Core Module

The Core module provides fundamental and defining DeltaSpike API and utility classes. As such, this module must be included in every project that uses DeltaSpike. Instructions for configuring your projects to use the DeltaSpike Core module are detailed in Configure DeltaSpike in Your Projects as part of the general instructions for configuring your projects for DeltaSpike.

Project Setup

Add Core-API and Core-Impl to your project as detailed in Configure DeltaSpike Core in Your Projects. For Maven-independent projects, see Configure DeltaSpike Core in Maven-independent Projects.

DeltaSpike Configuration

DeltaSpike provides a very flexible application configuration mechanism. Its main goal is to make it possible to never have to rebuild a project just for the sake of adjusting configuration values. It respects the usual software development cycle and takes into account project stages, various configuration sources and resolution mechanisms. Configuration can be overridden just by dropping a JAR into the classpath.

Detailed documentation is available on a separate page: Configuration.

Internal configuration

The functionality of DeltaSpike itself and its modules is adjustable using the same mechanism. There are two main types of internal configuration:

  • static configuration: certain configurable options, like the maximum number of windows in the window scope or the priority for DeltaSpike’s global interceptors, are defined using the DeltaSpike configuration mechanism and can be adjusted by redefining their value in the apache-deltaspike.properties file, as described in the documentation. All of these configuration options and their corresponding properties can be found as members of interfaces extending DeltaSpikeBaseConfig, e.g. CoreBaseConfig which configures the core module.

  • dynamic configuration: certain values can change dynamically during runtime and some may differ even among contexts. For example, much of the behaviour of the JSF module is configured in JsfModuleConfig. To override any of the default configuration options, the JsfModuleConfig bean can be overridden using a custom implementation of JsfModuleConfig which would be annotated @Alternative or @Specializes. All of the dynamic CDI-based configuration beans can be found as implementations of the DeltaSpikeConfig interface.

BeanProvider

The BeanProvider utility class provides static methods for manual lookup of bean instances in places where standard injection is not available or if the lookup depends on dynamic conditions.

BeanProvider is only used to look up normal-scoped contextual instances. To obtain instances of dependent-scoped beans, use DependentProvider.
The term 'contextual instance' is used instead of 'bean' because that’s the term used by CDI itself.

The following example shows a simple lookup. With the second parameter it is possible to specify if the contextual instance is optional. If it is not expected that the contextual instance is optional, but no instance has been found, an IllegalStateException will be thrown.

Resolving a Simple Contextual Instance
MyBean myBean = BeanProvider.getContextualReference(MyBean.class, false);

Pass true as second argument, if you look for an implementation of the given interface and an implementation is not required or it is not required that there is an instance with the given qualifier (see the qualifier example for further details).

Resolving an Optional Contextual Instance
MyServiceInterface optionalService = BeanProvider.getContextualReference(MyServiceInterface.class, true);

Optionally you can provide a qualifier for the contextual instance in question. CDI qualifiers are annotations, therefore you need to implement a corresponding literal for providing an instance.

Literal Implementation for '@MyQualifier'
import javax.enterprise.util.AnnotationLiteral;
//...

public class MyQualifierLiteral extends AnnotationLiteral<MyQualifier> implements MyQualifier
{
}

The following example will return a contextual instance with the qualifier @MyQualifier.

Resolving a Simple Contextual Instance with Qualifier
MyBean myBean = BeanProvider.getContextualReference(MyBean.class, false, new MyQualifierLiteral());

The @Named qualifier has a special role and allows to specify a string based name (e.g. for referencing CDI beans in EL-expressions). However, the following examples show how to do a manual lookup by name.

Resolving a Simple Contextual Instance by Name
Object myBean = BeanProvider.getContextualReference("myBean", false);
Resolving a Simple Contextual Instance by Name and Expected Type
MyBean myBean = BeanProvider.getContextualReference("myBean", false, MyBean.class);

Sometimes it is essential to resolve all contextual instances which implement for example an interface or all beans with the same type but a different qualifier. The following example shows how to do such a lookup which returns all contextual instances (independent of the scope → also dependent scoped instances).

