition error.

If a disposer method has an injection point of type InjectionPoint and qualifier Default, the container automatically detects the problem and treats it as a definition error.

If a class supporting injection that is not a bean has an injection point of type InjectionPoint and qualifier @Default, the container automatically detects the problem and treats it as a definition error.

5.5.8. Bean metadata

The interfaces Bean, Interceptor and Decorator provide metadata about a bean.

The container must provide beans allowing a bean instance to obtain a Bean, Interceptor or Decorator instance containing its metadata:

  • a bean with scope @Dependent, qualifier @Default and type Bean which can be injected into any bean instance

  • a bean with scope @Dependent, qualifier @Default and type Interceptor which can be injected into any interceptor instance

  • a bean with scope @Dependent, qualifier @Default and type Decorator which can be injected into any decorator instance

Additionally, the container must provide beans allowing interceptors and decorators to obtain information about the beans they intercept and decorate:

  • a bean with scope @Dependent, qualifier @Intercepted and type Bean which can be injected into any interceptor instance, and

  • a bean with scope @Dependent, qualifier @Decorated and type Bean which can be injected into any decorator instance.

These beans are passivation capable dependencies, as defined in Passivation capable dependencies.

If an Interceptor instance is injected into a bean instance other than an interceptor instance, the container automatically detects the problem and treats it as a definition error.

If a Decorator instance is injected into a bean instance other than a decorator instance, the container automatically detects the problem and treats it as a definition error.

If a Bean instance with qualifier @Intercepted is injected into a bean instance other than an interceptor instance, the container automatically detects the problem and treats it as a definition error.

If a Bean instance with qualifier @Decorated is injected into a bean instance other than a decorator instance, the container automatically detects the problem and treats it as a definition error.

The injection of bean metadata is restricted. If:

  • the injection point is a field, an initializer method parameter or a bean constructor, with qualifier @Default, then the type parameter of the injected Bean, Interceptor or Decorator must be the same as the type declaring the injection point, or

  • the injection point is a field, an initializer method parameter or a bean constructor of an interceptor, with qualifier @Intercepted, then the type parameter of the injected Bean must be an unbounded wildcard, or

  • the injection point is a field, an initializer method parameter or a bean constructor of a decorator, with qualifier @Decorated, then the type parameter of the injected Bean must be the same as the delegate type, or

  • the injection point is a producer method parameter then the type parameter of the injected Bean must be the same as the producer method return type, or

  • the injection point is a parameter of a disposer method then the container automatically detects the problem and treats it as a definition error.

Otherwise, the container automatically detects the problem and treats it as a definition error.

@Named("Order") public class OrderProcessor {

    @Inject Bean<OrderProcessor> bean;

    public void getBeanName() {
       return bean.getName();
    }

}

5.6. Programmatic lookup

In certain situations, injection is not the most convenient way to obtain a contextual reference. For example, it may not be used when:

  • the bean type or qualifiers vary dynamically at runtime, or

  • depending upon the deployment, there may be no bean which satisfies the type and qualifiers, or

  • we would like to iterate over all beans of a certain type.

In these situations, an instance of the javax.enterprise.inject.Instance interface may be injected:

@Inject Instance<PaymentProcessor> paymentProcessor;

The method get() returns a contextual reference:

PaymentProcessor pp = paymentProcessor.get();

Any combination of qualifiers may be specified at the injection point:

@Inject @PayBy(CHEQUE) Instance<PaymentProcessor> chequePaymentProcessor;

Or, the @Any qualifier may be used, allowing the application to specify qualifiers dynamically:

@Inject @Any Instance<PaymentProcessor> anyPaymentProcessor;
...
Annotation qualifier = synchronously ? new SynchronousQualifier() : new AsynchronousQualifier();
PaymentProcessor pp = anyPaymentProcessor.select(qualifier).get().process(payment);

In this example, the returned bean has qualifier @Synchronous or @Asynchronous depending upon the value of synchronously.

Finally, the @New qualifier may be used, allowing the application to obtain a @New qualified bean, as defined in @New qualified beans:

@Inject @New(ChequePaymentProcessor.class) Instance<PaymentProcessor> chequePaymentProcessor;

It’s even possible to iterate over a set of beans:

@Inject @Any Instance<PaymentProcessor> anyPaymentProcessor;
...
for (PaymentProcessor pp: anyPaymentProcessor) pp.test();

5.6.1. The Instance interface

The Instance interface provides a method for obtaining instances of beans with a specified combination of required type and qualifiers, and inherits the ability to iterate beans with that combination of required type and qualifiers from java.lang.Iterable:

public interface Instance<T> extends Iterable<T>, Provider<T> {

    public Instance<T> select(Annotation... qualifiers);
    public <U extends T> Instance<U> select(Class<U> subtype, Annotation... qualifiers);
    public <U extends T> Instance<U> select(TypeLiteral<U> subtype, Annotation... qualifiers);

    public boolean isUnsatisfied();
    public boolean isAmbiguous();

    public void destroy(T instance);

}

For an injected Instance:

  • the required type is the type parameter specified at the injection point, and

  • the required qualifiers are the qualifiers specified at the injection point.

For example, this injected Instance has required type PaymentProcessor and required qualifier @Any:

@Inject @Any Instance<PaymentProcessor> anyPaymentProcessor;

The select() method returns a child Instance for a given required type and additional required qualifiers. If no required type is given, the required type is the same as the parent.

For example, this child Instance has required type AsynchronousPaymentProcessor and additional required qualifier @Asynchronous:

Instance<AsynchronousPaymentProcessor> async = anyPaymentProcessor.select(
            AsynchronousPaymentProcessor.class, new AsynchronousQualifier() );

If an injection point of raw type Instance is defined, the container automatically detects the problem and treats it as a definition error.

If two instances of the same qualifier type are passed to select(), an IllegalArgumentException is thrown.

If an instance of an annotation that is not a qualifier type is passed to select(), an IllegalArgumentException is thrown.

The get() method must:

  • Identify a bean that has the required type and required qualifiers and is eligible for injection into the class into which the parent Instance was injected, according to the rules of typesafe resolution, as defined in Performing typesafe resolution, resolving ambiguities according to Unsatisfied and ambiguous dependencies.

  • If typesafe resolution results in an unsatisfied dependency, throw an UnsatisfiedResolutionException. If typesafe resolution results in an unresolvable ambiguous dependency, throw an AmbiguousResolutionException.

  • Otherwise, obtain a contextual reference for the bean and the required type, as defined in Contextual reference for a bean.

The iterator() method must:

  • Identify the set of beans that have the required type and required qualifiers and are eligible for injection into the class into which the parent Instance was injected, according to the rules of typesafe resolution, as defined in Performing typesafe resolution, resolving ambiguities according to Unsatisfied and ambiguous dependencies.

  • Return an Iterator, that iterates over the set of contextual references for the resulting beans and required type, as defined in Contextual reference for a bean.

The method isUnsatisfied() returns true if there is no bean that has the required type and qualifiers and is eligible for injection into the class into which the parent Instance was injected, or false otherwise.

The method isAmbiguous() returns true if there is more than one bean that has the required type and qualifiers and is eligible for injection into the class into which the parent Instance was injected, or false otherwise.

The method destroy() instructs the container to destroy the instance. The bean instance passed to destroy() should be a dependent scoped bean instance, or a client proxy for a normal scoped bean. Applications are encouraged to always call destroy() when they no longer require an instance obtained from Instance. All built-in normal scoped contexts support destroying bean instances. An UnsupportedOperationException is thrown if the active context object for the scope type of the bean does not support destroying bean instances.

5.6.2. The built-in Instance

The container must provide a built-in bean with:

  • Instance<X> and Provider<X> for every legal bean type X in its set of bean types,

  • every qualifier type in its set of qualifier types,

  • scope @Dependent,

  • no bean name, and

  • an implementation provided automatically by the container.

The built-in implementation must be a passivation capable dependency, as defined in Passivation capable dependencies.

5.6.3. Using AnnotationLiteral and TypeLiteral

javax.enterprise.util.AnnotationLiteral makes it easier to specify qualifiers when calling select():

public PaymentProcessor getSynchronousPaymentProcessor(PaymentMethod paymentMethod) {

    class SynchronousQualifier extends AnnotationLiteral<Synchronous>
            implements Synchronous {}

    class PayByQualifier extends AnnotationLiteral<PayBy>
            implements PayBy {
        public PaymentMethod value() { return paymentMethod; }
    }

    return anyPaymentProcessor.select(new SynchronousQualifier(), new PayByQualifier()).get();
}

javax.enterprise.util.TypeLiteral makes it easier to specify a parameterized type with actual type parameters when calling select():

public PaymentProcessor<Cheque> getChequePaymentProcessor() {
    PaymentProcessor<Cheque> pp = anyPaymentProcessor
        .select( new TypeLiteral<PaymentProcessor<Cheque>>() {} ).get();
}

6. Scopes and contexts

Associated with every scope type is a context object. The context object determines the lifecycle and visibility of instances of all beans with that scope. In particular, the context object defines:

  • When a new instance of any bean with that scope is created

  • When an existing instance of any bean with that scope is destroyed

  • Which injected references refer to any instance of a bean with that scope

The context implementation collaborates with the container via the Context and Contextual interfaces to create and destroy contextual instances.

6.1. The Contextual interface

The interface javax.enterprise.context.spi.Contextual defines operations to create and destroy contextual instances of a certain type. Any implementation of Contextual is called a contextual type. In particular, the Bean interface defined in The Bean interface extends Contextual, so all beans are contextual types.

public interface Contextual<T> {
    public T create(CreationalContext<T> creationalContext);
    public void destroy(T instance, CreationalContext<T> creationalContext);
}
  • create() is responsible for creating new contextual instances of the type.

  • destroy() is responsible for destroying instances of the type. In particular, it is responsible for destroying all dependent objects of an instance.

If an exception occurs while creating an instance, the exception is rethrown by the create() method. If the exception is a checked exception, it must be wrapped and rethrown as an (unchecked) CreationException.

If an exception occurs while destroying an instance, the exception must be caught by the destroy() method.

If the application invokes a contextual instance after it has been destroyed, the behavior is undefined.

The container and portable extensions may define implementations of the Contextual interface that do not extend Bean, but it is not recommended that applications directly implement Contextual.

6.1.1. The CreationalContext interface

The interface javax.enterprise.context.spi.CreationalContext provides operations that are used by the Contextual implementation during instance creation and destruction.

public interface CreationalContext<T> {
    public void push(T incompleteInstance);
    public void release();
}
  • push() registers an incompletely initialized contextual instance the with the container. A contextual instance is considered incompletely initialized until it is returned by the create() method.

  • release() destroys all dependent objects, as defined in Dependent objects, of the instance which is being destroyed, by passing each dependent object to the destroy() method of its Contextual object.

The implementation of Contextual is not required to call push(). However, for certain bean scopes, invocation of push() between instantiation and injection helps the container minimize the use of client proxy objects (which would otherwise be required to allow circular dependencies).

If Contextual.create() calls push(), it must also return the instance passed to push().

Contextual.create() should use the given CreationalContext when obtaining contextual references to inject, as defined in Contextual reference for a bean, in order to ensure that any dependent objects are associated with the contextual instance that is being created.

Contextual.destroy() should call release() to allow the container to destroy dependent objects of the contextual instance.

6.2. The Context interface

The javax.enterprise.context.spi.Context interface provides an operation for obtaining contextual instances with a particular scope of any contextual type. Any instance of Context is called a context object.

The context object is responsible for creating and destroying contextual instances by calling operations of the Contextual interface.

The Context interface is called by the container and may be called by portable extensions. It should not be called directly by the application.

public interface Context {
    public Class<? extends Annotation> getScope();
    boolean isActive();
    public <T> T get(Contextual<T> bean);
    public <T> T get(Contextual<T> bean, CreationalContext<T> creationalContext);
}
public interface AlterableContext extends Context {
    public void destroy(Contextual<?> contextual);
}

The method getScope() returns the scope type of the context object.

A context object may be defined for any of the built-in scopes and registered with the container using the AfterBeanDiscovery event as described in AfterBeanDiscovery event.

At a particular point in the execution of the program a context object may be active with respect to the current thread. When a context object is active the isActive() method returns true. Otherwise, we say that the context object is inactive and the isActive() method returns false.

The get() method obtains contextual instances of the contextual type represented by the given instance of Contextual. The get() method may either:

  • return an existing instance of the given contextual type, or

  • if no CreationalContext is given, return a null value, or

  • if a CreationalContext is given, create a new instance of the given contextual type by calling Contextual.create(), passing the given CreationalContext, and return the new instance.

The get() method may not return a null value unless no CreationalContext is given, or Contextual.create() returns a null value.

The get() method may not create a new instance of the given contextual type unless a CreationalContext is given.

The destroy() method destroys an existing contextual instance, removing it from the context instance.

The AlterableContext interface was introduced in CDI 1.1 to allow bean instances to be destroyed by the application. Extensions providing context implementations for normal scopes should implement AlterableContext instead of Context.

If the context object is inactive, the get() and destroy() methods must throw a ContextNotActiveException.

When the container calls get() or destroy() for a context that is associated with a passivating scope it must ensure that the given instance of Contextual and the instance of CreationalContext, if given, are serializable.

The context object is responsible for destroying any contextual instance it creates by passing the instance to the destroy() method of the Contextual object representing the contextual type. A destroyed instance must not subsequently be returned by the get() method.

The context object must pass the same instance of CreationalContext to Contextual.destroy() that it passed to Contextual.create() when it created the instance.

6.3. Normal scopes and pseudo-scopes

Most scopes are normal scopes. The context object for a normal scope type is a mapping from each contextual type with that scope to an instance of that contextual type. There may be no more than one mapped instance per contextual type per thread. The set of all mapped instances of contextual types with a certain scope for a certain thread is called the context for that scope associated with that thread.

A context may be associated with one or more threads. A context with a certain scope is said to propagate from one point in the execution of the program to another when the set of mapped instances of contextual types with that scope is preserved.

The context associated with the current thread is called the current context for the scope. The mapped instance of a contextual type associated with a current context is called the current instance of the contextual type.

The get() operation of the context object for an active normal scope returns the current instance of the given contextual type.

At certain points in the execution of the program a context may be destroyed. When a context is destroyed, all mapped instances belonging to that context are destroyed by passing them to the Contextual.destroy() method.

Contexts with normal scopes must obey the following rule:

Suppose beans A, B and Z all have normal scopes. Suppose A has an injection point x, and B has an injection point y. Suppose further that both x and y resolve to bean Z according to the rules of typesafe resolution. If a is the current instance of A, and b is the current instance of B, then both a.x and b.y refer to the same instance of Z. This instance is the current instance of Z.

Any scope that is not a normal scope is called a pseudo-scope. The concept of a current instance is not well-defined in the case of a pseudo-scope.

All normal scopes must be explicitly declared @NormalScope, to indicate to the container that a client proxy is required.

All pseudo-scopes must be explicitly declared @Scope, to indicate to the container that no client proxy is required.

All scopes defined by this specification, except for the @Dependent pseudo-scope, are normal scopes.

6.4. Dependent pseudo-scope

The @Dependent scope type is a pseudo-scope. Beans declared with scope type @Dependent behave differently to beans with other built-in scope types.

When a bean is declared to have @Dependent scope:

  • No injected instance of the bean is ever shared between multiple injection points.

  • Any instance of the bean injected into an object that is being created by the container is bound to the lifecycle of the newly created object.

  • Any instance of the bean that receives a producer method, producer field, disposer method or observer method invocation exists to service that invocation only.

  • Any instance of the bean injected into method parameters of a disposer method or observer method exists to service the method invocation only (except for observer methods of container lifecycle events).

Every invocation of the get() operation of the Context object for the @Dependent scope with a CreationalContext returns a new instance of the given bean.

Every invocation of the get() operation of the Context object for the @Dependent scope with no CreationalContext returns a null value.

The @Dependent scope is always active.

6.4.1. Dependent objects

Many instances of beans with scope @Dependent belong to some other bean and are called dependent objects.

  • Instances of decorators and interceptors are dependent objects of the bean instance they decorate.

  • An instance of a bean with scope @Dependent injected into a field, bean constructor or initializer method is a dependent object of the bean into which it was injected.

  • An instance of a bean with scope @Dependent injected into a producer method is a dependent object of the producer method bean instance that is being produced.

  • An instance of a bean with scope @Dependent obtained by direct invocation of an Instance is a dependent object of the instance of Instance.

6.4.2. Destruction of objects with scope @Dependent

Dependent objects of a contextual instance are destroyed when Contextual.destroy() calls CreationalContext.release(), as defined in The CreationalContext interface.

Additionally, the container must ensure that:

  • all dependent objects of a non-contextual instance of a bean are destroyed when the instance is destroyed by the container,

  • all @Dependent scoped contextual instances injected into method parameters of a disposer method or an observer method are destroyed when the invocation completes,

  • all @Dependent scoped contextual instances injected into method or constructor parameters that are annotated with @TransientReference are destroyed when the invocation completes, and

  • any @Dependent scoped contextual instance created to receive a producer method, producer field, disposer method or observer method invocation is destroyed when the invocation completes.

Finally, the container is permitted to destroy any @Dependent scoped contextual instance at any time if the instance is no longer referenced by the application (excluding weak, soft and phantom references).

6.5. Contextual instances and contextual references

The Context object is the ultimate source of the contextual instances that underly contextual references.

6.5.1. The active context object for a scope

From time to time, the container must obtain an active context object for a certain scope type. The container must search for an active instance of Context associated with the scope type.

  • If no active context object exists for the scope type, the container throws a ContextNotActiveException.

  • If more than one active context object exists for the given scope type, the container must throw an IllegalStateException.

If there is exactly one active instance of Context associated with the scope type, we say that the scope is active.

6.5.2. Contextual instance of a bean

From time to time, the container must obtain a contextual instance of a bean. The container must:

  • obtain the active context object for the bean scope, then

  • obtain an instance of the bean by calling Context.get(), passing the Bean instance representing the bean and an instance of CreationalContext.

From time to time, the container attempts to obtain a contextual instance of a bean that already exists, without creating a new contextual instance. The container must determine if the scope of the bean is active and if it is:

  • obtain the active context object for the bean scope, then

  • attempt to obtain an existing instance of the bean by calling Context.get(), passing the Bean instance representing the bean without passing any instance of CreationalContext.

If the scope is not active, or if Context.get() returns a null value, there is no contextual instance that already exists.

A contextual instance of any of the built-in kinds of bean defined in Programming model is considered an internal container construct, and it is therefore not strictly required that a contextual instance of a built-in kind of bean directly implement the bean types of the bean. However, in this case, the container is required to transform its internal representation to an object that does implement the bean types expected by the application before injecting or returning a contextual instance to the application.

For a custom implementation of the Bean interface defined in The Bean interface, the container calls getScope() to determine the bean scope.

6.5.3. Contextual reference for a bean

From time to time, the container must obtain a contextual reference for a bean and a given bean type of the bean. A contextual reference implements the given bean type and all bean types of the bean which are Java interfaces. A contextual reference is not, in general, required to implement all concrete bean types of the bean.

Contextual references must be obtained with a given CreationalContext, allowing any instance of scope @Dependent that is created to be later destroyed.

  • If the bean has a normal scope and the given bean type cannot be proxied by the container, as defined in Unproxyable bean types, the container throws an UnproxyableResolutionException.

  • If the bean has a normal scope, then the contextual reference for the bean is a client proxy, as defined in Client proxies, created by the container, that implements the given bean type and all bean types of the bean which are Java interfaces.

  • Otherwise, if the bean has a pseudo-scope, the container must obtain a contextual instance of the bean. If the bean has scope @Dependent, the container must associate it with the CreationalContext.

The container must ensure that every injection point of type InjectionPoint and qualifier @Default of any dependent object instantiated during this process receives:

  • an instance of InjectionPoint representing the injection point into which the dependent object will be injected, or

  • a null value if it is not being injected into any injection point.

6.5.4. Contextual reference validity

A contextual reference for a bean is valid only for a certain period of time. The application should not invoke a method of an invalid reference.

The validity of a contextual reference for a bean depends upon whether the scope of the bean is a normal scope or a pseudo-scope.

  • Any reference to a bean with a normal scope is valid as long as the application maintains a hard reference to it. However, it may only be invoked when the context associated with the normal scope is active. If it is invoked when the context is inactive, a ContextNotActiveException is thrown by the container.

  • Any reference to a bean with a pseudo-scope (such as @Dependent) is valid until the bean instance to which it refers is destroyed. It may be invoked even if the context associated with the pseudo-scope is not active. If the application invokes a method of a reference to an instance that has already been destroyed, the behavior is undefined.

6.5.5. Injectable references

From time to time, the container must obtain an injectable reference for an injection point. The container must:

For certain combinations of scopes, the container is permitted to optimize the above procedure:

  • The container is permitted to directly inject a contextual instance of the bean, as defined in Contextual instance of a bean.

  • If an incompletely initialized instance of the bean is registered with the current CreationalContext, as defined in The Contextual interface, the container is permitted to directly inject this instance.

However, in performing these optimizations, the container must respect the rules of injectable reference validity.

6.5.6. Injectable reference validity

Injectable references to a bean must respect the rules of contextual reference validity, with the following exceptions:

  • A reference to a bean injected into a field, bean constructor or initializer method is only valid until the object into which it was injected is destroyed.

  • A reference to a bean injected into a producer method is only valid until the producer method bean instance that is being produced is destroyed.

  • A reference to a bean injected into a disposer method or observer method is only valid until the invocation of the method completes.

The application should not invoke a method of an invalid injected reference. If the application invokes a method of an invalid injected reference, the behavior is undefined.

6.6. Passivation and passivating scopes

The temporary transfer of the state of an idle object held in memory to some form of secondary storage is called passivation. The transfer of the passivated state back into memory is called activation.

6.6.1. Passivation capable beans

A bean is called passivation capable if the container is able to temporarily transfer the state of any idle instance to secondary storage.

  • A managed bean is passivation capable if and only if the bean class is serializable and all interceptors and decorators of the bean are passivation capable.

  • A producer method is passivation capable if and only if it never returns a value which is not passivation capable at runtime.

  • A producer field is passivation capable if and only if it never refers to a value which is not passivation capable at runtime.

A custom implementation of Bean is passivation capable if it implements the interface PassivationCapable. An implementation of Contextual that is not a bean is passivation capable if it implements both PassivationCapable and Serializable.

public interface PassivationCapable {
    public String getId();
}

The getId() method must return a value that uniquely identifies the instance of Bean or Contextual. It is recommended that the string contain the package name of the class that implements Bean or Contextual.

6.6.2. Passivation capable injection points

We call an injection point of a bean passivation capable if the injection point is:

  • a transient field, or

  • a non-transient field which resolves to a bean that is a passivation capable dependency, or

  • a bean constructor parameter which is annotated with @TransientReference, or

  • a bean constructor parameter which resolves to a bean that is a passivation capable dependency, or

  • a method parameter which is annotated with @TransientReference, or

  • a method parameter which resolves to a bean that is a passivation capable dependency.

6.6.3. Passivation capable dependencies

A bean is called a passivation capable dependency if any contextual reference for that bean is preserved when the object holding the reference is passivated and then activated.

The container must guarantee that:

  • all beans with normal scope are passivation capable dependencies,

  • all passivation capable beans with scope @Dependent are passivation capable dependencies,

  • the built-in beans of type Instance, Event, InjectionPoint and BeanManager are passivation capable dependencies.

A custom implementation of Bean is a passivation capable dependency if it implements PassivationCapable.

6.6.4. Passivating scopes

A passivating scope requires that:

  • beans with the scope are passivation capable, and

  • implementations of Contextual passed to any context object for the scope are passivation capable.

Passivating scopes must be explicitly declared @NormalScope(passivating=true).

For example, the built-in session and conversation scopes defined in Context management for built-in scopes are passivating scopes. No other built-in scopes are passivating scopes.

6.6.5. Validation of passivation capable beans and dependencies

For every bean which declares a passivating scope, the container must validate that the bean truly is passivation capable and that, in addition, its dependencies are passivation capable.

If a managed bean which declares a passivating scope, or a built-in bean:

  • is not passivation capable,

  • has an injection point that is not passivation capable,

  • has an interceptor or decorator that is not passivation capable

  • has an interceptor or decorator with an injection point that is not passivation capable

then the container automatically detects the problem and treats it as a deployment problem.

If a producer method declares a passivating scope and:

  • has a return type that is declared final and does not implement or extend Serializable, or,

  • has an injection point that is not passivation capable

then the container automatically detects the problem and treats it as a deployment problem.

If a producer method declares a passivating scope and doesn’t only return Serializable types at runtime, then the container must throw an IllegalProductException.

If a producer field declares a passivating scope and has a type that is declared final and does not implement or extend Serializable then the container automatically detects the problem and treats it as a deployment problem.

If a producer field declares a passivating scope and doesn’t only contain Serializable values at runtime then the container must throw an IllegalProductException.

If a producer method or field of scope @Dependent returns an unserializable object for injection into an injection point that requires a passivation capable dependency, the container must throw an IllegalProductException

For a custom implementation of Bean, the container calls getInjectionPoints() to determine the injection points, and InjectionPoint.isTransient() to determine whether the injection point is a transient field.

If a managed bean which declares a passivating scope type, has a decorator or interceptor which is not a passivation capable dependency, the container automatically detects the problem and treats it as a deployment problem.

6.7. Context management for built-in scopes

The container must provide an implementation of the Context interface for each of the built-in scopes defined in Built-in scope types. These implementations depend on the platform the container is running.

A portable extension may define a custom context object for any or all of the built-in scopes. For example, a remoting framework might provide a request context object for the built-in request scope.

The context associated with a built-in normal scope propagates across local, synchronous Java method calls. The context does not propagate across remote method invocations or to asynchronous processes.

Portable extensions are encouraged to fire an event with qualifier @Initialized(X.class) when a custom context is initialized, and an event with qualifier @Destroyed(X.class) when a custom context is destroyed, where X is the scope type associated with the context. A suitable event payload should be chosen.

6.7.1. Request context lifecycle

The request context is provided by a built-in context object for the built-in scope type @RequestScoped.

An event with qualifier @Initialized(RequestScoped.class) is fired when the request context is initialized and an event with qualifier @Destroyed(RequestScoped.class) when the request context is destroyed.

6.7.2. Session context lifecycle

The session context is provided by a built-in context object for the built-in passivating scope type @SessionScoped.

6.7.3. Application context lifecycle

The application context is provided by a built-in context object for the built-in scope type @ApplicationScoped.

An event with qualifier @Initialized(ApplicationScoped.class) is fired when the application context is initialized and an event with qualifier @Destroyed(ApplicationScoped.class) is fired when the application is destroyed.

6.7.4. Conversation context lifecycle

The conversation context is provided by a built-in context object for the built-in passivating scope type @ConversationScoped.

6.7.5. The Conversation interface

The container provides a built-in bean with bean type Conversation, scope @RequestScoped, and qualifier @Default, named javax.enterprise.context.conversation.

public interface Conversation {
   public void begin();
   public void begin(String id);
   public void end();
   public String getId();
   public long getTimeout();
   public void setTimeout(long milliseconds);
   public boolean isTransient();
}
  • begin() marks the current transient conversation long-running. A conversation identifier may, optionally, be specified. If no conversation identifier is specified, an identifier is generated by the container.

  • end() marks the current long-running conversation transient.

  • getId() returns the identifier of the current long-running conversation, or a null value if the current conversation is transient.

  • getTimeout() returns the timeout, in milliseconds, of the current conversation.

  • setTimeout() sets the timeout of the current conversation.

  • isTransient() returns true if the conversation is marked transient, or false if it is marked long-running.

If any method of Conversation is called when the conversation scope is not active, a ContextNotActiveException is thrown.

If end() is called, and the current conversation is marked transient, an IllegalStateException is thrown.

If begin() is called, and the current conversation is already marked long-running, an IllegalStateException is thrown.

If begin() is called with an explicit conversation identifier, and a long-running conversation with that identifier already exists, an IllegalArgumentException is thrown.

7. Lifecycle of contextual instances

The lifecycle of a contextual instance of a bean is managed by the context object for the bean’s scope, as defined in Scopes and contexts.

Every bean in the system is represented by an instance of the Bean interface defined in The Bean interface. This interface is a subtype of Contextual. To create and destroy contextual instances, the context object calls the create() and destroy() operations defined by the interface Contextual, as defined in The Contextual interface.

7.1. Restriction upon bean instantiation

There are very few programming restrictions upon the bean class of a bean. In particular, the class is a concrete class and is not required to implement any special interface or extend any special superclass. Therefore, bean classes are easy to instantiate and unit test.

However, if the application directly instantiates a bean class, instead of letting the container perform instantiation, the resulting instance is not managed by the container and is not a contextual instance as defined by Contextual instance of a bean. Furthermore, the capabilities listed in Functionality provided by the container to the bean will not be available to that particular instance. In a deployed application, it is the container that is responsible for instantiating beans and initializing their dependencies.

If the application requires more control over instantiation of a contextual instance, a producer method or field may be used. Any Java object may be returned by a producer method or field. It is not required that the returned object be a contextual reference for a bean. However, if the object is not a contextual reference for another bean, the object will be contextual instance of the producer method bean, and therefore available for injection into other objects and use in name resolution, but the other capabilities listed in Functionality provided by the container to the bean will not be available to the object.

In the following example, a producer method returns instances of other beans:

@SessionScoped
public class PaymentStrategyProducer implements Serializable {

   private PaymentStrategyType paymentStrategyType;

   public void setPaymentStrategyType(PaymentStrategyType type) {
      paymentStrategyType = type;
   }

   @Produces PaymentStrategy getPaymentStrategy(@CreditCard PaymentStrategy creditCard,
                                                @Cheque PaymentStrategy cheque,
                                                @Online PaymentStrategy online) {
      switch (paymentStrategyType) {
         case CREDIT_CARD: return creditCard;
         case CHEQUE: return cheque;
         case ONLINE: return online;
         default: throw new IllegalStateException();
      }
    }

}

In this case, any object returned by the producer method has already had its dependencies injected, receives lifecycle callbacks and event notifications and may have interceptors.

But in this example, the returned objects are not contextual instances:

@SessionScoped
public class PaymentStrategyProducer implements Serializable {

   private PaymentStrategyType paymentStrategyType;

   public void setPaymentStrategyType(PaymentStrategyType type) {
      paymentStrategyType = type;
   }

   @Produces PaymentStrategy getPaymentStrategy() {
      switch (paymentStrategyType) {
         case CREDIT_CARD: return new CreditCardPaymentStrategy();
         case CHEQUE: return new ChequePaymentStrategy();
         case ONLINE: return new OnlinePaymentStrategy();
         default: throw new IllegalStateException();
      }
    }

}

In this case, any object returned by the producer method will not have any dependencies injected by the container, receives no lifecycle callbacks or event notifications and does not have interceptors or decorators.

7.2. Container invocations and interception

When the application invokes:

  • a method of a bean via a contextual reference to the bean, as defined in Contextual reference for a bean, or

  • a method of a bean via a non-contextual reference to the bean, if the instance was created by the container (e.g. using InjectionTarget.produce() or UnmanagedInstance.produce()),

the invocation is treated as a business method invocation.

When the container invokes a method of a bean, the invocation may or may not be treated as a business method invocation:

  • Invocations of initializer methods by the container are not business method invocations.

  • Invocations of producer, disposer and observer methods by the container are business method invocations and are intercepted by method interceptors and decorators.

  • Invocation of lifecycle callbacks by the container are not business method invocations, but are intercepted by interceptors for lifecycle callbacks.

  • Invocations of interceptors and decorator methods during method or lifecycle callback interception are not business method invocations, and therefore no recursive interception occurs.

  • Invocations of methods declared by java.lang.Object are not business method invocations.

A method invocation passes through method interceptors and decorators if, and only if, it is a business method invocation.

Otherwise, the invocation is treated as a normal Java method call and is not intercepted by the container.

7.3. Lifecycle of contextual instances

The actual mechanics of bean creation and destruction varies according to what kind of bean is being created or destroyed.

7.3.1. Lifecycle of managed beans

When the create() method of the Bean object that represents a managed bean is called, the container obtains an instance of the bean, as defined by the Managed Beans specification, calling the bean constructor as defined by Injection using the bean constructor, and performing dependency injection as defined in Injection of fields and initializer methods.

When the destroy() method is called, the container destroys the instance, as defined by the Managed Beans specification, and any dependent objects, as defined in Destruction of dependent objects.

7.3.2. Lifecycle of producer methods

When the create() method of a Bean object that represents a producer method is called, the container must invoke the producer method as defined by Invocation of producer or disposer methods. The return value of the producer method, after method interception completes, is the new contextual instance to be returned by Bean.create().

If the producer method returns a null value and the producer method bean has the scope @Dependent, the create() method returns a null value.

Otherwise, if the producer method returns a null value, and the scope of the producer method is not @Dependent, the create() method throws an IllegalProductException.

When the destroy() method is called, and if there is a disposer method for this producer method, the container must invoke the disposer method as defined by Invocation of producer or disposer methods, passing the instance given to destroy() to the disposed parameter. Finally, the container destroys dependent objects, as defined in Destruction of dependent objects.

7.3.3. Lifecycle of producer fields

When the create() method of a Bean object that represents a producer field is called, the container must access the producer field as defined by Access to producer field values to obtain the current value of the field. The value of the producer field is the new contextual instance to be returned by Bean.create().

If the producer field contains a null value and the producer field bean has the scope @Dependent, the create() method returns a null value.

Otherwise, if the producer field contains a null value, and the scope of the producer field is not @Dependent, the create() method throws an IllegalProductException.

When the destroy() method is called, and if there is a disposer method for this producer field, the container must invoke the disposer method as defined by Invocation of producer or disposer methods, passing the instance given to destroy() to the disposed parameter.

8. Decorators

A decorator implements one or more bean types and intercepts business method invocations of beans which implement those bean types. These bean types are called decorated types.

Decorators are superficially similar to interceptors, but because they directly implement operations with business semantics, they are able to implement business logic and, conversely, unable to implement the cross-cutting concerns for which interceptors are optimized.

Decorators may be associated with any managed bean that is not itself an interceptor or decorator, or with any built-in bean other than the built-in bean with type BeanManager and qualifier @Default. Decorators are not applied to the return value of a producer method or the current value of a producer field. A decorator instance is a dependent object of the object it decorates.

8.1. Decorator beans

A decorator is a managed bean. The set of decorated types of a decorator includes all bean types of the managed bean which are Java interfaces, except for java.io.Serializable. The decorator bean class and its superclasses are not decorated types of the decorator. The decorator class may be abstract.

If the set of decorated types of a decorator is empty, the container automatically detects the problem and treats it as a definition error.

8.1.1. Declaring a decorator

A decorator is declared by annotating the bean class with the @javax.decorator.Decorator stereotype.

@Decorator @Priority(APPLICATION)
class TimestampLogger implements Logger { ... }

8.1.2. Decorator delegate injection points

All decorators have a delegate injection point. A delegate injection point is an injection point of the bean class. The type and qualifiers of the injection point are called the delegate type and delegate qualifiers. The decorator applies to beans that are assignable to the delegate injection point.

The delegate injection point must be declared by annotating the injection point with the annotation @javax.decorator.Delegate:

@Decorator @Priority(APPLICATION)
class TimestampLogger implements Logger {
    @Inject @Delegate @Any Logger logger;
    ...
}
@Decorator @Priority(APPLICATION)
class TimestampLogger implements Logger {
    private Logger logger;

    @Inject
    public TimestampLogger(@Delegate @Debug Logger logger) {
        this.logger=logger;
    }
    ...
}

A decorator must have exactly one delegate injection point. If a decorator has more than one delegate injection point, or does not have a delegate injection point, the container automatically detects the problem and treats it as a definition error.

The delegate injection point must be an injected field, initializer method parameter or bean constructor method parameter. If an injection point that is not an injected field, initializer method parameter or bean constructor method parameter is annotated @Delegate, the container automatically detects the problem and treats it as a definition error.

If a bean class that is not a decorator has an injection point annotated @Delegate, the container automatically detects the problem and treats it as a definition error.

The container must inject a delegate object to the delegate injection point. The delegate object implements the delegate type and delegates method invocations to remaining uninvoked decorators and eventually to the bean. When the container calls a decorator during business method interception, the decorator may invoke any method of the delegate object.

@Decorator @Priority(APPLICATION)
class TimestampLogger implements Logger {
    @Inject @Delegate @Any Logger logger;

    void log(String message) {
        logger.log( timestamp() + ": " + message );
    }
    ...
}

If a decorator invokes the delegate object at any other time, the invoked method throws an IllegalStateException.

8.1.3. Decorated types of a decorator

The delegate type of a decorator must implement or extend every decorated type (with exactly the same type parameters). If the delegate type does not implement or extend a decorated type of the decorator (or specifies different type parameters), the container automatically detects the problem and treats it as a definition error.

A decorator is not required to implement the delegate type.

A decorator may be an abstract Java class, and is not required to implement every method of every decorated type. Whenever the decorator does not implement a method of the decorated type, the container will provide an implicit implementation that calls the method on the delegate. If a decorator has abstract methods that are not declared by a decorated type, the container automatically detects the problem and treats it as a definition error.

The decorator intercepts every method which is declared by a decorated type of the decorator and is implemented by the bean class of the decorator.

8.2. Decorator enablement and ordering

This specification defines two methods of enabling and ordering decorators. From Contexts and Dependency Injection 1.1 onwards the @Priority annotation allows a decorator to be enabled and ordered for an entire application. Contexts and Dependency Injection 1.0 allowed only for a decorator to be enabled and ordered for a bean archive.

Decorators are called after interceptors. Decorators enabled using @Priority are called before decorators enabled using beans.xml.

A decorator is said to be enabled if it is enabled in at least one bean archive or is listed in the final list of decorators for the application, as defined in AfterTypeDiscovery event.

8.2.1. Decorator enablement and ordering for an application

A decorator may be enabled for the entire application by applying the @Priority annotation, along with a priority value, on the decorator class. Decorators with the smaller priority values are called first. The order of more than one decorator with the same priority is undefined.

@Decorator @Priority(APPLICATION)
class TimestampLogger implements Logger {

   ...
}

The priority value ranges defined in the Java Interceptors specification section 5.5 should be used when defining decorator priorities.

8.2.2. Decorator enablement and ordering for a bean archive

A decorator may be explicitly enabled by listing its bean class under the <decorators> element of the beans.xml file of the bean archive.

<beans xmlns="http://xmlns.jcp.org/xml/ns/javaee"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://xmlns.jcp.org/xml/ns/javaee http://xmlns.jcp.org/xml/ns/javaee/beans_1_1.xsd">
   <decorators>
      <class>com.acme.myfwk.TimestampLogger</class>
      <class>com.acme.myfwk.IdentityLogger</class>
   </decorators>
</beans>

The order of the decorator declarations determines the decorator ordering. Decorators which occur earlier in the list are called first.