Resolving All Contextual Instances of a Given Type
List<MyServiceInterface> myServiceList = BeanProvider.getContextualReferences(MyServiceInterface.class, false);

Since dependent scoped beans have a special role in CDI (you have to destroy them manually - especially if you get them via a manual lookup), you can also call the previous util method with an additional parameter to filter dependent scoped instances.

Resolving All Contextual Instances of a Given Type without Dependent-scoped Instances
List<MyServiceInterface> myServiceList = BeanProvider.getContextualReferences(MyServiceInterface.class, false, false);

Furthermore, it is possible to trigger the injection of fields of any given instance, if it was not done by the container (e.g. because the class is in a jar-file without beans.xml) and @Inject is used for 1-n fields.

Manually Inject Fields
BeanProvider.injectFields(myObject);

DependentProvider

DependentProvider must be used instead of BeanProvider to obtain instances of dependent-scoped beans to allow for their proper destruction.

When obtaining contextual instances using @Inject, the normal-scoped ones get destroyed along with their associated context. However, instances of dependent-scoped beans — as implied by their name — depend on the lifecycle of the contextual instance which declares them and get destroyed along with this declaring bean’s instance.

However, if dependent-scoped instances are obtained programmatically using DependentProvider, there’s no "declaring bean" to speak of and they must be destroyed manually.

Obtaining and destroying an instance of a dependent-scoped bean using DependentProvider
DependentProvider<MyBean> myBeanProvider = BeanProvider.getDependent(MyBean.class);
MyBean myBean = myBeanProvider.get();

// ...work with myBean...

myBeanProvider.destroy();

BeanManagerProvider

This mechanism provides access to the BeanManager by registering the current BeanManager during the startup. This is really handy if you like to access CDI functionality from places where no CDI based injection is available. If a simple but manual bean-lookup is needed, it is easier to use the BeanProvider.

Resolving the BeanManager
//in most cases the following works without problems:
@Inject
private BeanManager beanManager;

//use
BeanManager beanManager = BeanManagerProvider.getInstance().getBeanManager();

//if CDI based injection is not available.

BeanManagerProvider uses a different approach to find the correct BeanManager, because a portable API for it has only been available from CDI 1.1. However, once you are using CDI 1.1+ DeltaSpike delegates the lookup to the CDI container instead of using its own approach. If you migrate from CDI 1.0 to a later version of CDI and you would like to keep the lookup strategy you used before, you can deactivate the delegation to the container by adding deltaspike.bean-manager.delegate_lookup=false to your config-source (e.g. in /META-INF/apache-deltaspike.properties).

AnnotationInstanceProvider

Java EE provides a standard mechanism for obtaining annotation instances — the AnnotationLiteral class.

public class CurrentUserLiteral extends AnnotationLiteral<CurrentUser> implements CurrentUser {}
CurrentUser user = new CurrentUserLiteral() {};

AnnotationLiteral can however be used only if the annotation class name is known beforehand. AnnotationInstanceProvider is the solution for dynamic creation of annotation instances, with the option to provide a map of values for annotation members. That might be useful in many situations, especially for CDI extension authors. For example:

  • avoiding a compile-time dependency on an annotation class

    Class<? extends Annotation> priorityAnnotationClass = ClassUtils.tryToLoadClassForName("javax.annotation.Priority");
    priorityAnnotationInstance = AnnotationInstanceProvider.of(priorityAnnotationClass, mapOfMemberValues);
  • getting an instance of a dynamically obtained annotation class

    Annotation exceptionQualifier = AnnotationInstanceProvider.of(jsfModuleConfig.getExceptionQualifier());
  • or simply for the sake of a prettier syntax compared to AnnotationLiteral

    CurrentUser principal = AnnotationInstanceProvider.of(CurrentUser.class);

Type-safe ProjectStage

The DeltaSpike ProjectStage mechanism allows to use configuration and implementations depending on the server environment you currently run on.

DeltaSpike provides some pre-defined ProjectStages but it’s also possible to create your own Custom Project Stage, Please, check the DeltaSpike ProjectStage page for more details.

@Exclude

With @Exclude it is possible to annotate beans which should be ignored by CDI even if they are in a CDI enabled archive.