Each child <class> element must specify the name of a decorator bean class. If there is no class with the specified name, or if the class with the specified name is not a decorator bean class, the container automatically detects the problem and treats it as a deployment problem.

If the same class is listed twice under the <decorators> element, the container automatically detects the problem and treats it as a deployment problem.

8.3. Decorator resolution

The process of matching decorators to a certain bean is called decorator resolution. A decorator is bound to a bean if:

If a decorator matches a managed bean, the managed bean class must be a proxyable bean type, as defined in Unproxyable bean types.

For a custom implementation of the Decorator interface defined in The Decorator interface, the container calls getDelegateType(), getDelegateQualifiers() and getDecoratedTypes() to determine the delegate type and qualifiers and decorated types of the decorator.

8.3.1. Assignability of raw and parameterized types for delegate injection points

Decorator delegate injection points have a special set of rules for determining assignability of raw and parameterized types, as an exception to Assignability of raw and parameterized types.

A raw bean type is considered assignable to a parameterized delegate type if the raw types are identical and all type parameters of the delegate type are either unbounded type variables or java.lang.Object.

A parameterized bean type is considered assignable to a parameterized delegate type if they have identical raw type and for each parameter:

  • the delegate type parameter and the bean type parameter are actual types with identical raw type, and, if the type is parameterized, the bean type parameter is assignable to the delegate type parameter according to these rules, or

  • the delegate type parameter is a wildcard, the bean type parameter is an actual type and the actual type is assignable to the upper bound, if any, of the wildcard and assignable from the lower bound, if any, of the wildcard, or

  • the delegate type parameter is a wildcard, the bean type parameter is a type variable and the upper bound of the type variable is assignable to the upper bound, if any, of the wildcard and assignable from the lower bound, if any, of the wildcard, or

  • the delegate type parameter and the bean type parameter are both type variables and the upper bound of the bean type parameter is assignable to the upper bound, if any, of the delegate type parameter, or

  • the delegate type parameter is a type variable, the bean type parameter is an actual type, and the actual type is assignable to the upper bound, if any, of the type variable.

8.4. Decorator invocation

Whenever a business method is invoked on an instance of a bean with decorators, the container intercepts the business method invocation and, after processing all interceptors of the method, invokes decorators of the bean.

The container searches for the first decorator of the instance that implements the method that is being invoked as a business method. If no such decorator exists, the container invokes the business method of the intercepted instance. Otherwise, the container calls the method of the decorator.

When any decorator is invoked by the container, it may in turn invoke a method of the delegate. The container intercepts the delegate invocation and searches for the first decorator of the instance such that:

  • the decorator occurs after the decorator invoking the delegate, and

  • the decorator implements the method that is being invoked upon the delegate.

If no such decorator exists, the container invokes the business method of the intercepted instance. Otherwise, the container calls the method of the decorator.

9. Interceptor bindings

Managed beans support interception. Interceptors are used to separate cross-cutting concerns from business logic. The Java Interceptors specification defines the basic programming model and semantics, and how to associate interceptors with target classes. This specification defines various extensions to the Java Interceptors specification, including how to override the interceptor order defined by the @Priority annotation.

9.1. Interceptor binding types

This specification extends the Java Interceptors specification and allows interceptor bindings to be applied to CDI stereotypes.

9.1.1. Interceptor bindings for stereotypes

Interceptor bindings may be applied to a stereotype by annotating the stereotype annotation:

@Transactional
@Secure
@RequestScoped
@Stereotype
@Target(TYPE)
@Retention(RUNTIME)
public @interface Action {}

An interceptor binding declared by a stereotype is inherited by any bean that declares that stereotype.

If a stereotype declares interceptor bindings, it must be defined as @Target(TYPE).

9.2. Declaring the interceptor bindings of an interceptor

This specification extends the Java Interceptors specification and defines how the container must behave if a definition error is encountered.

9.3. Binding an interceptor to a bean

This specification extends the Java Interceptors specification and defines:

  • additional restrictions about the type of bean to which an interceptor can be bound, and

  • how the container must behave if a definition error is encountered, and

  • how interceptors are bound using stereotypes.

Interceptor bindings may be used to associate interceptors with any managed bean that is not a decorator.

The set of interceptor bindings for a method declared at class level includes those declared on stereotypes. An interceptor binding declared on a bean class replaces an interceptor binding of the same type declared by a stereotype that is applied to the bean class.

The set of interceptor bindings for a producer method is not used to associate interceptors with the return value of the producer method.

If a managed bean has a class-level or method-level interceptor binding, the managed bean must be a proxyable bean type, as defined in Unproxyable bean types.

9.4. Interceptor enablement and ordering

This specification extends the Java Interceptors specification and defines:

  • support for enabling interceptors only for a bean archive, as defined by Contexts and Dependency Injection 1.0, and

  • the ability to override the interceptor order using the portable extension SPI, defined in AfterTypeDiscovery event.

An interceptor may be explicitly enabled for a bean archive by listing its class under the <interceptors> element of the beans.xml file of the bean archive.

<beans xmlns="http://xmlns.jcp.org/xml/ns/javaee"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://xmlns.jcp.org/xml/ns/javaee http://xmlns.jcp.org/xml/ns/javaee/beans_1_1.xsd"">
   <interceptors>
      <class>com.acme.myfwk.TransactionInterceptor</class>
      <class>com.acme.myfwk.LoggingInterceptor</class>
   </interceptors>
</beans>

The order of the interceptor declarations determines the interceptor ordering. Interceptors which occur earlier in the list are called first.

Each child <class> element must specify the name of an interceptor class. If there is no class with the specified name, or if the class with the specified name is not an interceptor class, the container automatically detects the problem and treats it as a deployment problem.

If the same class is listed twice under the <interceptors> element, the container automatically detects the problem and treats it as a deployment problem.

Interceptors enabled using @Priority are called before interceptors enabled using beans.xml.

An interceptor is said to be enabled if it is enabled in at least one bean archive or is listed in the final list of interceptors for the application, as defined in AfterTypeDiscovery event.

9.5. Interceptor resolution

This specification extends the Java Interceptors specification and defines:

  • the effect of applying @Nonbinding to an interceptor binding member, and

  • how the interceptor bindings of a custom implementation of the Interceptor interface are determined.

If any interceptor binding has a member annotated @javax.enterprise.util.Nonbinding, the member is ignored when performing interceptor resolution, and the method does not need to have the same annotation member value.

For a custom implementation of the Interceptor interface defined in The Interceptor interface, the container calls getInterceptorBindings() to determine the interceptor bindings of the interceptor and intercepts() to determine if the interceptor intercepts a given kind of lifecycle callback or business method.

10. Events

Beans may produce and consume events. This facility allows beans to interact in a completely decoupled fashion, with no compile-time dependency between the interacting beans. Most importantly, it allows stateful beans in one architectural tier of the application to synchronize their internal state with state changes that occur in a different tier.

An event comprises:

  • A Java object - the event object

  • A set of instances of qualifier types - the event qualifiers

The event object acts as a payload, to propagate state from producer to consumer. The event qualifiers act as topic selectors, allowing the consumer to narrow the set of events it observes.

An observer method acts as event consumer, observing events of a specific type - the observed event type - with a specific set of qualifiers - the observed event qualifiers. An observer method will be notified of an event if the event object is assignable to the observed event type, and if the set of observed event qualifiers is a subset of all the event qualifiers of the event.

10.1. Event types and qualifier types

An event object is an instance of a concrete Java class with no unresolvable type variables. The event types of the event include all superclasses and interfaces of the runtime class of the event object.

An event type may not contain an unresolvable type variable.

An event qualifier type is just an ordinary qualifier type as specified in Defining new qualifier types, typically defined as @Target({METHOD, FIELD, PARAMETER, TYPE}) or @Target({FIELD, PARAMETER}).

Every event has the qualifier @javax.enterprise.inject.Any, even if it does not explicitly declare this qualifier.

Any Java type may be an observed event type.

10.2. Firing events

Beans fire events via an instance of the javax.enterprise.event.Event interface, which may be injected:

@Inject Event<LoggedInEvent> loggedInEvent;

Any combination of qualifiers may be specified at the injection point:

@Inject @Admin Event<LoggedInEvent> adminLoggedInEvent;

Or the application may specify qualifiers dynamically:

@Inject Event<LoggedInEvent> loggedInEvent;
...
LoggedInEvent event = new LoggedInEvent(user);
if ( user.isAdmin() ) {
    loggedInEvent.select( new AdminQualifier() ).fire(event);
}
else {
    loggedInEvent.fire(event);
}

In this example, the event sometimes has the qualifier @Admin, depending upon the value of user.isAdmin().

10.2.1. Firing events synchronously

The method fire() accepts an event object:

public void login() {
    ...
    loggedInEvent.fire( new LoggedInEvent(user) );
}

Event fired with the fire() method is fired synchronously. All the resolved synchronous observers are called in the thread in which fire() was called. A synchronous observer notification blocks the calling thread until it completes.

10.2.2. Firing events asynchronously

Events may also be fired asynchronously using one of the methods fireAsync()

@Inject Event<LoggedInEvent> loggedInEvent;

public void login() {
    ...
    loggedInEvent.fireAsync( new LoggedInEvent(user) );
}

Event fired with the fireAsync() method is fired asynchronously. All resolved synchronous observers are called in the same thread in which fireAsync() was called, and all the resolved asynchronous observers are called in one or more different threads.

If synchronous observer have to be notified, fireAsync returns immediately after the last synchronous observer has returned. Otherwise it returns immediately.

The method fireAsync accepts an event object which can be qualified with the rules defined in Firing events.

10.2.3. The Event interface

The Event interface provides a method for firing events with a specified combination of type and qualifiers:

public interface Event<T> {

    public void fire(T event);
    public <U extends T> CompletionStage<U> fireAsync(U event);
    public <U extends T> CompletionStage<U> fireAsync(U event, Executor executor);

    public Event<T> select(Annotation... qualifiers);
    public <U extends T> Event<U> select(Class<U> subtype, Annotation... qualifiers);
    public <U extends T> Event<U> select(TypeLiteral<U> subtype, Annotation... qualifiers);

}

For an injected Event:

  • the specified type is the type parameter specified at the injection point, and

  • the specified qualifiers are the qualifiers specified at the injection point.

For example, this injected Event has specified type LoggedInEvent and specified qualifier @Admin:

@Inject @Admin Event<LoggedInEvent> any;

The select() method returns a child Event for a given specified type and additional specified qualifiers. If no specified type is given, the specified type is the same as the parent.

For example, this child Event has required type AdminLoggedInEvent and additional specified qualifier @Admin:

Event<AdminLoggedInEvent> admin = any.select(
            AdminLoggedInEvent.class,
            new AdminQualifier() );

If the specified type contains a type variable, an IllegalArgumentException is thrown.

If two instances of the same qualifier type are passed to select(), an IllegalArgumentException is thrown.

If an instance of an annotation that is not a qualifier type is passed to select(), an IllegalArgumentException is thrown.

The methods fire() and fireAsync() fire an event with the specified qualifiers and notify observers, as defined by Observer notification. If the container is unable to resolve the parameterized type of the event object, it uses the specified type to infer the parameterized type of the event types.

The method fireAsync() may be called with a specific Executor object to be used to invoke observers method. The container should provide a default Executor when fireAsync() is called without specifying one.

If the runtime type of the event object contains an unresolvable type variable, an IllegalArgumentException is thrown.

If the runtime type of the event object is assignable to the type of a container lifecycle event, an IllegalArgumentException is thrown.

If the provided executor cannot execute observers notification, an IllegalArgumentException is thrown.

10.2.4. The built-in Event

The container must provide a built-in bean with:

  • Event<X> in its set of bean types, for every Java type X that does not contain a type variable,

  • every event qualifier type in its set of qualifier types,

  • scope @Dependent,

  • no bean name, and

  • an implementation provided automatically by the container.

If an injection point of raw type Event is defined, the container automatically detects the problem and treats it as a definition error.

The built-in implementation must be a passivation capable dependency, as defined in Passivation capable dependencies.

10.3. Observer resolution

The process of matching an event to its observer methods is called observer resolution. The container considers event type and qualifiers when resolving observers.

Observer resolution usually occurs at runtime.

An event is delivered to an observer method if:

  • The observer method belongs to an enabled bean.

  • An event type is assignable to the observed event type, taking type parameters into consideration.

  • The observer method has no event qualifiers or has a subset of the event qualifiers. An observer method has an event qualifier if it has an observed event qualifier with (a) the same type and (b) the same annotation member value for each member which is not annotated @javax.enterprise.util.Nonbinding.

  • Either the event is not a container lifecycle event, as defined in Container lifecycle events, or the observer method belongs to an extension.

  • The event is fired synchronously and the observer is defined with @Observes.

  • The event is fired asynchronously and the observer is defined with @Observes or @ObservesAsync.

If the runtime type of the event object contains an unresolvable type variable, the container must throw an IllegalArgumentException.

For a custom implementation of the ObserverMethod interface defined in The ObserverMethod interface, the container must call getObservedType() and getObservedQualifiers() to determine the observed event type and qualifiers.

10.3.1. Assignability of type variables, raw and parameterized types

An event type is considered assignable to a type variable if the event type is assignable to the upper bound, if any.

A parameterized event type is considered assignable to a raw observed event type if the raw types are identical.

A parameterized event type is considered assignable to a parameterized observed event type if they have identical raw type and for each parameter:

  • the observed event type parameter is an actual type with identical raw type to the event type parameter, and, if the type is parameterized, the event type parameter is assignable to the observed event type parameter according to these rules, or

  • the observed event type parameter is a wildcard and the event type parameter is assignable to the upper bound, if any, of the wildcard and assignable from the lower bound, if any, of the wildcard, or

  • the observed event type parameter is a type variable and the event type parameter is assignable to the upper bound, if any, of the type variable.

10.3.2. Event qualifier types with members

As usual, the qualifier type may have annotation members:

@Qualifier
@Target(PARAMETER)
@Retention(RUNTIME)
public @interface Role {
    String value();
}

Consider the following event:

@Inject Event<LoggedInEvent> loggedInEvent;
...
public void login() {
    final User user = ...;
    loggedInEvent.select(new RoleQualifier() { public String value() { return user.getRole(); } }).fire(new LoggedInEvent(user));
}

Where RoleQualifier is an implementation of the qualifier type Role:

public abstract class RoleQualifier
        extends AnnotationLiteral<Role>
        implements Role {}

Then the following observer method will always be notified of the event:

public void afterLogin(@Observes LoggedInEvent event) { ... }

Whereas this observer method may or may not be notified, depending upon the value of user.getRole():

public void afterAdminLogin(@Observes @Role("admin") LoggedInEvent event) { ... }

As usual, the container uses equals() to compare event qualifier type member values.

10.3.3. Multiple event qualifiers

An event parameter may have multiple qualifiers.

public void afterDocumentUpdatedByAdmin(@Observes @Updated @ByAdmin Document doc) { ... }

Then this observer method will only be notified if all the observed event qualifiers are specified when the event is fired:

@Inject Event<Document> documentEvent;
...
documentEvent.select(new UpdatedQualifier(), new ByAdminQualifier()).fire(document);

Other, less specific, observers will also be notified of this event:

public void afterDocumentUpdated(@Observes @Updated Document doc) { ... }
public void afterDocumentEvent(@Observes Document doc) { ... }

10.4. Observer methods

An observer method allows the application to receive and respond to event notifications.

An observer method is a non-abstract method of a managed bean class or of an extension, as defined in Container lifecycle events. An observer method may be either static or non-static.

There may be arbitrarily many observer methods with the same event parameter type and qualifiers.

A bean (or extension) may declare multiple observer methods.

10.4.1. Event parameter of an observer method

Each observer method must have exactly one event parameter, of the same type as the event type it observes. When searching for observer methods for an event, the container considers the type and qualifiers of the event parameter.

If the event parameter does not explicitly declare any qualifier, the observer method observes events with no qualifier.

The event parameter type may contain a type variable or wildcard.

The event parameter may be an array type whose component type contains a type variable or a wildcard.

Modifications made to the event parameter in an observer method are propagated to following observers. The container is not required to guarantee a consistent state for an event parameter modified by asynchronous observers.

10.4.2. Declaring an observer method

An observer method may be declared by annotating a parameter @javax.enterprise.event.Observes or @javax.enterprise.event.ObservesAsync+` of a default-access, public, protected or private method. That parameter is the event parameter. The declared type of the parameter is the observed event type.

If @Observes is used the observer method is a synchronous observer method.

If @ObservesAsync is used the observer method is an asynchronous observer method.

public void afterLogin(@Observes LoggedInEvent event) { ... }

public void asyncAfterLogin(@ObservesAsync LoggedInEvent event) { ... }

If a method has more than one parameter annotated @Observes or @ObservesAsync, the container automatically detects the problem and treats it as a definition error.

Observed event qualifiers may be declared by annotating the event parameter:

public void afterLogin(@Observes @Admin LoggedInEvent event) { ... }

If an observer method is annotated @Produces or @Inject or has a parameter annotated @Disposes, the container automatically detects the problem and treats it as a definition error.

Interceptors and decorators may not declare observer methods. If an interceptor or decorator has a method with a parameter annotated @Observes, the container automatically detects the problem and treats it as a definition error.

In addition to the event parameter, observer methods may declare additional parameters, which may declare qualifiers. These additional parameters are injection points.

public void afterLogin(@Observes LoggedInEvent event, @Manager User user, Logger log) { ... }

10.4.3. The EventMetadata interface

The interface javax.enterprise.inject.spi.EventMetadata provides access to metadata about an observed event.

public interface EventMetadata {
    public Set<Annotation> getQualifiers();
    public InjectionPoint getInjectionPoint();
    public Type getType();
    public boolean isAsync();
}
  • getQualifiers() returns the set of qualifiers with which the event was fired.

  • getInjectionPoint() returns the InjectionPoint from which this event payload was fired, or null if it was fired from BeanManager.fireEvent(…​).

  • getType() returns the type representing runtime class of the event object with type variables resolved.

  • isAsync() returns true if the event was fired with fireAsync() or fireAsyncEvent() methods otherwise returns false.

The container must provide a bean with scope @Dependent, bean type EventMetadata and qualifier @Default, allowing observer methods to obtain information about the events they observe.

If an injection point of type EventMetadata and qualifier @Default which is not a parameter of an observer method exists, the container automatically detects the problem and treats it as a definition error.

public void afterLogin(@Observes LoggedInEvent event, EventMetadata metadata) { ... }

10.4.4. Conditional observer methods

A conditional observer method is an observer method which is notified of an event only if an instance of the bean that defines the observer method already exists in the current context.

A conditional observer method may be declared by specifying notifyObserver=IF_EXISTS.

public void refreshOnDocumentUpdate(@Observes(notifyObserver=IF_EXISTS) @Updated Document doc) { ... }

public void asyncRefreshOnDocumentUpdate(@ObservesAsync(notifyObserver=IF_EXISTS) @Updated Document doc) { ... }

Beans with scope @Dependent may not have conditional observer methods. If a bean with scope @Dependent has an observer method declared notifyObserver=IF_EXISTS, the container automatically detects the problem and treats it as a definition error.

The enumeration javax.enterprise.event.Reception identifies the possible values of notifyObserver:

public enum Reception { IF_EXISTS, ALWAYS }

10.4.5. Transactional observer methods

Transactional observer methods are observer methods which receive event notifications during the before or after completion phase of the transaction in which the event was fired. If no transaction is in progress when the event is fired, they are notified at the same time as other observers.

  • A before completion observer method is called during the before completion phase of the transaction.

  • An after completion observer method is called during the after completion phase of the transaction.

  • An after success observer method is called during the after completion phase of the transaction, only when the transaction completes successfully.

  • An after failure observer method is called during the after completion phase of the transaction, only when the transaction fails.

The enumeration javax.enterprise.event.TransactionPhase identifies the kind of transactional observer method:

public enum TransactionPhase {
    IN_PROGRESS,
    BEFORE_COMPLETION,
    AFTER_COMPLETION,
    AFTER_FAILURE,
    AFTER_SUCCESS
}

A transactional observer method may be declared by specifying any value other than IN_PROGRESS for during:

void onDocumentUpdate(@Observes(during=AFTER_SUCCESS) @Updated Document doc) { ... }

Asynchronous observer cannot be declared transactional.

10.5. Observer notification

When an event is fired by the application, the container must:

  • determine the observer methods for that event according to the rules of observer resolution defined by Observer resolution, then,

  • for each observer method, either invoke the observer method immediately, or register the observer method for later invocation during the transaction completion phase, using a JTA Synchronization.

  • honor the priority of observer methods as defined in Observer ordering.

The container calls observer methods as defined in Invocation of observer methods.

  • If the observer method is a transactional observer method and there is currently a JTA transaction in progress, the container calls the observer method during the appropriate transaction completion phase.

  • If there is no context active for the scope to which the bean declaring the observer method belongs, then the observer method should not be called.

  • Otherwise, the container calls the observer immediately.

Any observer method called before completion of a transaction may call setRollbackOnly() to force a transaction rollback. An observer method may not directly initiate, commit or rollback JTA transactions.

Observer methods may throw exceptions:

  • If the observer method is a transactional observer method, any exception is caught and logged by the container.

  • If the observer method is asynchronous, the exception aborts processing of the observer but not of the event. Exception management during an asynchronous event is defined in Handling multiple exceptions thrown during an asynchronous event.

  • Otherwise, the exception aborts processing of the event. No other observer methods of that event will be called. The BeanManager.fireEvent() or Event.fire() method rethrows the exception. If the exception is a checked exception, it is wrapped and rethrown as an (unchecked) ObserverException.

For a custom implementation of the ObserverMethod interface defined in The ObserverMethod interface, the container must call getTransactionPhase() to determine if the observer method is transactional observer method, and notify() to invoke the method.

10.5.1. Handling multiple exceptions thrown during an asynchronous event

If an event is asynchronous, and an exception is thrown by any of its notified observers, the CompletionStage returned by fireAsync will complete exceptionally with FireAsyncException exception. FireAsyncException contains all exceptions thrown by observers as suppressed exceptions. They can be accessed as an array of throwable with the getSuppressed method.

It can be handled with one of the CompletionStage methods related to exceptions:

myEvent.fireAsync(anEventObject)
       .handle((ok, ex) -> {
           if (ok != null) {
               return ok;
           } else {
             for (Throwable t : ex.getSuppressed()) {
                      ...
             }
              ...
           } });

If no exception is thrown by observers then the resulting CompletionStage is completed normally with the event object.

10.5.2. Observer ordering

Before the actual observer notification, the container determines an order in which the observer methods for a certain event are invoked. The priority of an observer method may be declared using the @Priority annotation. If no @Priority annotation is specified, the default priority APPLICATION + 500 is assumed. Observers with smaller priority values are called first.

void afterLogin(@Observes @Priority(APPLICATION) LoggedInEvent event) { ... }

The order of more than one observer with the same priority is undefined and the observer methods are notified therefore in a non predictable order.

10.5.3. Observer method invocation context

The transaction context, client security context and lifecycle contexts active when an observer method is invoked depend upon what kind of observer method it is.

  • If the observer method is asynchronous, it is called in a new lifecycle contexts, a new transaction context but the same security context as the invocation of Event.fireAsync().

  • If the observer method is a before completion transactional observer method, it is called within the context of the transaction that is about to complete and with the same client security context and lifecycle contexts.

  • Otherwise, if the observer method is any other kind of transactional observer method, it is called in an unspecified transaction context, but with the same client security context and lifecycle contexts as the transaction that just completed.

  • Otherwise, the observer method is called in the same transaction context, client security context and lifecycle contexts as the invocation of Event.fire() or BeanManager.fireEvent().

11. Portable extensions

A portable extension may integrate with the container by:

  • Providing its own beans, interceptors and decorators to the container

  • Injecting dependencies into its own objects using the dependency injection service

  • Providing a context implementation for a custom scope

  • Augmenting or overriding the annotation-based metadata with metadata from some other source

11.1. The Bean interface

The BeanAttributes interface exposes the basic attributes of a bean.

public interface BeanAttributes<T> {
    public Set<Type> getTypes();
    public Set<Annotation> getQualifiers();
    public Class<? extends Annotation> getScope();
    public String getName();
    public Set<Class<? extends Annotation>> getStereotypes();
    public boolean isAlternative();
}
  • getTypes(), getQualifiers(), getScope(), getName() and getStereotypes() must return the bean types, qualifiers, scope type, bean name and stereotypes of the bean, as defined in Concepts.

  • isAlternative() must return true if the bean is an alternative, and false otherwise.

The interface javax.enterprise.inject.spi.Bean defines everything the container needs to manage instances of a certain bean.

public interface Bean<T> extends Contextual<T>, BeanAttributes<T> {
    public Class<?> getBeanClass();
    public Set<InjectionPoint> getInjectionPoints();
    public boolean isNullable();
}
  • getBeanClass() returns the bean class of the managed bean or of the bean that declares the producer method or field.

  • getInjectionPoints() returns a set of InjectionPoint objects, defined in Injection point metadata, representing injection points of the bean, that will be validated by the container at initialization time.

  • isNullable() is deprecated in CDI 1.1 and should be ignored by the container.

Note that implementations of Bean must also implement the inherited operations defined by the Contextual interface defined in The Contextual interface.

An instance of Bean must exist for every enabled bean.

A portable extension may add support for new kinds of beans beyond those defined by the this specification by implementing Bean and registering beans with the container, using the mechanism defined in AfterBeanDiscovery event.

Custom implementations of Bean are encouraged to implement PassivationCapable and may be required to in later revisions of this specification.

11.1.1. The Decorator interface

The Bean object for a decorator must implement the interface javax.enterprise.inject.spi.Decorator.

public interface Decorator<T> extends Bean<T> {
    public Set<Type> getDecoratedTypes();
    public Type getDelegateType();
    public Set<Annotation> getDelegateQualifiers();
}
  • getDecoratedTypes() returns the decorated types of the decorator.

  • getDelegateType() and getDelegateQualifiers() return the delegate type and qualifiers of the decorator.

An instance of Decorator exists for every enabled decorator.

11.1.2. The Interceptor interface

The Bean object for an interceptor must implement javax.enterprise.inject.spi.Interceptor.

public interface Interceptor<T> extends Bean<T> {
    public Set<Annotation> getInterceptorBindings();
    public boolean intercepts(InterceptionType type);
    public Object intercept(InterceptionType type, T instance, InvocationContext ctx) throws Exception;
}
  • getInterceptorBindings() returns the interceptor bindings of the interceptor.

  • intercepts() returns true if the interceptor intercepts the specified kind of lifecycle callback or method invocation, and false otherwise.

  • intercept() invokes the specified kind of lifecycle callback or method invocation interception upon the given instance of the interceptor.

An InterceptionType identifies the kind of lifecycle callback or business method.

public enum InterceptionType {
    AROUND_INVOKE, AROUND_CONSTRUCT, POST_CONSTRUCT, PRE_DESTROY, PRE_PASSIVATE, POST_ACTIVATE, AROUND_TIMEOUT
}

An instance of Interceptor exists for every enabled interceptor.

11.1.3. The ObserverMethod interface

The interface javax.enterprise.inject.spi.ObserverMethod defines everything the container needs to know about an observer method.

public interface ObserverMethod<T> {
    public Class<?> getBeanClass();
    public Type getObservedType();
    public Set<Annotation> getObservedQualifiers();
    public Reception getReception();
    public TransactionPhase getTransactionPhase();
    public int getPriority();
    public void notify(T event);
    public boolean isAsync();
}
  • getBeanClass() returns the class of the type that declares the observer method.

  • getObservedType() and getObservedQualifiers() return the observed event type and qualifiers.

  • getReception() returns IF_EXISTS for a conditional observer and ALWAYS otherwise.

  • getTransactionPhase() returns the appropriate transaction phase for a transactional observer method or IN_PROGRESS otherwise.

  • getPriority() returns the priority that will be used by the container to determine the notification order.

  • notify() calls the observer method, as defined in Observer notification.

  • isAsync() returns true if the method is an asynchronous observer method otherwise returns false.

An instance of ObserverMethod must exist for every observer method of every enabled bean.

11.2. The Producer and InjectionTarget interfaces

The interface javax.enterprise.inject.spi.Producer provides a generic operation for producing an instance of a type.

public interface Producer<T> {
    public T produce(CreationalContext<T> ctx);
    public void dispose(T instance);
    public Set<InjectionPoint> getInjectionPoints();
}

For a Producer that represents a class:

  • produce() calls the constructor annotated @Inject if it exists, or the constructor with no parameters otherwise, as defined in Injection using the bean constructor, and returns the resulting instance. If the class has interceptors, produce() is responsible for building the interceptors and decorators of the instance. The instance returned by produce() may be a proxy.

  • dispose() does nothing.

  • getInjectionPoints() returns the set of InjectionPoint objects representing all injected fields, bean constructor parameters and initializer method parameters.

For a Producer that represents a producer method or field:

  • produce() calls the producer method on, or accesses the producer field of, a contextual instance of the bean that declares the producer method, as defined in Invocation of producer or disposer methods.

  • dispose() calls the disposer method, if any, on a contextual instance of the bean that declares the disposer method, as defined in Invocation of producer or disposer methods, or performs any additional required cleanup, if any, to destroy state associated with a resource.

  • getInjectionPoints() returns the set of InjectionPoint objects representing all parameters of the producer method.

The subinterface javax.enterprise.inject.spi.InjectionTarget provides operations for performing dependency injection and lifecycle callbacks on an instance of a type.

public interface InjectionTarget<T>
        extends Producer<T> {
    public void inject(T instance, CreationalContext<T> ctx);
    public void postConstruct(T instance);
    public void preDestroy(T instance);
}
  • inject() performs dependency injection upon the given object. The container sets the value of all injected fields, and calls all initializer methods, as defined in Injection of fields and initializer methods.

  • postConstruct() calls the @PostConstruct callback, if it exists.

  • preDestroy() calls the @PreDestroy callback, if it exists.

Implementations of Producer and InjectionTarget must ensure that the set of injection points returned by getInjectionPoints() are injected by produce() or inject().

11.3. The BeanManager object

The interface javax.enterprise.inject.spi.BeanManager provides operations for obtaining contextual references for beans, along with many other operations of use to portable extensions.

The container provides a built-in bean with bean type BeanManager, scope @Dependent and qualifier @Default. The built-in implementation must be a passivation capable dependency, as defined in Passivation capable dependencies. Thus, any bean may obtain an instance of BeanManager by injecting it:

@Inject BeanManager manager;

Note that, an exception is thrown if the following operations are called before the AfterBeanDiscovery event is fired:

  • getBeans(String),

  • getBeans(Type, Annotation…​),

  • getPassivationCapableBean(String)

  • resolve(Set),

  • resolveDecorators(Set, Annotation…​),

  • resolveInterceptors(InterceptionType, Annotation…​),

  • resolveObserverMethods(Object, Annotation…​),

  • validate(InjectionPoint),

and if the following operations are called before the AfterDeploymentValidation event is fired:

  • getReference(Bean, Type, CreationalContext)

  • getInjectableReference(InjectionPoint, CreationalContext).

All other operations of BeanManager may be called at any time during the execution of the application.

11.3.1. Obtaining a reference to the CDI container

Portable extensions and other objects sometimes interact directly with the container via programmatic API call. The abstract javax.enterprise.inject.spi.CDI provides access to the BeanManager as well providing lookup of bean instances.

public abstract class CDI<T> implements Instance<T> {
   public static CDI<Object> current() { ... }
   public static void setCDIProvider(CDIProvider provider);
   public abstract BeanManager getBeanManager();
}

A portable extension or other object may obtain a reference to the current container by calling CDI.current(). CDI.getBeanManager() may be called at any time after the container fires the BeforeBeanDiscovery container lifecycle event until the container fires the BeforeShutdown container lifecycle event. Other methods on CDI may be called after the application initialization is completed until the application shutdown starts. If methods on CDI are called at any other time, non-portable behavior results.

When CDI.current() is called, getCDI() method is called on javax.enterprise.inject.spi.CDIProvider.

The CDIProvider to use may be set by the application or container using the setCDIProvider() method. If the setCDIProvider() has not been called, the first service provider of the service javax.enterprise.inject.spi.CDIProvider declared in META-INF/services is used. If no provider is available an IllegalStateException is thrown.

public interface CDIProvider {
   public CDI<Object> getCDI();
   boolean isInitialized();
   default CDI<Object> initialize();
   CDI<Object> initialize(Map<String,Object> params);
}

11.3.2. Obtaining a contextual reference for a bean

The method BeanManager.getReference() returns a contextual reference for a given bean and bean type, as defined in Contextual reference for a bean.

public Object getReference(Bean<?> bean, Type beanType, CreationalContext<?> ctx);

The first parameter is the Bean object representing the bean. The second parameter represents a bean type that must be implemented by any client proxy that is returned. The third parameter is an instance of CreationalContext that may be used to destroy any object with scope @Dependent that is created.

If the given type is not a bean type of the given bean, an IllegalArgumentException is thrown.

11.3.3. Obtaining an injectable reference

The method BeanManager.getInjectableReference() returns an injectable reference for a given injection point, as defined in Injectable references.

public Object getInjectableReference(InjectionPoint ij, CreationalContext<?> ctx);

The first parameter represents the target injection point. The second parameter is an instance of CreationalContext that may be used to destroy any object with scope @Dependent that is created.

If the InjectionPoint represents a decorator delegate injection point, getInjectableReference() returns a delegate, as defined in Decorator delegate injection points.

If typesafe resolution results in an unsatisfied dependency, the container must throw an UnsatisfiedResolutionException. If typesafe resolution results in an unresolvable ambiguous dependency, the container must throw an AmbiguousResolutionException.

Implementations of Bean usually maintain a reference to an instance of BeanManager. When the Bean implementation performs dependency injection, it must obtain the contextual instances to inject by calling BeanManager.getInjectableReference(), passing an instance of InjectionPoint that represents the injection point and the instance of CreationalContext that was passed to Bean.create().

11.3.4. Obtaining non-contextual instance

A non-contextual instance can be obtained and injected from an InjectionTarget, however the InjectionTarget interface is designed to work on contextual instances. A helper class, Unmanaged provides a set of methods optimized for working with non-contextual instances.

For example:

Unmanaged<Foo> unmanagedFoo = new Unmanaged<Foo>(Foo.class);
UnmanagedInstance<Foo> fooInstance = unmanagedFoo.newInstance();
Foo foo = fooInstance.produce().inject().postConstruct().get();
... // Use the foo instance
fooInstance.preDestroy().dispose();

11.3.5. Obtaining a CreationalContext

An instance of CreationalContext for a certain instance of Contextual may be obtained by calling BeanManager.createCreationalContext().

public <T> CreationalContext<T> createCreationalContext(Contextual<T> contextual);

An instance of CreationalContext for a non-contextual object may be obtained by passing a null value to createCreationalContext().

11.3.6. Obtaining a Bean by type

The method BeanManager.getBeans() returns the set of beans which have the given required type and qualifiers and are available for injection in the module or library containing the class into which the BeanManager was injected or the class from whose JNDI environment namespace the BeanManager was obtained, according to the rules for candidates of typesafe resolution defined in Performing typesafe resolution.

public Set<Bean<?>> getBeans(Type beanType, Annotation... qualifiers);

The first parameter is a required bean type. The remaining parameters are required qualifiers.

If no qualifiers are passed to getBeans(), the default qualifier @Default is assumed.

If the given type represents a type variable, an IllegalArgumentException is thrown.

If two instances of the same qualifier type are given, an IllegalArgumentException is thrown.

If an instance of an annotation that is not a qualifier type is given, an IllegalArgumentException is thrown.

11.3.7. Obtaining a Bean by name

The method BeanManager.getBeans() which accepts a string returns the set of beans which have the given bean name and are available for injection in the module or library containing the class into which the BeanManager was injected or the class from whose JNDI environment namespace the BeanManager was obtained, according to the rules of name resolution defined in Name resolution.

public Set<Bean<?>> getBeans(String name);

The parameter is a bean name.

11.3.8. Obtaining a passivation capable bean by identifier

The method BeanManager.getPassivationCapableBean() returns the PassivationCapable bean with the given identifier (see Passivation capable beans).

public Bean<?> getPassivationCapableBean(String id);

11.3.9. Resolving an ambiguous dependency

The method BeanManager.resolve() applies the ambiguous dependency resolution rules defined in Unsatisfied and ambiguous dependencies to a set of Bean s.

public <X> Bean<? extends X> resolve(Set<Bean<? extends X>> beans);

If the ambiguous dependency resolution rules fail (as defined in Unsatisfied and ambiguous dependencies, the container must throw an AmbiguousResolutionException.

BeanManager.resolve() must return null if:

  • null is passed to resolve(), or

  • no beans are passed to resolve().

11.3.10. Validating an injection point

The BeanManager.validate() operation validates an injection point and throws an InjectionException if there is a deployment problem (for example, an unsatisfied or unresolvable ambiguous dependency) associated with the injection point.

public void validate(InjectionPoint injectionPoint);

11.3.11. Firing an event

The method BeanManager.fireEvent() fires an event and notifies observers, according to Observer notification.

public void fireEvent(Object event, Annotation... qualifiers);

The first argument is the event object. The remaining parameters are event qualifiers.

If the runtime type of the event object contains a type variable, an IllegalArgumentException is thrown.

If two instances of the same qualifier type are given, an IllegalArgumentException is thrown.

If an instance of an annotation that is not a qualifier type is given, an IllegalArgumentException is thrown.

If the runtime type of the event object is assignable to the type of a container lifecycle event, an IllegalArgumentException is thrown.

11.3.12. Observer method resolution

The method BeanManager.resolveObserverMethods() resolves observer methods for an event according to the rules of observer resolution defined in Observer resolution.

public <T> Set<ObserverMethod<? super T>> resolveObserverMethods(T event, Annotation... qualifiers);

The first parameter of resolveObserverMethods() is the event object. The remaining parameters are event qualifiers.

If the runtime type of the event object contains a type variable, an IllegalArgumentException is thrown.

If two instances of the same qualifier type are given, an IllegalArgumentException is thrown.

If an instance of an annotation that is not a qualifier type is given, an IllegalArgumentException is thrown.

11.3.13. Decorator resolution

The method BeanManager.resolveDecorators() returns the ordered list of decorators for a set of bean types and a set of qualifiers and which are enabled in the module or library containing the class into which the BeanManager was injected or the class from whose JNDI environment namespace the BeanManager was obtained, as defined in Decorator resolution.

List<Decorator<?>> resolveDecorators(Set<Type> types, Annotation... qualifiers);

The first argument is the set of bean types of the decorated bean. The annotations are qualifiers declared by the decorated bean.

If two instances of the same qualifier type are given, an IllegalArgumentException is thrown.

If an instance of an annotation that is not a qualifier type is given, an IllegalArgumentException is thrown.

If the set of bean types is empty, an IllegalArgumentException is thrown.

11.3.14. Interceptor resolution

The method BeanManager.resolveInterceptors() returns the ordered list of interceptors for a set of interceptor bindings and a type of interception and which are enabled in the module or library containing the class into which the BeanManager was injected or the class from whose JNDI environment namespace the BeanManager was obtained, as defined in Interceptor resolution.