Excluding a Bean in any Case
@Exclude
public class NoBean
{
}
Excluding a Bean in Case of ProjectStageDevelopment
@Exclude(ifProjectStage = ProjectStage.Development.class)
public class MyBean
{
}
Excluding a Bean if the ProjectStage is different from Development
@Exclude(exceptIfProjectStage = ProjectStage.Development.class)
public class MyDevBean
{
}

The following usage allows to exclude a bean based on a configured value (see the supported configuration sources).

Excluding a Bean based on an Expression which Evaluates to True
@Exclude(onExpression = "db==prodDB")
public class DevDbBean
{
}

By default a simple syntax is supported ([TODO]), however, it is possible to provide a custom ExpressionInterpreter for interpreting custom expressions.

Excluding a Bean based on a Custom Expression
@Exclude(onExpression = "db eq prodDB", interpretedBy = SimpleExpressionInterpreter.class)
public class DevDbBean
{
}

public class SimpleExpressionInterpreter implements ExpressionInterpreter<String, Boolean>
{
    @Override
    public Boolean evaluate(String expression)
    {
        if(expression.contains(" eq "))
        {
            //...
        }
        //...
    }
}

In several cases it is also useful to combine this feature with the @Alternative annotation provided by CDI.

In addition to the following snippet, it is required to configure the implementation as alternative in the beans.xml file. This configuration entry will not be changed, for example for different environments, because it just gets active if it is not excluded during the bootstrapping process.

Excluding an Alternative implementation if the ProjectStage is different from Development
@Exclude(exceptIfProjectStage = ProjectStage.Development.class)
@Alternative
public class MyDevBean
{
}

Custom ExpressionInterpreter

By default only a very simple and limited syntax is supported. In real projects there are usually quite concrete requirements. Since it would be very complex to support most of them, it is easier for users to implement an optimized syntax. For such cases a custom ExpressionInterpreter is needed:

@Alternative
@Exclude(onExpression = "environment!=HSQL", interpretedBy = ConfigAwareExpressionInterpreter.class)
public class DevDbBean implements DbBean
{
}

public class ConfigAwareExpressionInterpreter implements ExpressionInterpreter<String, Boolean>
{
    public Boolean evaluate(String expression)
    {
        if (expression == null)
        {
            return false;
        }

        String[] values = expression.split("!=");

        if (values.length != 2)
        {
            throw new IllegalArgumentException("'" + expression + "' is not a supported syntax");
        }

        String configuredValue = ConfigResolver.getPropertyValue(values[0], null);

        //exclude if null or the configured value is different
        return configuredValue == null || !values[1].trim().equalsIgnoreCase(configuredValue);
    }
}

Type-safe View-Config

TODO (Overview)

Literals

Literals allow the instantiation of annotations by extending the abstract class javax.enterprise.util.AnnotationLiteral

Example
public abstract class PayByQualifier
       extends AnnotationLiteral<PayBy>
       implements PayBy {}

 PayBy paybyCheque = new PayByQualifier() { public PaymentMethod value() { return CHEQUE; } };

DeltaSpike provides many annotation literals that you can use, including the following:

  • AlternativeLiteral

  • AnyLiteral

  • ApplicationScopedLiteral

  • ConversationScopedLiteral

  • DefaultLiteral

  • DependentScopeLiteral

  • ModelLiteral

  • NamedLiteral

  • NewLiteral

  • RequestedScopeLiteral

  • SessionScopeLiteral

  • Singleton

  • SpecializesLiteral

  • TypedLiteral

Messages and i18n

The following implementation is the minimal effort to use type-safe messages (which are hardcoded in this case).

Simple Type-safe Message
@MessageBundle
public interface SimpleMessage
{
    @MessageTemplate("Welcome to DeltaSpike")
    String welcomeToDeltaSpike();
}

The following implementation uses the key welcome_to_deltaspike to do a lookup in the default message bundle. The default bundle has the same name as the interface (but .properties instead of .java (/.class) as file extension).

Internationalized Type-safe Message
@MessageBundle
public interface SimpleMessage
{
    @MessageTemplate("{welcome_to_deltaspike}")
    String welcomeToDeltaSpike();
}

org.apache.deltaspike.example.message.SimpleMessage

->

org/apache/deltaspike/example/message/SimpleMessage.properties
org/apache/deltaspike/example/message/SimpleMessage_en.properties
org/apache/deltaspike/example/message/SimpleMessage_de.properties
...