List<Interceptor<?>> resolveInterceptors(InterceptionType type,
                                         Annotation... interceptorBindings);

If two instances of the same interceptor binding type are given, an IllegalArgumentException is thrown.

If no interceptor binding type instance is given, an IllegalArgumentException is thrown.

If an instance of an annotation that is not an interceptor binding type is given, an IllegalArgumentException is thrown.

11.3.15. Determining if an annotation is a qualifier type, scope type, stereotype or interceptor binding type

A portable extension may test an annotation to determine if it is a qualifier type, scope type, stereotype or interceptor binding type, obtain the set of meta-annotations declared by a stereotype or interceptor binding type, or determine if a scope type is a normal or passivating scope.

public boolean isScope(Class<? extends Annotation> annotationType);
public boolean isQualifier(Class<? extends Annotation> annotationType);
public boolean isInterceptorBinding(Class<? extends Annotation> annotationType);
public boolean isStereotype(Class<? extends Annotation> annotationType);

public boolean isNormalScope(Class<? extends Annotation> scopeType);
public boolean isPassivatingScope(Class<? extends Annotation> scopeType);
public Set<Annotation> getInterceptorBindingDefinition(Class<? extends Annotation> qualifierType);
public Set<Annotation> getStereotypeDefinition(Class<? extends Annotation> stereotype);

11.3.16. Determining the hash code and equivalence of qualifiers and interceptor bindings

A portable extension may determine if two qualifiers or two interceptor bindings are considered equivalent for the purposes of typesafe resolution, as defined in Performing typesafe resolution.

public boolean areQualifiersEquivalent(Annotation qualifier1, Annotation qualifier2);
public boolean areInterceptorBindingsEquivalent(Annotation interceptorBinding1, Annotation interceptorBinding2);

A portable extension may determine the hash code of a qualifier or interceptor binding, ignoring any members annotated with @Nonbinding.

public int getQualifierHashCode(Annotation qualifier);
public int getInterceptorBindingHashCode(Annotation interceptorBinding);

11.3.17. Obtaining the active Context for a scope

The method BeanManager.getContext() retrieves an active context object associated with the given scope, as defined in The active context object for a scope.

public Context getContext(Class<? extends Annotation> scopeType);

11.3.18. Obtaining the ELResolver

The method BeanManager.getELResolver() returns the javax.el.ELResolver specified in Integration with Unified EL.

public ELResolver getELResolver();

11.3.19. Obtaining an AnnotatedType for a class

The method BeanManager.createAnnotatedType() returns an AnnotatedType that may be used to read the annotations of the given Java class or interface.

public <T> AnnotatedType<T> createAnnotatedType(Class<T> type);

11.3.20. Obtaining an InjectionTarget for a class

The method BeanManager.getInjectionTargetFactory() returns a factory capable of creating container provided implementations of InjectionTarget for a given AnnotatedType or throws an IllegalArgumentException if there is a definition error associated with any injection point of the type.

public <T> InjectionTargetFactory<T> getInjectionTargetFactory(AnnotatedType<T> type);
public interface InjectionTargetFactory<T> {

    public InjectionTarget<T> createInjectionTarget(Bean<T> bean);

}

Null should be passed to InjectionTargetFactory.createInjectionTarget() to create a non-contextual injection target. The method BeanManager.createInjectionTarget() is deprecated since version 1.1 of Contexts and Dependency Injection.

11.3.21. Obtaining a Producer for a field or method

The method BeanManager.getProducerFactory() returns a factory capable of creating container provided implementations of Producer for a given AnnotatedMethod or AnnotatedField, and declaring bean, or throws an IllegalArgumentException if there is a definition error associated with the producer method or field.

public <X> ProducerFactory<X> getProducerFactory(AnnotatedField<? super X> field, Bean<X> declaringBean);
public <X> ProducerFactory<X> getProducerFactory(AnnotatedMethod<? super X> method, Bean<X> declaringBean);
public interface ProducerFactory<X> {

    public <T> Producer<T> createProducer(Bean<T> bean);

}

Null should be passed to ProducerFactory.createProducer() to create a producer of non-contextual objects.

11.3.22. Obtaining an InjectionPoint

The method BeanManager.createInjectionPoint() returns a container provided implementation of InjectionPoint for a given AnnotatedField or AnnotatedParameter or throws an IllegalArgumentException if there is a definition error associated with the injection point.

public InjectionPoint createInjectionPoint(AnnotatedField<?> field);
public InjectionPoint createInjectionPoint(AnnotatedParameter<?> parameter);

11.3.23. Obtaining a BeanAttributes

The method BeanManager.createBeanAttributes() returns a container provided implementation of BeanAttributes by reading the annotations of a given AnnotatedType or AnnotatedMember, according to the rules defined in Concepts, or throws an IllegalArgumentException if there is a definition error associated with the declared bean attributes.

public <T> BeanAttributes<T> createBeanAttributes(AnnotatedType<T> type);
public BeanAttributes<?> createBeanAttributes(AnnotatedMember<?> member);

11.3.24. Obtaining a Bean

The method BeanManager.createBean() returns a container provided implementation of Bean. The methods accept:

  • a BeanAttributes, which determines the bean types, qualifiers, scope, name and stereotypes of the returned Bean, and the return values of isAlternative(), and

  • a class, which determines the return value of Bean.getClass().

  • an InjectionTargetFactory, which is used to obtain an InjectionTarget. The InjectionTarget is used to create and destroy instances of the bean, to perform dependency injection and lifecycle callbacks, and which determines the return value of Bean.getInjectionPoints().

public <T> Bean<T> createBean(BeanAttributes<T> attributes, Class<T> beanClass,
                          InjectionTargetFactory<T> injectionTargetFactory);

A second version of the method is provided to create a Bean from a producer. The method accepts:

  • a BeanAttributes, which determines the bean types, qualifiers, scope, name and stereotypes of the returned Bean, and the return values of isAlternative(), and

  • a class, which determines the return value of Bean.getClass().

  • a ProducerFactory, which is used to obtain a Producer. The Producer is used to create and destroy instances of the bean, and which determines the return value of Bean.getInjectionPoints().

public <T, X> Bean<T> createBean(BeanAttributes<T> attributes, Class<X> beanClass,
                          ProducerFactory<X> producer);

11.3.25. Obtaining the instance of an Extension

The method BeanManager.getExtension() returns the container’s instance of an Extension class declared in META-INF/services, or throws an IllegalArgumentException if the container has no instance of the given class.

public <T extends Extension> T getExtension(Class<T> extensionClass);

11.4. Alternative metadata sources

A portable extension may provide an alternative metadata source, such as configuration by XML.

The interfaces AnnotatedType, AnnotatedField, AnnotatedMethod, AnnotatedConstructor and AnnotatedParameter in the package javax.enterprise.inject.spi allow a portable extension to specify metadata that overrides the annotations that exist on a bean class. The portable extension is responsible for implementing the interfaces, thereby exposing the metadata to the container.

In general, the behavior is as defined by the Java Language Specification, and only deviations from the Java Language Specification are noted.

The interface javax.enterprise.inject.spi.AnnotatedType exposes the Class object and members.

public interface AnnotatedType<X>
        extends Annotated {
    public Class<X> getJavaClass();
    public Set<AnnotatedConstructor<X>> getConstructors();
    public Set<AnnotatedMethod<? super X>> getMethods();
    public Set<AnnotatedField<? super X>> getFields();
}
  • getConstructors() returns all default-access, public, protected or private constructors declared for the type.

  • getMethods() returns all default-access, public, protected or private methods declared on the type and those declared on any supertypes. The container should call AnnotatedMethod.getJavaMember().getDeclaringClass() to determine the type in the type hierarchy that declared the method.

  • getFields() returns all default-access, public, protected or private fields declared on the type and those declared on any supertypes. The container should call AnnotatedField.getJavaMember().getDeclaringClass() to determine the type in the type hierarchy that declared the field.

When determining annotations on a type, the container must only consider the special inheritance rules defined for scope types in Inheritance of type-level metadata.

The interface javax.enterprise.inject.spi.AnnotatedField exposes the Field object.

public interface AnnotatedField<X>
        extends AnnotatedMember<X> {
    public Field getJavaMember();
}

The interface javax.enterprise.inject.spi.AnnotatedMethod exposes the Method object.

public interface AnnotatedMethod<X>
        extends AnnotatedCallable<X> {
    public Method getJavaMember();
}

The interface javax.enterprise.inject.spi.AnnotatedConstructor exposes the Constuctor object.

public interface AnnotatedConstructor<X>
        extends AnnotatedCallable<X> {
    public Constructor<X> getJavaMember();
}

The interface javax.enterprise.inject.spi.AnnotatedParameter exposes the position of the parameter object and the declaring program element.

public interface AnnotatedParameter<X>
        extends Annotated {
    public int getPosition();
    public AnnotatedCallable<X> getDeclaringCallable();
}

The interface javax.enterprise.inject.spi.AnnotatedMember exposes the Member object and the AnnotatedType that defines the member.

public interface AnnotatedMember<X>
        extends Annotated {
    public Member getJavaMember();
    public boolean isStatic();
    public AnnotatedType<X> getDeclaringType();
}

The interface javax.enterprise.inject.spi.AnnotatedCallable exposes the parameters of an invokable object.

CDI 1.1 deprecated the method AnnotatedMember.isStatic(). The container should instead call AnnotatedMember.getJavaMember().getModifiers() to determine if the member is static.

public interface AnnotatedCallable<X>
        extends AnnotatedMember<X> {
    public List<AnnotatedParameter<X>> getParameters();
}

The interface javax.enterprise.inject.spi.Annotated exposes the overriding annotations and type declarations.

public interface Annotated {
    public Type getBaseType();
    public Set<Type> getTypeClosure();
    public <T extends Annotation> T getAnnotation(Class<T> annotationType);
    public Set<Annotation> getAnnotations();
    public boolean isAnnotationPresent(Class<? extends Annotation> annotationType);
}
  • getBaseType() returns the type of the program element.

  • getTypeClosure() returns all types to which the base type should be considered assignable.

  • getAnnotation() returns the program element annotation of the given annotation type, or a null value.

  • getAnnotations() returns all annotations of the program element.

  • isAnnotationPresent() returns true if the program element has an annotation of the given annotation type, or false otherwise.

The container must use the operations of Annotated and its subinterfaces to discover program element types and annotations. The container must not directly call the Java Reflection API. In particular, the container must:

  • call Annotated.getBaseType() to determine the type of an injection point, event parameter or disposed parameter,

  • call Annotated.getTypeClosure() to determine the bean types of any kind of bean,

  • call Annotated.getAnnotations() to determine the scope, qualifiers, stereotypes and interceptor bindings of a bean,

  • call Annotated.isAnnotationPresent() and Annotated.getAnnotation() to read any bean annotations defined by this specification, and

  • call AnnotatedType.getConstructors(), AnnotatedType.getMethods() and AnnotatedType.getFields() to determine the members of a bean class.

11.5. Container lifecycle events

During the application initialization process, the container fires a series of events, allowing portable extensions to integrate with the container initialization process defined in Application initialization lifecycle. These events are fired synchronously.

Observer methods of these events must belong to extensions. An extension is a service provider of the service javax.enterprise.inject.spi.Extension declared in META-INF/services.

public interface Extension {}

If any method on the event object is called outside of the observer method invocation, an IllegalStateException is thrown.

Service providers may have observer methods, which may observe any event, including any container lifecycle event, and obtain an injected BeanManager reference. Any decorators associated with BeanManager will not be applied. If other beans are injected into an extension’s observer methods, non-portable behavior results. An extension may use BeanManager.fireEvent() to deliver events to observer methods defined on extensions. The container is not required to deliver events fired during application initialization to observer methods defined on beans.

The container instantiates a single instance of each extension at the beginning of the application initialization process and maintains a reference to it until the application shuts down. The container delivers event notifications to this instance by calling its observer methods.

The notification order for observer methods within extensions follows the same ordering rule as defined in Observer ordering for non-extension observers. The priority of an observer method may be declared using the @Priority annotation.

void beforeBeanDiscovery(@Observes @Priority(LIBRARY) BeforeBeanDiscovery event) { ... }

For each service provider, the container must provide a bean of scope @ApplicationScoped and qualifier @Default, supporting injection of a reference to the service provider instance. The bean types of this bean include the class of the service provider and all superclasses and interfaces.

Lifecycle events described below can be grouped in to two categories:

  • Application lifecycle events, that are fired once:

    • BeforeBeanDiscovery

    • AfterTypeDiscovery

    • AfterBeanDiscovery

    • AfterDeploymentValidation

    • BeforeShutdown

  • Bean discovery events, that are fired multiple times:

    • ProcessAnnotatedType

    • ProcessInjectionPoint

    • ProcessInjectionTarget

    • ProcessBeanAttributes

    • ProcessBean

    • ProcessProducer

    • ProcessObserverMethod

Note that the chronological order of these events is specified in Application initialization lifecycle.

11.5.1. BeforeBeanDiscovery event

The container must fire an event before it begins the type discovery process. The event object must be of type javax.enterprise.inject.spi.BeforeBeanDiscovery:

public interface BeforeBeanDiscovery {
    public void addQualifier(Class<? extends Annotation> qualifier);
    public void addQualifier(AnnotatedType<? extends Annotation> qualifier);
    public void addScope(Class<? extends Annotation> scopeType, boolean normal, boolean passivating);
    public void addStereotype(Class<? extends Annotation> stereotype, Annotation... stereotypeDef);
    public void addInterceptorBinding(Class<? extends Annotation> bindingType, Annotation... bindingTypeDef);
    public void addInterceptorBinding(AnnotatedType<? extends Annotation> bindingType);
    public void addAnnotatedType(AnnotatedType<?> type);
    public void addAnnotatedType(AnnotatedType<?> type, String id);
}
  • addQualifier() declares an annotation type as a qualifier type.

  • addScope() declares an annotation type as a scope type.

  • addStereotype() declares an annotation type as a stereotype, and specifies its meta-annotations.

  • addInterceptorBinding() declares an annotation type as an interceptor binding type, and specifies its meta-annotations.

  • addAnnotatedType() adds a given AnnotatedType to the set of types which will be scanned during bean discovery, with an optional identifier. The first version of the method is deprecated since version 1.1 of Contexts and Dependency Injection.

void beforeBeanDiscovery(@Observes BeforeBeanDiscovery event) { ... }

If any observer method of the BeforeBeanDiscovery event throws an exception, the exception is treated as a definition error by the container.

If any BeforeBeanDiscovery method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.2. AfterTypeDiscovery event

The container must fire an event when it has fully completed the type discovery process and before it begins the bean discovery process. The event object must be of type javax.enterprise.inject.spi.AfterTypeDiscovery.

public interface AfterTypeDiscovery {
    public List<Class<?>> getAlternatives();
    public List<Class<?>> getInterceptors();
    public List<Class<?>> getDecorators();
    public void addAnnotatedType(AnnotatedType<?> type, String id);
}
  • getAlternatives() returns the ordered list of enabled alternatives for the application, sorted by priority in ascending order. Alternatives enabled for a bean archive are not included in the list.

  • getInterceptors() returns the ordered list of enabled interceptors for the application, sorted by priority in ascending order. Interceptors enabled for a bean archive are not included in the list.

  • getDecorators() returns the ordered list of enabled decorators for the application, sorted by priority in ascending order. Decorators enabled for a bean archive are not included in the list.

  • addAnnotatedType() adds a given AnnotatedType to the set of types which will be scanned during bean discovery, with an identifier.

If an alternative, interceptor or decorator is added using AfterTypeDiscovery.addAnnotatedType(), non-portable behavior results.

Any observer of this event is permitted to add classes to, or remove classes from, the list of alternatives, list of interceptors or list of decorators. The container must use the final values of these collections, after all observers of AfterTypeDiscovery have been called, to determine the order of the enabled alternatives, interceptors, and decorators for application. The initial values of these collections are defined by the @Priority annotation.

void afterTypeDiscovery(@Observes AfterTypeDiscovery event) { ... }

If any observer method of a AfterTypeDiscovery event throws an exception, the exception is treated as a definition error by the container.

If any AfterTypeDiscovery method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.3. AfterBeanDiscovery event

The container must fire an event when it has fully completed the bean discovery process, validated that there are no definition errors relating to the discovered beans, and registered Bean and ObserverMethod objects for the discovered beans.

The event object must be of type javax.enterprise.inject.spi.AfterBeanDiscovery:

public interface AfterBeanDiscovery {
    public void addDefinitionError(Throwable t);
    public void addBean(Bean<?> bean);
    public void addObserverMethod(ObserverMethod<?> observerMethod);
    public void addContext(Context context);
    public <T> AnnotatedType<T> getAnnotatedType(Class<T> type, String id);
    public <T> Iterable<AnnotatedType<T>> getAnnotatedTypes(Class<T> type);
}
  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after all observers have been notified.

  • addBean() fires an event of type ProcessBean containing the given Bean and then registers the Bean with the container, thereby making it available for injection into other beans. The given Bean may implement Interceptor or Decorator.

  • addObserverMethod() fires an event of type ProcessObserverMethod containing the given ObserverMethod and then registers the ObserverMethod with the container, thereby making it available for event notifications.

  • addContext() registers a custom Context object with the container.

  • getAnnotatedType() and getAnnotatedTypes() returns the AnnotatedType s discovered or added during container initialization. The id of an AnnotatedType added by the container is not defined. If the id passed is null, the container should substitute the container generated id.

A portable extension may take advantage of this event to register beans, interceptors, decorators, observer methods and custom context objects with the container.

void afterBeanDiscovery(@Observes AfterBeanDiscovery event, BeanManager manager) { ... }

If any observer method of the AfterBeanDiscovery event throws an exception, the exception is treated as a definition error by the container.

If any AfterBeanDiscovery method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.4. AfterDeploymentValidation event

The container must fire an event after it has validated that there are no deployment problems and before creating contexts or processing requests.

The event object must be of type javax.enterprise.inject.spi.AfterDeploymentValidation:

public interface AfterDeploymentValidation {
    public void addDeploymentProblem(Throwable t);
}
  • addDeploymentProblem() registers a deployment problem with the container, causing the container to abort deployment after all observers have been notified.

void afterDeploymentValidation(@Observes AfterDeploymentValidation event, BeanManager manager) { ... }

If any observer method of the AfterDeploymentValidation event throws an exception, the exception is treated as a deployment problem by the container.

If any AfterDeploymentValidation method is called outside of the observer method invocation, an IllegalStateException is thrown.

The container must not allow any request to be processed by the deployment until all observers of this event return.

11.5.5. BeforeShutdown event

The container must fire a final event after it has finished processing requests and destroyed all contexts.

The event object must be of type javax.enterprise.inject.spi.BeforeShutdown:

public interface BeforeShutdown {}
void beforeShutdown(@Observes BeforeShutdown event, BeanManager manager) { ... }

If any observer method of the BeforeShutdown event throws an exception, the exception is ignored by the container.

11.5.6. ProcessAnnotatedType event

The container must fire an event, before it processes a type, for every Java class, interface (excluding the special kind of interface declaration annotation type) or enum discovered

as defined in Type discovery

An event is not fired for any type annotated with @Vetoed, or in a package annotated with @Vetoed.

The event object must be of type javax.enterprise.inject.spi.ProcessAnnotatedType<X>, where X is the class, for types discovered in a bean archive, or of type javax.enterprise.inject.spi.ProcessSyntheticAnnotatedType<X> for types added by BeforeBeanDiscovery.addAnnotatedType() or AfterTypeDiscovery.addAnnotatedType().

The annotation @WithAnnotations may be applied to the event parameter. If the annotation is applied, the container must only deliver ProcessAnnotatedType events for types which contain at least one of the annotations specified. The annotation can appear on the annotated type, or on any member, or any parameter of any member of the annotated type, as defined in Alternative metadata sources. The annotation may be applied as a meta-annotation on any annotation considered.

If the @WithAnnotations annotation is applied to any other event parameter, the container automatically detects the problem and treats it as a definition error.

public interface ProcessAnnotatedType<X> {
    public AnnotatedType<X> getAnnotatedType();
    public void setAnnotatedType(AnnotatedType<X> type);
    public void veto();
}
interface ProcessSyntheticAnnotatedType<X> extends ProcessAnnotatedType<X> {
    public Extension getSource();
}
  • getAnnotatedType() returns the AnnotatedType object that will be used by the container to read the declared annotations.

  • setAnnotatedType() replaces the AnnotatedType.

  • veto() forces the container to ignore the type.

  • getSource() returns the Extension instance that added the annotated type.

Any observer of this event is permitted to wrap and/or replace the AnnotatedType. The container must use the final value of this property, after all observers have been called, as the only source of types and annotations for the program elements.

For example, the following observer decorates the AnnotatedType for every class that is discovered by the container.

<T> void decorateAnnotatedType(@Observes ProcessAnnotatedType<T> pat) {
    pat.setAnnotatedType( decorate( pat.getAnnotatedType() ) );
}

If any observer method of a ProcessAnnotatedType event throws an exception, the exception is treated as a definition error by the container.

If any ProcessAnnotatedType method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.7. ProcessInjectionPoint event

The container must fire an event for every injection point of every bean, interceptor or decorator.

The event object must be of type javax.enterprise.inject.spi.ProcessInjectionPoint<T, X> where T is the bean class, and X is the declared type of the injection point.

public interface ProcessInjectionPoint<T, X> {
    public InjectionPoint getInjectionPoint();
    public void setInjectionPoint(InjectionPoint injectionPoint);
    public void addDefinitionError(Throwable t);
}
  • getInjectionPoint() returns the InjectionPoint object that will be used by the container to perform injection.

  • setInjectionPoint() replaces the InjectionPoint.

  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after bean discovery is complete.

Any observer of this event is permitted to wrap and/or replace the InjectionPoint. The container must use the final value of this property, after all observers have been called, whenever it performs injection upon the injection point.

If any observer method of a ProcessInjectionPoint event throws an exception, the exception is treated as a definition error by the container.

If any ProcessInjectionPoint method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.8. ProcessInjectionTarget event

The container must fire an event for every bean, interceptor or decorator.

The event object must be of type javax.enterprise.inject.spi.ProcessInjectionTarget<X>, where X is the bean class.

public interface ProcessInjectionTarget<X> {
    public AnnotatedType<X> getAnnotatedType();
    public InjectionTarget<X> getInjectionTarget();
    public void setInjectionTarget(InjectionTarget<X> injectionTarget);
    public void addDefinitionError(Throwable t);
}
  • getAnnotatedType() returns the AnnotatedType representing the bean class.

  • getInjectionTarget() returns the InjectionTarget object that will be used by the container to perform injection.

  • setInjectionTarget() replaces the InjectionTarget.

  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after bean discovery is complete.

Any observer of this event is permitted to wrap and/or replace the InjectionTarget. The container must use the final value of this property, after all observers have been called, whenever it performs injection upon the managed bean.

If any observer method of a ProcessInjectionTarget event throws an exception, the exception is treated as a definition error by the container.

If any ProcessInjectionTarget method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.9. ProcessBeanAttributes event

The container must fire an event for each managed bean, producer, interceptor or decorator deployed in a bean archive, before registering the Bean object. No event is fired for any:

  • @New qualified bean, defined in @New qualified beans, or,

  • beans added programmatically using AfterBeanDiscovery.addBean(), or,

  • for any built-in beans.

The event object must be of type javax.enterprise.inject.spi.ProcessBeanAttributes<T> where T is the bean class of the bean, the return type of the producer method, or the type of the producer field.

Resources are considered to be producer fields.

public interface ProcessBeanAttributes<T> {
    public Annotated getAnnotated();
    public BeanAttributes<T> getBeanAttributes();
    public void setBeanAttributes(BeanAttributes<T> beanAttributes);
    public void addDefinitionError(Throwable t);
    public void veto();
}
  • getAnnotated() returns the AnnotatedType representing the bean class, the AnnotatedMethod representing the producer field, or the AnnotatedField representing the producer field.

  • getBeanAttributes() returns the BeanAttributes object that will be used by the container to manage instances of the bean.

  • setBeanAttributes() replaces the BeanAttributes.

  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after bean discovery is complete.

  • veto() forces the container to ignore the bean.

Any observer of this event is permitted to wrap and/or replace the BeanAttributes. The container must use the final value of this property, after all observers have been called, to manage instances of the bean. Changes to BeanAttributes are not propagated to the annotated type from which the bean definition was created.

Any bean which has its bean attributes altered must have it’s definition validated during deployment validation.

If any observer method of a ProcessBeanAttributes event throws an exception, the exception is treated as a definition error by the container.

If any ProcessBeanAttributes method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.10. ProcessBean event

The container must fire an event for each bean, interceptor or decorator deployed in a bean archive, after firing the ProcessBeanAttributes for the bean and before registering the Bean object. No event is fired for any @New qualified bean, defined in @New qualified beans.

The event object type in the package javax.enterprise.inject.spi depends upon what kind of bean was discovered:

  • For a managed bean with bean class X, the container must raise an event of type ProcessManagedBean<X>.

  • For a producer method with method return type T of a bean with bean class X, the container must raise an event of type ProcessProducerMethod<T, X>.

  • For a producer field with field type T of a bean with bean class X, the container must raise an event of type ProcessProducerField<T, X>.

The interface javax.enterprise.inject.spi.ProcessBean is a supertype of all these event types:

public interface ProcessBean<X> {
    public Annotated getAnnotated();
    public Bean<X> getBean();
    public void addDefinitionError(Throwable t);
}
  • getAnnotated() returns the AnnotatedType representing the bean class, the AnnotatedMethod representing the producer method, or the AnnotatedField representing the producer field.

  • getBean() returns the Bean object that is about to be registered. The Bean may implement Interceptor or Decorator.

  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after bean discovery is complete.

public interface ProcessManagedBean<X>
        extends ProcessBean<X> {
    public AnnotatedType<X> getAnnotatedBeanClass();
}
public interface ProcessProducerMethod<T, X>
        extends ProcessBean<X> {
    public AnnotatedMethod<T> getAnnotatedProducerMethod();
    public AnnotatedParameter<T> getAnnotatedDisposedParameter();
}
public interface ProcessProducerField<T, X>
        extends ProcessBean<X> {
    public AnnotatedField<T> getAnnotatedProducerField();
    public AnnotatedParameter<T> getAnnotatedDisposedParameter();
}

If any observer method of a ProcessBean event throws an exception, the exception is treated as a definition error by the container.

If any ProcessBean method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.11. ProcessProducer event

The container must fire an event for each producer method or field of each bean, including resources.

The event object must be of type javax.enterprise.inject.spi.ProcessProducer<T, X>, where T is the bean class of the bean that declares the producer method or field and X is the return type of the producer method or the type of the producer field.

public interface ProcessProducer<T, X> {
    public AnnotatedMember<T> getAnnotatedMember();
    public Producer<X> getProducer();
    public void setProducer(Producer<X> producer);
    public void addDefinitionError(Throwable t);
}
  • getAnnotatedMember() returns the AnnotatedField representing the producer field or the AnnotatedMethod representing the producer method.

  • getProducer() returns the Producer object that will be used by the container to call the producer method or read the producer field.

  • setProducer() replaces the Producer.

  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after bean discovery is complete.

Any observer of this event is permitted to wrap and/or replace the Producer. The container must use the final value of this property, after all observers have been called, whenever it calls the producer or disposer.

For example, this observer decorates the Producer for all producer methods and fields of type EntityManager.

void decorateEntityManager(@Observes ProcessProducer<?, EntityManager> pp) {
    pit.setProducer( decorate( pp.getProducer() ) );
}

If any observer method of a ProcessProducer event throws an exception, the exception is treated as a definition error by the container.

If any ProcessProducer method is called outside of the observer method invocation, an IllegalStateException is thrown.

11.5.12. ProcessObserverMethod event

The container must fire an event for each observer method of each bean, before registering the ObserverMethod object.

The event object must be of type javax.enterprise.inject.spi.ProcessObserverMethod<T, X>, where T is the observed event type of the observer method and X is the bean class of the bean that declares the observer method.

public interface ProcessObserverMethod<T, X> {
    public AnnotatedMethod<X> getAnnotatedMethod();
    public ObserverMethod<T> getObserverMethod();
    public void addDefinitionError(Throwable t);
}
  • getAnnotatedMethod() returns the AnnotatedMethod representing the observer method.

  • getObserverMethod() returns the ObserverMethod object that will be used by the container to call the observer method.

  • addDefinitionError() registers a definition error with the container, causing the container to abort deployment after bean discovery is complete.

If any observer method of a ProcessObserverMethod event throws an exception, the exception is treated as a definition error by the container.

If any ProcessObserverMethod method is called outside of the observer method invocation, an IllegalStateException is thrown.

12. Packaging and deployment

When an application is started, the container must perform bean discovery, detect definition errors and deployment problems and raise events that allow portable extensions to integrate with the deployment lifecycle.

Bean discovery is the process of determining:

  • The bean archives that exist in the application, and the beans they contain

  • Which alternatives, interceptors and decorators are enabled for each bean archive

  • The ordering of enabled interceptors and decorators

Additional beans may be registered programmatically with the container by the application or a portable extension after the automatic bean discovery completes. Portable extensions may even integrate with the process of building the Bean object for a bean, to enhance the container’s built-in functionality.

12.1. Bean archives

Bean classes of enabled beans must be deployed in bean archives.

A bean archive has a bean discovery mode of all, annotated or none. A bean archive which contains a beans.xml file with no version has a default bean discovery mode of all. A bean archive which contains a beans.xml file with version 1.1 (or later) must specify the bean-discovey-mode attribute. The default value for the attribute is annotated.

An archive which:

  • contains a beans.xml file with the bean-discovery-mode of none, or,

  • contains an extension and no beans.xml file

is not a bean archive.

An explicit bean archive is an archive which contains a beans.xml file:

  • with a version number of 1.1 (or later), with the bean-discovery-mode of all, or,

  • with no version number, or,

  • that is an empty file.

An implicit bean archive is any other archive which contains one or more bean classes with a bean defining annotation as defined in Bean defining annotations.

When determining which archives are bean archives, the container must consider:

  • Library jars

  • Directories in the JVM classpath

Non Java EE containers may or may not provide support for war, EJB jar or rar bean archives.

The beans.xml file must be named:

  • META-INF/beans.xml.

For compatibility with Contexts and Dependency 1.0, products must contain an option to cause an archive to be ignored by the container when no beans.xml is present.

The container searches for beans in all bean archives in the application classpath.

If a bean class is deployed in two different bean archives, non-portable behavior results. Portable applications must deploy each bean class in no more than one bean archive.

Explicit bean archives may contain classes which are not deployed as beans. For example a bean archive might contain an abstract class not annotated with @Decorator.

Implicit bean archives are likely to contain classes which are not deployed as beans.

An extension may be deployed in any archive, including those that are not bean archives.

12.2. Application initialization lifecycle

When an application is started, the container performs the following steps:

  • First, the container must search for service providers for the service javax.enterprise.inject.spi.Extension defined in Container lifecycle events, instantiate a single instance of each service provider, and search the service provider class for observer methods of initialization events.

  • Next, the container must fire an event of type BeforeBeanDiscovery, as defined in BeforeBeanDiscovery event.

  • Next, the container must perform type discovery, as defined in Type discovery.

  • Next, the container must fire an event of type AfterTypeDiscovery, as defined in AfterTypeDiscovery event.

  • Next, the container must perform bean discovery, as defined in Bean discovery.

  • Next, the container must fire an event of type AfterBeanDiscovery, as defined in AfterBeanDiscovery event, and abort initialization of the application if any observer registers a definition error.

  • Next, the container must detect deployment problems by validating bean dependencies and specialization and abort initialization of the application if any deployment problems exist, as defined in Problems detected automatically by the container.

  • Next, the container must fire an event of type AfterDeploymentValidation, as defined in AfterDeploymentValidation event, and abort initialization of the application if any observer registers a deployment problem.

  • Finally, the container begins directing requests to the application.

12.3. Application shutdown lifecycle

When an application is stopped, the container performs the following steps:

  • First, the container must destroy all contexts.

  • Finally, the container must fire an event of type BeforeShutdown, as defined in BeforeShutdown event.

12.4. Type and Bean discovery

The container automatically discovers managed beans (according to the rules of Which Java classes are managed beans?) in bean archives and searches the bean classes for producer methods, producer fields, disposer methods and observer methods.

12.4.1. Type discovery

First the container must discover types. The container discovers:

  • each Java class, interface (excluding the special kind of interface declaration annotation type) or enum deployed in an explicit bean archive, and

  • each Java class with a bean defining annotation in an implicit bean archive.

that is not excluded from discovery by an exclude filter as defined in Exclude filters.

Then, for every type discovered the container must create an AnnotatedType representing the type and fire an event of type ProcessAnnotatedType, as defined in ProcessAnnotatedType event.

If an extension calls BeforeBeanDiscovery.addAnnotatedType() or AfterTypeDiscovery.addAnnotatedType(), the type passed must be added to the set of discovered types and the container must fire an event of type ProcessSyntheticAnnotatedType for every type added, as defined in ProcessAnnotatedType event+

12.4.2. Exclude filters

Exclude filters are defined by <exclude> elements in the beans.xml for the bean archive as children of the <scan> element. By default an exclude filter is active. If the exclude filter definition contains:

  • a child element named <if-class-available> with a name attribute, and the classloader for the bean archive can not load a class for that name, or

  • a child element named <if-class-not-available> with a name attribute, and the classloader for the bean archive can load a class for that name, or

  • a child element named <if-system-property> with a name attribute, and there is no system property defined for that name, or

  • a child element named <if-system-property> with a name attribute and a value attribute, and there is no system property defined for that name with that value.

then the filter is inactive.

If the filter is active, and:

  • the fully qualified name of the type being discovered matches the value of the name attribute of the exclude filter, or

  • the package name of the type being discovered matches the value of the name attribute with a suffix ".*" of the exclude filter, or

  • the package name of the type being discovered starts with the value of the name attribute with a suffix ".**" of the exclude filter

then we say that the type is excluded from discovery.

For example, consider the follow beans.xml file:

<?xml version="1.0" encoding="UTF-8"?>
<beans xmlns="http://xmlns.jcp.org/xml/ns/javaee">

    <scan>
        <exclude name="com.acme.rest.*" />

        <exclude name="com.acme.faces.**">
            <if-class-not-available name="javax.faces.context.FacesContext"/>
        </exclude>

        <exclude name="com.acme.verbose.*">
            <if-system-property name="verbosity" value="low"/>
        </exclude>

        <exclude name="com.acme.ejb.**">
            <if-class-available name="javax.enterprise.inject.Model"/>
            <if-system-property name="exclude-ejbs"/>
        </exclude>
    </scan>

</beans>

The first exclude filter will exclude all classes in com.acme.rest package. The second exclude filter will exclude all classes in the com.acme.faces package, and any subpackages, but only if JSF is not available. The third exclude filter will exclude all classes in the com.acme.verbose package if the system property verbosity has the value low. The fourth exclude filter will exclude all classes in the com.acme.ejb package, and any subpackages if the system property exclude-ejbs is set (with any value) and at the same time, the javax.enterprise.inject.Model class is available to the classloader.

12.4.3. Bean discovery

For every type in the set of discovered types (as defined in Type discovery), the container must:

  • inspect the type metadata to determine if it is a bean, and then

  • detect definition errors by validating the class and its metadata, and then

  • if the class is a managed bean, fire an event of type ProcessInjectionPoint for each injection point in the class, as defined in ProcessInjectionPoint event, and then

  • if the class is a managed bean, fire an event of type ProcessInjectionTarget, as defined in ProcessInjectionTarget event, and then

  • determine which alternatives, interceptors and decorators are enabled, according to the rules defined in Enabled and disabled beans, Interceptor enablement and ordering and Decorator enablement and ordering, and then

  • if the class is an enabled bean, interceptor or decorator, fire an event of type ProcessBeanAttributes, as defined in ProcessBeanAttributes event, and then

  • if the class is an enabled bean, interceptor or decorator and if ProcessBeanAttributes.veto() wasn’t called in previous step, fire an event which is a subtype of ProcessBean, as defined in ProcessBean event.

For each enabled bean, the container must search the class for producer methods and fields, as defined in Producer methods and in Producer fields, including resources, and for each producer:

  • if it is a producer method, fire an event of type ProcessInjectionPoint for each injection point in the method parameters, as defined in ProcessInjectionPoint event, and then

  • fire an event of type ProcessProducer, as defined in ProcessProducer event, and then

  • if the producer method or field is enabled, fire an event of type ProcessBeanAttributes, as defined in ProcessBeanAttributes event, and then

  • if the producer method or field is enabled and if ProcessBeanAttributes.veto() wasn’t called in previous step, fire an event which is a subtype of ProcessBean, as defined in ProcessBean event.

For each enabled bean, the container must search for disposer methods as defined in Disposer methods, and for each disposer method:

  • fire an event of type ProcessInjectionPoint for each injection point in the method parameters, as defined in ProcessInjectionPoint event.

For each enabled bean, the container must search the class for observer methods, and for each observer method:

Then, the container registers the Bean and ObserverMethod objects:

  • For each enabled bean that is not an interceptor or decorator, the container registers an instance of the Bean interface defined in The Bean interface.

  • For each enabled interceptor, the container registers an instance of the Interceptor interface defined in The Interceptor interface.

  • For each enabled decorator, the container registers an instance of the Decorator interface defined in The Decorator interface.

  • For each observer method of every enabled bean, the container registers an instance of the ObserverMethod interface defined in The ObserverMethod interface.

Part II - CDI in Java SE

This part of the document, specifies additional rules or features when using CDI in a Java SE. All content defined in Part I - Core CDI apply in this part.

13. Bootstrapping a CDI container in Java SE

In Java SE, the CDI container must be explicitly bootstrapped by the application. This is performed with CDIProvider and CDI utility classes.

13.1. CDIProvider lookup

The CDI utility class provides static lookup for CDIProvider. By default, CDIProvider will be looked up via a ServiceLoader but this can be over-written by calling setCDIProvider on the CDI utility class. Once a CDIProvider has been located, it should not be instantiated again, instead the same one returned again.

13.2. Starting the container

Container can be initialized from a CDIProvider instance. Calling one of the initialize methods will start a container and return a CDI utility object to interact with it.

Each time initialize method is called on the CDIProvider it returns a new CDI instance. The application context is started automatically by the container on start up. An implementation need not support calling initialize multiple times on the same provider without the prior container being shutdown.

13.3. Stopping the container

The CDI can be stopped when it’s not needed. To help with this, CDI implements AutoCloseable, so when dereferenced it should shutdown automatically (the close method invokes shutdown).

The CDI instance may also be stopped manually by calling the shutdown method on the CDI object. If the shutdown method is explicitly invoked and the container was not initialized or already stopped , an IllegalStateException will be thrown. The close method should not throw an IllegalStateException if the container was never initialized.