//content (as usual in message bundle files):
welcome_to_deltaspike=Welcome to DeltaSpike

The following implementation uses the key welcome_to_deltaspike to do a lookup in a custom message bundle known by CustomMessageResolver.

Internationalized Type-safe Message
@MessageBundle
@MessageContextConfig(messageResolver = CustomMessageResolver.class)
public interface SimpleMessage
{
    @MessageTemplate("{welcome_to_deltaspike}")
    String welcomeToDeltaSpike();
}

@MessageContextConfig allows to provide a custom MessageResolver, MessageInterpolator and LocaleResolver.

The following implementation shows the usage of an internationalized simple type-safe message.

Internationalized Type-safe Message with Parameter/s
@MessageBundle
@MessageContextConfig(messageInterpolator = CustomMessageInterpolator.class)
public interface SimpleMessage
{
    //in the message bundle: welcome_to=Welcome to %s

    @MessageTemplate("{welcome_to}")
    String welcomeTo(String name);
}

//...
public class MyBean
{
    @Inject
    private SimpleMessage messages;

    public String welcomeToDeltaSpike
    {
        return this.messages.welcomeTo("DeltaSpike");
    }
}

Dynamic Message Builder

Creating Message Instances

The following implementation creates an instance of Message for the key hello. The final text will be resolved and interpolated lazily. Later on it might be supported to provide a different MessageContext via #toString(MessageContext) like it is in MyFaces CODI right now.

You can use #argument(String) to pass these arguments to the message template specified on #template(String) method. The template pattern uses printf-style format strings.

public class MyBean
{

    @Inject
    private MessageContext messageContext;

    public void action()
    {
        Message message = this.messageContext.message();
        write(message.template("Hello %s from %s").argument("World").argument("DeltaSpike"));
    }
    //...
}

Besides the static configuration via @MessageContextConfig#messageSource, you can also specify the message sources dynamically.

@Inject
private MessageContext messageContext;

public void action()
{
    Message message = this.messageContext.messageSource("org.apache.deltaspike.example.message.Messages").message();
    write(message.template("{hello}").argument("World").argument("DeltaSpike"));
}
//...
->

org/apache/deltaspike/example/message/Messages.properties
org/apache/deltaspike/example/message/Messages_en.properties
org/apache/deltaspike/example/message/Messages_de.properties
...

//content (as usual) in message bundle files:
hello=Hello %s from %s

Customizing the Message Context

MessageResolver

A message-resolver is responsible for creating the message-text based on the message-descriptor (key or inline-text), the current locale (and in some cases the message-payload). (The supported format, for example, if it is required to escape a key, if inline-text is supported,…​ depends on the concrete implementation.) In case of a message-key, the message-resolver has to transform it to the message-text by looking it up in a message source like a resource-bundle.

Configuration of a message-resolver

Besides the static configuration via @MessageContextConfig#messageResolver, you can use it dynamically via passing a custom message-resolver instance to the current messageContext:

@Inject
private MessageContext messageContext;

//...

Message message = this.messageContext.messageResolver(new CustomMessageResolver()).message();

The result of a MessageResolver is the message-text. The text might contain placeholders which are processed by a MessageInterpolator

MessageInterpolator

A MessageInterpolator replaces the placeholders in a message-text with the arguments of the message.

Configuration of a message-interpolator

Besides the static configuration via `@MessageContextConfig#messageInterpolator, you can use it dynamically via passing a custom message-interpolator instance to the current messageContext:

@Inject
private MessageContext messageContext;

//...
Message message = this.messageContext.messageInterpolator(new CustomMessageInterpolator()).message();
LocaleResolver

A locale resolver provides the current locale. The locale is, for example, used to by a MessageResolver to choose the correct language for the message-text.

Configuration of a locale-resolver

Besides the static configuration via `@MessageContextConfig#localeResolver, you can use it dynamically via passing a custom locale-resolver instance to the current messageContext:

@Inject
private MessageContext messageContext;

//...
Message message = this.messageContext.localeResolver(new CustomLocaleResolver()).message();

Injecting Resources

DeltaSpike has simple APIs for performing basic resource loading and property file reading.