The following code example are equivalent:

public static void main(String... args) {
    CDIProvider provider = CDI.getCDIProvider();
    CDI<Object> cdi = provider.initialize();
    // retrieve a bean and do work with it
    MyBean myBean = cdi.select(MyBean.class).get();
    myBean.doWork();
    // when done
    cdi.shutdown();
}
public static void main(String... args) {
    try(CDI<Object> cdi = CDI.getCDIProvider().initialize()) {
        // start the container, retrieve a bean and do work with it
        MyBean myBean = cdi.select(MyBean.class).get();
        myBean.doWork();
    }
    // shuts down automatically after the try with resources block.
}

14. Scopes and contexts in Java SE

14.1. Context management for built-in scopes in Java SE

When running in Java SE, the container must extend the rules defined in Context management for built-in scopes and is also required to ensure the following rules for built-in context implementation.

14.1.1. Application context lifecycle in Java SE

When running in Java SE the container must extend the rules defined in Application context lifecycle and is also required to ensure the following rules.

The application scope is active:

  • during any method invocation

The application context is shared between all method invocations that execute within the same container.

The application context is destroyed when the container is shut down.

The payload of the event fired when the application context is initialized or destroyed is:

  • any java.lang.Object.

15. Packaging and deployment in Java SE

15.1. Bean archive in Java SE

When running in Java SE, the container must extend the rules defined in Bean archives and also ensure that :

An archive which doesn’t contain a beans.xml file can’t be discovered as an implicit bean archive unless:

  • the application is launched with system property javax.enterprise.inject.scan.implicit set to true, or

  • the container was initialized with a parameter map containing an entry with javax.enterprise.inject.scan.implicit as key and Boolean.TRUE as value.

Part III - CDI in Java EE

This part of the document, specifies additional rules or features when using CDI in a Java EE container. All content defined in Part I - Core CDI apply in this part.

16. Extended Concepts for Java EE

When running in Java EE, the container must extend the concepts defined in Concepts with:

  • A Java EE component is a bean if the lifecycle of its instances may be managed by the container according to the lifecycle context model defined in Scopes and contexts

  • Contextual instances of a bean may be used in EL expressions that are evaluated in the same context

  • For some Java EE components - like environment resources, defined in Resources - the developer provides only the annotations and the bean implementation is provided by the container.

16.1. Functionality provided by the container to the bean in Java EE

When running in Java EE, the container must extend the capabilities defined in Functionality provided by the container to the bean, by providing:

  • scoped resolution by bean name when used in a Unified EL expression, as defined by Name resolution.

16.2. Bean types for Java EE component

As managed beans, EJB session beans may have multiple bean types depending on their client-visible types. For instance, this session bean has only the local interfaces BookShop and Auditable, along with Object, as bean types, since the bean class is not a client-visible type.

@Stateful
public class BookShopBean
        extends Business
        implements BookShop, Auditable {
    ...
}

The rules for determining the (unrestricted) set of bean types for Java EE components are defined in Bean types of a session bean and Bean types of a resource.

16.3. Scopes

Java EE components such as servlets, EJBs and JavaBeans do not have a well-defined scope. These components are either:

  • singletons, such as EJB singleton session beans, whose state is shared between all clients,

  • stateless objects, such as servlets and stateless session beans, which do not contain client-visible state, or

  • objects that must be explicitly created and destroyed by their client, such as JavaBeans and stateful session beans, whose state is shared by explicit reference passing between clients.

CDI scopes add to Java EE these missing well-defined lifecycle context as defined in Scopes.

16.3.1. Built-in scope types in Java EE

When running in Java EE, the implementations of the @RequestScoped, @ApplicationScoped and @SessionScoped annotations provided by the container, represent the standard scopes defined by the Java Servlets specification.

16.4. Default bean discovery mode for Java EE

When running in Java EE, If the bean discovery mode is annotated, the container must extend the rules defined in Default bean discovery mode with:

  • bean classes of EJB sessions beans, are discovered, and

  • producer methods that are on an EJB session bean are discovered, and

  • producer fields that are on an EJB session bean are discovered, and

  • disposer methods that are on an EJB session bean are discovered, and

  • observer methods that are on an EJB session bean are discovered.

16.5. Bean names in Java EE

A bean with a name may be referred to by its name in Unified EL expressions.

There is no relationship between the bean name of an EJB session bean and the EJB name of the bean.

Bean names allow the direct use of beans in JSP or JSF pages, as defined in Integration with Unified EL. For example, a bean with the name products could be used like this:

<h:outputText value="#{products.total}"/>

16.5.1. Default bean names for EJB session beans

In the circumstances listed in Default bean names, the rule for determining default name for an EJB session bean are defined in Default bean name for a session bean.

17. Addition to programming model for Java EE

When running in Java EE, the container must extend the rules defined in Programming model, and must also provide built-in support for injection and contextual lifecycle management of the following kinds of bean:

  • Session beans

  • Resources (Java EE resources, persistence contexts, persistence units, remote EJBs and web services)

Java EE and embeddable EJB containers are required by the Java EE and EJB specifications to support EJB session beans and the Java EE component environment. Other containers are not required to provide support for injection or lifecycle management of session beans or resources.

17.1. Managed beans in Java EE

17.1.1. Which Java classes are managed beans in Java EE?

When running in Java EE, A top-level Java class is a managed bean if it meets requirements described in Which Java classes are managed beans? or if it is defined to be a managed bean by any other Java EE specification and if

  • It is not annotated with an EJB component-defining annotation or declared as an EJB bean class in ejb-jar.xml.

17.2. EJB Session beans

A session bean is a bean that is implemented by a session bean with an EJB 3.x client view that is not annotated with @Vetoed or in a package annotated @Vetoed. The basic lifecycle and semantics of EJB session beans are defined by the EJB specification.

A stateless session bean must belong to the @Dependent pseudo-scope. A singleton session bean must belong to either the @ApplicationScoped scope or to the @Dependent pseudo-scope. If a session bean specifies an illegal scope, the container automatically detects the problem and treats it as a definition error. A stateful session bean may have any scope.

When a contextual instance of a session bean is obtained via the dependency injection service, the behavior of SessionContext.getInvokedBusinessInterface() is specific to the container implementation. Portable applications should not rely upon the value returned by this method.

If the bean class of a session bean is annotated @Interceptor or @Decorator, the container automatically detects the problem and treats it as a definition error.

If the session bean class is a generic type, it must have scope @Dependent. If a session bean with a parameterized bean class declares any scope other than @Dependent, the container automatically detects the problem and treats it as a definition error.

If packages annotated @Vetoed are split across classpath entries, non-portable behavior results. An application can prevent packages being split across jars by sealing the package as defined by the Extension Mechanism Architecture.

17.2.1. EJB remove methods of session beans

If a session bean is a stateful session bean:

  • If the scope is @Dependent, the application may call any EJB remove method of a contextual instance of the session bean.

  • Otherwise, the application may not directly call any EJB remove method of any contextual instance of the session bean.

The session bean is not required to have an EJB remove method in order for the container to destroy it.

If the application directly calls an EJB remove method of a contextual instance of a session bean that is a stateful session bean and declares any scope other than @Dependent, an UnsupportedOperationException is thrown.

If the application directly calls an EJB remove method of a contextual instance of a session bean that is a stateful session bean and has scope @Dependent then no parameters are passed to the method by the container. Furthermore, the container ignores the instance instead of destroying it when Contextual.destroy() is called, as defined in Lifecycle of EJB stateful session beans.

17.2.2. Bean types of a session bean

The unrestricted set of bean types for a session bean contains all local interfaces of the bean and their superinterfaces. If the session bean has a no-interface view, the unrestricted set of bean types contains the bean class and all superclasses. In addition, java.lang.Object is a bean type of every session bean.

Remote interfaces are not included in the set of bean types.

17.2.3. Declaring a session bean

A session bean does not require any special annotations apart from the component-defining annotation (or XML declaration) required by the EJB specification. The following EJBs are beans:

@Singleton
class Shop { .. }
@Stateless
class PaymentProcessorImpl implements PaymentProcessor { ... }

A bean class may also specify a scope, bean name, stereotypes and/or qualifiers:

@ConversationScoped @Stateful @Default @Model
public class ShoppingCart { ... }

A session bean class may extend another bean class:

@Stateless
@Named("loginAction")
public class LoginActionImpl implements LoginAction { ... }
@Stateless
@Mock
@Named("loginAction")
public class MockLoginActionImpl extends LoginActionImpl { ... }

17.2.4. Specializing a session bean

If a bean class of a session bean X is annotated @Specializes, then the bean class of X must directly extend the bean class of another session bean Y. Then X directly specializes Y, as defined in Specialization.

If the bean class of X does not directly extend the bean class of another session bean, the container automatically detects the problem and treats it as a definition error.

For example, MockLoginActionBean directly specializes LoginActionBean:

@Stateless
public class LoginActionBean implements LoginAction { ... }
@Stateless @Mock @Specializes
public class MockLoginActionBean extends LoginActionBean implements LoginAction { ... }

17.2.5. Default bean name for a session bean

The default name for a session bean is the unqualified class name of the session bean class, after converting the first character to lower case.

For example, if the bean class is named ProductList, the default bean name is productList.

17.3. Producer methods on EJB session bean

A producer method defined in an EJB session bean follows the rules defined in <<producer_method> with the following addition:

  • A producer method defined in an EJB session bean must be either a business method of the EJB or a static method of the bean class.

17.3.1. Declaring a producer method in an EJB session bean

A producer method declaration in an EJB session bean follows the rules defined in Declaring a producer method with the following addition:

  • if a non-static method of a session bean class is annotated @Produces, and the method is not a business method of the session bean, the container automatically detects the problem and treats it as a definition error.

17.4. Producer field on EJB session bean

A producer field defined in an EJB session bean follows the rules defined in <<producer_field> with the following addition:

  • A producer field defined in an EJB session bean must be a static field of the bean class.

17.4.1. Declaring a producer field in an EJB session bean

A producer field declaration in an EJB session bean follows the rules defined in Declaring a producer field with the following addition:

  • If a non-static field of an EJB session bean class is annotated @Produces, the container automatically detects the problem and treats it as a definition error.

17.5. Disposer methods on EJB session bean

A disposer method defined in an EJB session bean follows the rules defined in <<disposer_method> with the following addition:

  • A disposer method defined in an EJB session bean must be either a business method of the EJB or a static method of the bean class.

17.5.1. Declaring a disposer method on an EJB session bean

A disposer method declaration in an EJB session bean follows the rules defined in Declaring a disposer method with the following addition:

  • If a non-static method of an EJB session bean class has a parameter annotated @Disposes, and the method is not a business method of the session bean, the container automatically detects the problem and treats it as a definition error.

17.6. Java EE components

Most Java EE components support injection and interception, as defined in the Java Platform, Enterprise Edition Specification 7, table EE.5-1, but are not considered beans (as defined by this specification). EJBs, as defined in EJB Session beans are the exception.

The instance used by the container to service an invocation of a Java EE component will not be the same instance obtained when using @Inject, instantiated by the container to invoke a producer method, observer method or disposer method, or instantiated by the container to access the value of a producer field. It is recommended that Java EE components should not define observer methods, producer methods, producer fields or disposer methods. It is safe to annotate Java EE components with @Vetoed to prevent them being considered beans.

17.7. Resources

A resource is a bean that represents a reference to a resource, persistence context, persistence unit, remote EJB or web service in the Java EE component environment.

By declaring a resource, we enable an object from the Java EE component environment to be injected by specifying only its type and qualifiers at the injection point. For example, if @CustomerDatabase is a qualifier:

@Inject @CustomerDatabase Datasource customerData;
@Inject @CustomerDatabase EntityManager customerDatabaseEntityManager;
@Inject @CustomerDatabase EntityManagerFactory customerDatabaseEntityManagerFactory;
@Inject PaymentService remotePaymentService;

The container is not required to support resources with scope other than @Dependent. Portable applications should not define resources with any scope other than @Dependent.

A resource may not have a bean name.

17.7.1. Declaring a resource

A resource may be declared by specifying a Java EE component environment injection annotation as part of a producer field declaration. The producer field may be static.

  • For a Java EE resource, @Resource must be specified.

  • For a persistence context, @PersistenceContext must be specified.

  • For a persistence unit, @PersistenceUnit must be specified.

  • For a remote EJB, @EJB must be specified.

  • For a web service, @WebServiceRef must be specified.

The injection annotation specifies the metadata needed to obtain the resource, entity manager, entity manager factory, remote EJB instance or web service reference from the component environment.

@Produces @WebServiceRef(lookup="java:app/service/PaymentService")
PaymentService paymentService;
@Produces @EJB(ejbLink="../their.jar#PaymentService")
PaymentService paymentService;
@Produces @Resource(lookup="java:global/env/jdbc/CustomerDatasource")
@CustomerDatabase Datasource customerDatabase;
@Produces @PersistenceContext(unitName="CustomerDatabase")
@CustomerDatabase EntityManager customerDatabasePersistenceContext;
@Produces @PersistenceUnit(unitName="CustomerDatabase")
@CustomerDatabase EntityManagerFactory customerDatabasePersistenceUnit;

The bean type and qualifiers of the resource are determined by the producer field declaration.

If the producer field declaration specifies a bean name, the container automatically detects the problem and treats it as a definition error.

If the matching object in the Java EE component environment is not of the same type as the producer field declaration, the container automatically detects the problem and treats it as a definition error.

17.7.2. Bean types of a resource

The unrestricted set of bean types of a resource is determined by the declared type of the producer field, as specified by Bean types of a producer field.

17.8. Additional built-in beans

A Java EE or embeddable EJB container must provide the following built-in beans, all of which have qualifier @Default:

  • a bean with bean type javax.transaction.UserTransaction, allowing injection of a reference to the JTA UserTransaction, and

  • a bean with bean type java.security.Principal, allowing injection of a Principal representing the current caller identity.

A servlet container must provide the following built-in beans, all of which have qualifier @Default:

  • a bean with bean type javax.servlet.http.HttpServletRequest, allowing injection of a reference to the HttpServletRequest

  • a bean with bean type javax.servlet.http.HttpSession, allowing injection of a reference to the HttpSession,

  • a bean with bean type javax.servlet.ServletContext, allowing injection of a reference to the ServletContext,

These beans are passivation capable dependencies, as defined in Passivation capable dependencies.

If a Java EE component class has an injection point of type UserTransaction and qualifier @Default, and may not validly make use of the JTA UserTransaction according to the Java EE platform specification, the container automatically detects the problem and treats it as a definition error.

17.9. Injected fields in Java EE

When running in Java EE, the container must extend the rules defined for bean classes in Injected fields to Java EE component classes supporting injection.

17.10. Initializer methods in Java EE

When running in Java EE, the container must extend the rules defined for bean classes in Initializer methods to Java EE component classes supporting injection. The container must also ensure that:

  • An intializer method defined in an EJB session bean is not required to be a business method of the session bean.

17.11. @New qualified beans in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in @New qualified beans to EJB session beans.

17.12. Inheritance of type-level metadata in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Inheritance of type-level metadata to EJB session beans.

17.13. Inheritance of member-level metadata in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Inheritance of member-level metadata to EJB session beans.

17.14. Specialization in Java EE

17.14.1. Direct and indirect specialization in Java EE

When running in Java EE, the container must extend the rules defined in Inheritance of member-level metadata and is also required to support specialization for EJB session beans as defined in Specializing a session bean.

18. Dependency injection, lookup and EL in Java EE

When running in Java EE, the container must extend the rules defined in Dependency injection and lookup and may also provide references to contextual instances by Unified EL expression evaluation.

When resolving a name in an EL expression, the container considers the bean name and selected alternatives.

18.1. Modularity in Java EE

In the Java EE module architecture, any Java EE module or library is a module. The Java EE module is a bean archive if it contains a beans.xml file, as defined in Bean archives.

When running in Java EE, the container must follow the same accessibility rules for beans and alternatives defined in Modularity for JSP/JSF pages using EL resolution and make sure that only beans available from injection in the module that defines the JSP/JSF pages are resolved.

In the Java EE module architecture, a bean class is accessible in a module if and only if it is required to be accessible according to the class loading requirements defined by the Java EE platform specification.

Note that, in some Java EE implementations, a bean class might be accessible to some other class even when this is not required by the Java EE platform specification. For the purposes of this specification, a class is not considered accessible to another class unless accessibility is explicitly required by the Java EE platform specification.

An alternative is not available for injection, lookup or EL resolution to classes or JSP/JSF pages in a module unless the module is a bean archive and the alternative is explicitly selected for the bean archive or the application.

18.1.1. Declaring selected alternatives for an application in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Declaring selected alternatives for an application to EJB session beans.

18.1.2. Declaring selected alternatives for a bean archive in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Declaring selected alternatives for a bean archive to EJB session beans.

18.1.3. Unsatisfied and ambiguous dependencies in Java EE

When running in Java EE, the container must extend the rules defined in Unsatisfied and ambiguous dependencies and must also validate all injection points of all Java EE component classes supporting injection

18.2. EL name resolution

When running in Java EE, the container must extend the rules defined in Name resolution and must also support name resolution for name used in Expression Language

An EL name resolves to a bean if:

  • the name can be resolved to a bean according to rules in Name resolution, and

  • the bean is available for injection in the war containing the JSP or JSF page with the EL expression.

18.2.1. Ambiguous EL names

When running in Java EE, the container must extend the rules defined in Ambiguous names to names used in Expression Language.

18.3. Dependency injection in Java EE

When running in Java EE, the container must extend the rules defined in Dependency injection and is also required to perform dependency injection whenever it creates the following contextual objects:

  • contextual instances of EJB session beans.

The container is also required to perform dependency injection whenever it instantiates any of the following non-contextual objects:

  • non-contextual instances of EJB session beans (for example, session beans obtained by the application from JNDI or injected using @EJB), and

  • instances of any other Java EE component class supporting injection.

A Java EE 5 container is not required to support injection for non-contextual objects.

18.3.1. Injection using the bean constructor in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Injection using the bean constructor to EJB session beans.

18.3.2. Injection of fields and initializer methods in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Injection of fields and initializer methods to EJB session beans and to any other Java EE component class supporting injection.

The container is also required to ensure that:

-* Initializer methods declared by a class X in the type hierarchy of the bean are called after all Java EE component environment resource dependencies declared by X or by superclasses of X have been injected. -* Any @PostConstruct callback declared by a class X in the type hierarchy of the bean is called after all Java EE component environment resource dependencies declared by X or by superclasses of X have been injected. -* Any servlet init() method is called after all initializer methods have been called, all injected fields have been initialized and all Java EE component environment resource dependencies have been injected.

18.3.3. Destruction of dependent objects in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Destruction of dependent objects to any other Java EE component class supporting injection and perform destruction after the servlet destroy() method is called.

19. Scopes and contexts in Java EE

19.1. Dependent pseudo-scope in Java EE

When running in Java EE, the container must extend the rules defined in Dependent pseudo-scope and must also ensure that if a bean is declared to have @Dependent scope:

  • When a Unified EL expression in a JSF or JSP page that refers to the bean by its bean name is evaluated, at most one instance of the bean is instantiated. This instance exists to service just a single evaluation of the EL expression. It is reused if the bean name appears multiple times in the EL expression, but is never reused when the EL expression is evaluated again, or when another EL expression is evaluated.

19.1.1. Dependent objects in Java EE

When running in Java EE, the container must extend the rules defined for bean in Dependent objects to Java EE component class instance.

19.1.2. Destruction of objects with scope @Dependent in Java EE

When running in Java EE, the container must extend the rules defined for bean in Destruction of objects with scope @Dependent to Java EE component class instance, and must also ensure that :

  • all @Dependent scoped contextual instances created during evaluation of a Unified EL expression in a JSP or JSF page are destroyed when the evaluation completes.