@Inject
@InjectableResource("myfile.properties")
private InputStream inputStream;

This can be used to read resources from the classpath or from the file system using the two default implementations — ClasspathResourceProvider and FileResourceProvider — or from any other source using a custom provider.

Custom resource providers

The InjectableResourceProvider interface can be implemented to allow reading from alternate sources if needed (e.g. database LOBs, NoSQL storage areas). A base class called AbstractResourceProvider is provided by DeltaSpike and contains most of the methods for potential implementations. The only method which must be provided is the readStream(InjectableResource) which returns an InputStream.

Exception Control

Exception handling in DeltaSpike is based around the CDI eventing model. While the implementation of exception handlers may not be the same as a CDI event, and the programming model is not exactly the same as specifying a CDI event observer, the concepts are very similar. DeltaSpike makes use of events for many of its features. Eventing is actually the only way to start using DeltaSpike’s exception handling.

This event is fired either by the application or a DeltaSpike exception handling integration. DeltaSpike then hands the exception off to a chain of registered handlers, which deal with the exception appropriately. The use of CDI events to connect exceptions to handlers makes this strategy of exception handling non-invasive and minimally coupled to the exception handling infrastructure.

The exception handling process remains mostly transparent to the developer. In most cases, you register an exception handler simply by annotating a handler method. Alternatively, you can handle an exception programmatically, just as you would observe an event in CDI.

Usage

The entire exception handling process starts with an event. This helps keep your application minimally coupled to DeltaSpike, but also allows for further extension. Exception handling in DeltaSpike is all about letting you take care of exceptions the way that makes the most sense for your application Events provide this delicate balance. Firing the event is the main way of starting the exception handling proccess.

Manually firing an event to use DeltaSpike’s exception handling is primarily used in your own try/catch blocks. It is very painless and also easy. Let’s examine a sample that might exist inside of a simple business logic lookup into an inventory database:

public class InventoryActions {
    @PersistenceContext private EntityManager em;
    @Inject private Event<ExceptionToCatchEvent> catchEvent;

    public Integer queryForItem(Item item) {
        try {
          Query q = em.createQuery("SELECT i from Item i where i.id = :id");
          q.setParameter("id", item.getId());
          return q.getSingleResult();
        } catch (PersistenceException e) {
          catchEvent.fire(new ExceptionToCatchEvent(e));
        }
    }
}

The Event of generic type ExceptionToCatchEvent is injected into your class for use later within a try/catch block.

The event is fired with a new instance of ExceptionToCatchEvent constructed with the exception to be handled.

Exception Handlers

As an application developer (i.e., an end user of DeltaSpike’s exception handling), you’ll be focused on writing exception handlers. An exception handler is a method on a CDI bean that is invoked to handle a specific type of exception. Within that method, you can implement any logic necessary to handle or respond to the exception.

If there are no exception handlers for an exception, the exception is rethrown - except ExceptionToCatchEvent#optinal is set to true

Given that exception handler beans are CDI beans, they can make use of dependency injection, be scoped, have interceptors or decorators and any other functionality available to CDI beans.

Exception handler methods are designed to follow the syntax and semantics of CDI observers, with some special purpose exceptions explained in this guide. The advantage of this design is that exception handlers will be immediately familiar to you if you are studying or well-versed in CDI.

In this and subsequent sections, you’ll learn how to define an exception handler, explore how and when it gets invoked, modify an exception and a stack trace, and even extend exception handling further through events that are fired during the handling workflow. We’ll begin by covering the two annotations that are used to declare an exception handler, @ExceptionHandler and @Handles, and @BeforeHandles to create a callback before the handler is called.

Exception handlers are considered equal if they both handle the same exception class, have the same qualifiers, the same ordinal and the same value for isBeforeHandler().

Exception handlers are contained within exception handler beans, which are CDI beans annotated with @ExceptionHandler. Exception handlers are methods which have a parameter which is an instance of ExceptionEvent<T extends Throwable> annotated with the @Handles annotation.