19.1.3. Dependent pseudo-scope and Unified EL

Suppose a Unified EL expression in a JSF or JSP page refers to a bean with scope @Dependent by its bean name. Each time the EL expression is evaluated:

  • the bean is instantiated at most once, and

  • the resulting instance is reused for every appearance of the bean name, and

  • the resulting instance is destroyed when the evaluation completes.

Portable extensions that integrate with the container via Unified EL should also ensure that these rules are enforced.

19.2. Passivation and passivating scopes in Java EE

19.2.1. Passivation capable beans in Java EE

  • As defined by the EJB specification, an EJB stateful session beans is passivation capable if:

    • interceptors and decorators of the bean are passivation capable, and,

    • the EJB stateful session bean does not have the passivationCapable flag set to false.

  • As defined by the EJB specification, an EJB stateless session bean or an EJB singleton session bean is not passivation capable.

19.2.2. Passivation capable dependencies in java EE

When running in Java EE, the container must extend the rules defined in Passivation capable dependencies, and must also guarantee that:

  • all EJB stateless session beans are passivation capable dependencies,

  • all EJB singleton session beans are passivation capable dependencies,

  • all passivation capable EJB stateful session beans are passivation capable dependencies, and

  • all Java EE resources are passivation capable dependencies.

19.2.3. Validation of passivation capable beans and dependencies in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Validation of passivation capable beans and dependencies to EJB session beans.

19.3. Context management for built-in scopes in Java EE

When running in Java EE, the container must extend the rules defined in Context management for built-in scopes and is also required to ensure the following rules for built-in context implementation.

The built-in request and application context objects are active during servlet, web service and EJB invocations, and the built in session and request context objects are active during servlet and web service invocations.

19.3.1. Request context lifecycle in Java EE

When running in Java EE the container must extend the rules defined in Request context lifecycle and is also required to implement request context with the following rules.

The request scope is active:

  • during the service() method of any servlet in the web application, during the doFilter() method of any servlet filter and when the container calls any ServletRequestListener or AsyncListener,

  • during any Java EE web service invocation,

  • during any asynchronous invocation of an event observer,

  • during any remote method invocation of any EJB, during any asynchronous method invocation of any EJB, during any call to an EJB timeout method and during message delivery to any EJB message-driven bean, and

  • during @PostConstruct callback of any bean.

The request context is destroyed:

  • at the end of the servlet request, after the service() method, all doFilter() methods, and all requestDestroyed() and onComplete() notifications return,

  • after the web service invocation completes,

  • after the EJB remote method invocation, asynchronous method invocation, timeout or message delivery completes if it did not already exist when the invocation occurred, or

  • after the @PostConstruct callback completes, if it did not already exist when the @PostConstruct callback occurred.

The payload of the event fired when the request context is initialized or destroyed is:

  • the ServletRequest if the context is initialized or destroyed due to a servlet request, or

  • the ServletRequest if the context is initialized or destroyed due to a web service invocation, or

  • any java.lang.Object for other types of request.

19.3.2. Session context lifecycle in Java EE

When running in Java EE the container is required to implement session context with the following rules.

The session scope is active:

  • during the service() method of any servlet in the web application, during the doFilter() method of any servlet filter and when the container calls any HttpSessionListener, AsyncListener or ServletRequestListener.

The session context is shared between all servlet requests that occur in the same HTTP session. The session context is destroyed when the HTTPSession times out, after all HttpSessionListener s have been called, and at the very end of any request in which invalidate() was called, after all filters and ServletRequestListener s have been called.

An event with the HttpSession as payload and with qualifier @Initialized(SessionScoped.class) is fired when the session context is initialized and an event with qualifier @Destroyed(SessionScoped.class) when the session context is destroyed.

19.3.3. Application context lifecycle in Java EE

When running in Java EE the container must extend the rules defined in Application context lifecycle and is also required to implement application context with the following rules.

The application scope is active:

  • during the service() method of any servlet in the web application, during the doFilter() method of any servlet filter and when the container calls any ServletContextListener, HttpSessionListener, AsyncListener or ServletRequestListener,

  • during any Java EE web service invocation,

  • during any asynchronous invocation of an event observer,

  • during any remote method invocation of any EJB, during any asynchronous method invocation of any EJB, during any call to an EJB timeout method and during message delivery to any EJB message-driven bean,

  • when the disposer method or @PreDestroy callback of any bean with any normal scope other than @ApplicationScoped is called, and

  • during @PostConstruct callback of any bean.

The application context is shared between all servlet requests, web service invocations, asynchronous invocation of an event observer, EJB remote method invocations, EJB asynchronous method invocations, EJB timeouts and message deliveries to message-driven beans that execute within the same application. The application context is destroyed when the application is shut down.

The payload of the event fired when the application context is initialized or destroyed is:

  • the ServletContext if the application is a web application deployed to a Servlet container, or

  • any java.lang.Object for other types of application.

19.3.4. Conversation context lifecycle in Java EE

When running in Java EE the container is required to implement conversation context with the following rules.

The conversation scope is active during all Servlet requests.

An event with qualifier @Initialized(ConversationScoped.class) is fired when the conversation context is initialized and an event with qualifier @Destroyed(ConversationScoped.class) is fired when the conversation is destroyed. The event payload is:

  • the conversation id if the conversation context is destroyed and is not associated with a current Servlet request, or

  • the ServletRequest if the application is a web application deployed to a Servlet container, or

  • any java.lang.Object for other types of application.

The conversation context provides access to state associated with a particular conversation. Every Servlet request has an associated conversation. This association is managed automatically by the container according to the following rules:

  • Any Servlet request has exactly one associated conversation.

  • The container provides a filter with the name "CDI Conversation Filter", which may be mapped in web.xml, allowing the user alter when the conversation is associated with the servlet request. If this filter is not mapped in any web.xml in the application, the conversation associated with a Servlet request is determined at the beginning of the request before calling any service() method of any servlet in the web application, calling the doFilter() method of any servlet filter in the web application and before the container calls any ServletRequestListener or AsyncListener in the web application.

  • The implementation should determine the conversation associated with the Servlet request in a way that does not prevent other filters or servlet from setting the request character encoding or parsing the request body themselves.

Any conversation is in one of two states: transient or long-running.

  • By default, a conversation is transient

  • A transient conversation may be marked long-running by calling Conversation.begin()

  • A long-running conversation may be marked transient by calling Conversation.end()

All long-running conversations have a string-valued unique identifier, which may be set by the application when the conversation is marked long-running, or generated by the container.

If the conversation associated with the current Servlet request is in the transient state at the end of a Servlet request, it is destroyed, and the conversation context is also destroyed.

If the conversation associated with the current Servlet request is in the long-running state at the end of a Servlet request, it is not destroyed. The long-running conversation associated with a request may be propagated to any Servlet request via use of a request parameter named cid containing the unique identifier of the conversation. In this case, the application must manage this request parameter.

If the current Servlet request is a JSF request, and the conversation is in long-running state, it is propagated according to the following rules:

  • The long-running conversation context associated with a request that renders a JSF view is automatically propagated to any faces request (JSF form submission) that originates from that rendered page.

  • The long-running conversation context associated with a request that results in a JSF redirect (a redirect resulting from a navigation rule or JSF NavigationHandler) is automatically propagated to the resulting non-faces request, and to any other subsequent request to the same URL. This is accomplished via use of a request parameter named cid containing the unique identifier of the conversation.

When no conversation is propagated to a Servlet request, or if a request parameter named conversationPropagation has the value none the request is associated with a new transient conversation.

All long-running conversations are scoped to a particular HTTP servlet session and may not cross session boundaries.

In the following cases, a propagated long-running conversation cannot be restored and reassociated with the request:

  • When the HTTP servlet session is invalidated, all long-running conversation contexts created during the current session are destroyed, after the servlet service() method completes.

  • The container is permitted to arbitrarily destroy any long-running conversation that is associated with no current Servlet request, in order to conserve resources.

The conversation timeout, which may be specified by calling Conversation.setTimeout() is a hint to the container that a conversation should not be destroyed if it has been active within the last given interval in milliseconds.

If the propagated conversation cannot be restored, the container must associate the request with a new transient conversation and throw an exception of type javax.enterprise.context.NonexistentConversationException.

The container ensures that a long-running conversation may be associated with at most one request at a time, by blocking or rejecting concurrent requests. If the container rejects a request, it must associate the request with a new transient conversation and throw an exception of type javax.enterprise.context.BusyConversationException.

20. Lifecycle of contextual instances

20.1. Container invocations and interception in Java EE

When the application invokes:

  • a business method of a session bean via an EJB remote or local reference,

the invocation is treated as a business method invocation.

When running in Java EE, the container must extend the rules defined in Container invocations and interception, with:

  • Invocation of EJB timer service timeouts by the container are not business method invocations, but are intercepted by interceptors for EJB timeouts.

  • Only an invocation of business method on an EJB session bean is subject to EJB services such as declarative transaction management, concurrency, security and asynchronicity, as defined by the EJB specification.

  • Additionally, invocations of message listener methods of message-driven beans during message delivery are passed through method interceptors.

20.1.1. Lifecycle of EJB stateful session beans

When the create() method of a Bean object that represents an EJB stateful session bean that is called, the container creates and returns a container-specific internal local reference to a new EJB session bean instance. The reference must be passivation capable. This reference is not directly exposed to the application. When the create() method of a Bean object that represents an EJB stateful session bean that is called, the container creates and returns a container-specific internal local reference to a new EJB session bean instance. The reference must be passivation capable. This reference is not directly exposed to the application.

Before injecting or returning a contextual instance to the application, the container transforms its internal reference into an object that implements the bean types expected by the application and delegates method invocations to the underlying EJB stateful session bean instance. This object must be passivation capable.

When the destroy() method is called, and if the underlying EJB was not already removed by direct invocation of a remove method by the application, the container removes the EJB stateful session bean. The @PreDestroy callback must be invoked by the container.

Note that the container performs additional work when the underlying EJB is created and removed, as defined in Dependency injection

20.1.2. Lifecycle of EJB stateless and singleton session beans

When the create() method of a Bean object that represents an EJB stateless session or singleton session bean is called, the container creates and returns a container-specific internal local reference to the EJB session bean. This reference is not directly exposed to the application.

Before injecting or returning a contextual instance to the application, the container transforms its internal reference into an object that implements the bean types expected by the application and delegates method invocations to the underlying EJB session bean. This object must be passivation capable.

When the destroy() method is called, the container simply discards this internal reference.

Note that the container performs additional work when the underlying EJB is created and removed, as defined in Dependency injection

20.1.3. Lifecycle of resources

When the create() method of a Bean object that represents a resource is called, the container creates and returns a container-specific internal reference to the Java EE component environment resource, entity manager, entity manager factory, remote EJB instance or web service reference. This reference is not directly exposed to the application.

Before injecting or returning a contextual instance to the application, the container transforms its internal reference into an object that implements the bean types expected by the application and delegates method invocations to the underlying resource, entity manager, entity manager factory, remote EJB instance or web service reference. This object must be passivation capable.

The container must perform ordinary Java EE component environment injection upon any non-static field that functions as a resource declaration, as defined by the Java EE platform and Common Annotations for the Java platform specifications. The container is not required to perform Java EE component environment injection upon a static field. Portable applications should not rely upon the value of a static field that functions as a resource declaration.

References to EJBs and web services are always dependent scoped and a new instance must be obtained for every injection performed.

For an entity manager associated with a resource definition, it must behave as though it were injected directly using @PersistenceContext.

When the destroy() method of a bean which represents a remote stateful EJB reference is called, the container will not automatically destroy the EJB reference. The application must explicitly call the method annotated @Remove. This behavior differs to that specified in Lifecycle of EJB stateful session beans for beans which represent a local stateful EJB reference

21. Decorators in Java EE

When running in Java EE, the container must extend the rules defined for managed beans in Decorators to EJB session beans.

21.1. Decorator beans in Java EE

Decorators of an EJB session bean must comply with the bean provider programming restrictions defined by the EJB specification. Decorators of an EJB stateful session bean must comply with the rules for instance passivation and conversational state defined by the EJB specification.

22. Interceptor bindings in Java EE

EJB session and message-driven beans support interception as defined in Interceptor bindings.

22.1. Interceptor resolution in Java EE

For a custom implementation of the Interceptor interface defined in The Interceptor interface, the container also calls intercepts() to determine if the interceptor intercepts an EJB timeout method invocation.

23. Events in Java EE

23.1. Observer methods in EJB session beans

An observer method may also be a non-abstract method of an EJB session bean class. It must be either a business method of the EJB or a static method of the bean class.

23.1.1. Declaring an observer method in an EJB

If a non-static method of a session bean class has a parameter annotated @Observes, and the method is not a business method of the EJB, the container automatically detects the problem and treats it as a definition error.

23.1.2. Observer method invocation context in Java EE

The transaction and security contexts for a business method of an EJB session bean also depend upon the transaction attribute and @RunAs descriptor, if any.

24. Portable extensions in Java EE

24.1. The Bean interface in Java EE

When Running in Java EE, the container must extend the rules defined in The Bean interface for managed bean to EJB session bean.

24.1.1. The Interceptor interface in Java EE

When Running in Java EE, the container must extend the rules defined in The Interceptor interface and must also ensure that

PRE_PASSIVATE, POST_ACTIVATE and AROUND_TIMEOUT InterceptorType values are linked to EJB lifecycle callback or timeout method.

24.2. InjectionTarget interface in Java EE

When running in Java EE, the container must extend the rules defined for InjectionTarget in The Producer and InjectionTarget interfaces and must also ensure that:

  • when inject() is called, The container performs Java EE component environment injection, according to the semantics required by the Java EE platform specification, sets the value of all injected fields, and calls all initializer methods, as defined in Injection of fields and initializer methods in Java EE.

  • @PostConstruct callback is called according to the semantics required by the Java EE platform specification.

  • @PreDestroy callback is called according to the semantics required by the Java EE platform specification.

24.3. The BeanManager object in Java EE

24.3.1. Obtaining a reference to the CDI container in Java EE

A Java EE container is required to provide a CDI provider that will allow access to the current container for any Java EE application or Java EE module which contains enabled beans.

Java EE Components may obtain an instance of BeanManager from JNDI by looking up the name java:comp/BeanManager.

When running in Java EE, when initialize, stop or shutdown methods of CDIProvider are called an UnsupportedOperationException is thrown.

24.3.2. Wrapping a Unified EL ExpressionFactory

The method BeanManager.wrapExpressionFactory() returns a wrapper javax.el.ExpressionFactory that delegates MethodExpression and ValueExpression creation to the given ExpressionFactory. When a Unified EL expression is evaluated using a MethodExpression or ValueExpression returned by the wrapper ExpressionFactory, the rules defined in Dependent pseudo-scope and Unified EL are enforced by the container.

public ExpressionFactory wrapExpressionFactory(ExpressionFactory expressionFactory);

24.4. Alternative metadata sources and EJB

When running in Java EE, the container must extend the rules defined in Alternative metadata sources and ensure that:

  • when an AnnotatedType represents an EJB session bean class, Annotated.getTypeClosure() must returns the EJB session bean types as defined in Bean types of a session bean.

24.5. Addition to Container lifecycle events in Java EE

24.5.1. ProcessInjectionPoint event and EJB

When running in Java EE, the container must also fire an event for every injection point of every Java EE component class supporting injection that may be instantiated by the container at runtime, including every managed bean declared using @ManagedBean, EJB session or message-driven bean.

24.5.2. ProcessInjectionTarget event and EJB

When running in Java EE, the container must also fire an event for every Java EE component class supporting injection that may be instantiated by the container at runtime, including every managed bean declared using @ManagedBean, EJB session or message-driven bean.

The container must extend the rules defined in ProcessInjectionTarget event for managed bean to EJB session bean and other Java EE component class supporting injection.

For example, this observer decorates the InjectionTarget for all servlets.

<T extends Servlet> void decorateServlet(@Observes ProcessInjectionTarget<T> pit) {
    pit.setInjectionTarget( decorate( pit.getInjectionTarget() ) );
}

24.5.3. ProcessBeanAttributes event and EJB

When running in Java EE, the container must extend the rules defined in ProcessBeanAttributes event to EJB session bean.

24.5.4. ProcessBean event and EJB

In addition to definition given in ProcessBean event the following apply:

  • For a session bean with bean class X, the container must raise an event of type ProcessSessionBean<X>.

Resources are considered to be producer fields.

When running in Java EE, the interface javax.enterprise.inject.spi.ProcessBean is also a supertype of javax.enterprise.inject.spi.ProcessSession :

public interface ProcessSessionBean<X>
        extends ProcessManagedBean<Object> {
    public String getEjbName();
    public SessionBeanType getSessionBeanType();
}
  • getEjbName() returns the EJB name of the session bean.

  • getSessionBeanType() returns a javax.enterprise.inject.spi.SessionBeanType representing the kind of session bean.

public enum SessionBeanType { STATELESS, STATEFUL, SINGLETON }

25. Packaging and deployment in Java EE

25.1. Bean archive with EJB Session Beans

When running in Java EE, the container must extend the rules defined in Bean archives with:

  • An implicit bean archive may also contain EJB session beans, and

  • EJB session bean should be considered as bean class with bean defining annotation when determining if and archive is an implicit bean archive.

When determining which archives are bean archives, the container must also consider:

  • EJB jars or application client jars

  • The WEB-INF/classes directory of a war

The container is not required to support application client jar bean archives.

A Java EE container is required by the Java EE specification to support Java EE modules.

In a war, the beans.xml file must be named:

  • WEB-INF/beans.xml or WEB-INF/classes/META-INF/beans.xml.

If a war has a file named beans.xml in both the WEB-INF directory and in the WEB-INF/classes/META-INF directory, then non-portable behavior results. Portable applications must have a beans.xml file in only one of the WEB-INF or the WEB-INF/classes/META-INF directories.

The following additional rules apply regarding container search for beans:

  • In an application deployed as an ear, the container searches every bean archive bundled with or referenced by the ear, including bean archives bundled with or referenced by wars, EJB jars and rars contained in the ear. The bean archives might be library jars, EJB jars or war WEB-INF/classes directories.

  • In an application deployed as a war, the container searches every bean archive bundled with or referenced by the war. The bean archives might be library jars or the WEB-INF/classes directory.

  • In an application deployed as an EJB jar, the container searches the EJB jar, if it is a bean archive, and every bean archive referenced by the EJB jar.

  • In an application deployed as a rar, the container searches every bean archive bundled with or referenced by the rar.

  • An embeddable EJB container searches each bean archive in the JVM classpath that is listed in the value of the embeddable container initialization property javax.ejb.embeddable.modules, or every bean archive in the JVM classpath if the property is not specified. The bean archives might be directories, library jars or EJB jars.

25.2. Type and Bean discovery for EJB

In Java EE, the container automatically discovers EJB session beans and other Java EE component class supporting injection, in bean archives like it does for managed bean as defined in Type and Bean discovery.

25.2.1. Bean discovery in Java EE

When running in Java EE, the container must extend the rules defined in Bean discovery and must also discover each EJB session beans.

25.3. Integration with Unified EL

The container must provide a Unified EL ELResolver to the servlet engine and JSF implementation that resolves bean names using the rules of name resolution defined in Name resolution and resolving ambiguities according to Ambiguous names.

  • If a name used in an EL expression does not resolve to any bean, the ELResolver must return a null value.

  • Otherwise, if a name used in an EL expression resolves to exactly one bean, the ELResolver must return a contextual instance of the bean, as defined in Contextual instance of a bean.