@ExceptionHandler

The @ExceptionHandler annotation is simply a marker annotation that instructs the DeltaSpike exception handling CDI extension to scan the bean for handler methods.

Let’s designate a CDI bean as an exception handler by annotating it with @ExceptionHandler.

@ExceptionHandler
public class MyHandlers {}

That’s all there is to it. Now we can begin defining exception handling methods on this bean.

@Handles and @BeforeHandles

@Handles is a method parameter annotation that designates a method as an exception handler. Exception handler methods are registered on beans annotated with @ExceptionHandler. DeltaSpike will discover all such methods at deployment time.

Let’s look at an example. The following method is invoked for every exception that DeltaSpike processes and prints the exception message to stdout. (Throwable is the base exception type in Java and thus represents all exceptions).

@ExceptionHandler
public class MyHandlers
{
    void printExceptions(@Handles ExceptionEvent<Throwable> evt)
    {
        System.out.println("Something bad happened:" +
        evt.getException().getMessage());
        evt.handleAndContinue();
    }
}

The @Handles annotation on the first parameter designates this method as an exception handler (though it is not required to be the first parameter). This parameter must be of type ExceptionEvent<T extends Throwable>, otherwise it is detected as a definition error. The type parameter designates which exception the method should handle. This method is notified of all exceptions (requested by the base exception type Throwable).

The ExceptionEvent instance provides access to information about the exception and can be used to control exception handling flow. In this case, it is used to read the current exception being handled in the exception chain, as returned by getException().

This handler does not modify the invocation of subsequent handlers, as designated by invoking handleAndContinue() on ExceptionEvent. As this is the default behavior, this line could be omitted.

The @Handles annotation must be placed on a parameter of the method, which must be of type ExceptionEvent<T extends Throwable>. Handler methods are similar to CDI observers and, as such, follow the same principles and guidelines as observers (such as invocation, injection of parameters, qualifiers, etc) with the following exceptions:

  • a parameter of a handler method must be a ExceptionEvent

  • handlers are ordered before they are invoked (invocation order of observers is non-deterministic)

  • any handler can prevent subsequent handlers from being invoked

In addition to designating a method as exception handler, the @Handles annotation specifies an ordinal about when the method should be invoked relative to other handler methods of the same type. Handlers with higher ordinal are invoked before handlers with a lower ordinal that handle the same exception type. The default ordinal (if not specified) is 0.

The @BeforeHandles designates a method as a callback to happen before handlers are called.

Let’s take a look at more sophisticated example that uses all the features of handlers to log all exceptions.

@ExceptionHandler
public class MyHandlers
{
   void logExceptions(@BeforeHandles @WebRequest ExceptionEvent<Throwable> evt, Logger log)
   {
      log.warn("Something bad happened: " + evt.getException().getMessage());
   }

   void logExceptions(@Handles @WebRequest ExceptionEvent<Throwable> evt, Logger log)
   {
      // possibly send a HTTP Error code
   }
}

This handler has a default ordinal of 0 (the default value of the ordinal attribute on @Handles).

This handler is qualified with @WebRequest. When DeltaSpike calculates the handler chain, it filters handlers based on the exception type and qualifiers. This handler will only be invoked for exceptions passed to DeltaSpike that carry the @WebRequest qualifier. We’ll assume this qualifier distinguishes a web page request from a REST request.

Any additional parameters of a handler method are treated as injection points. These parameters are injected into the handler when it is invoked by DeltaSpike. In this case, we are injecting a Logger bean that must be defined within the application (or by an extension).

A handler is guaranteed to only be invoked once per exception (automatically muted), unless it re-enables itself by invoking the unmute() method on the ExceptionEvent instance.

Handlers must not throw checked exceptions, and should avoid throwing unchecked exceptions. Should a handler throw an unchecked exception it will propagate up the stack and all handling done via DeltaSpike will cease. Any exception that was being handled will be lost.

Ordinal

When DeltaSpike finds more than one handler for the same exception type, it orders the handlers by ordinal. Handlers with higher ordinal are executed before handlers with a lower ordinal. If DeltaSpike detects two handlers for the same type with the same ordinal, the order is non-deterministic.

Let’s define two handlers with different ordinals:

void handleIOExceptionFirst(@Handles(ordinal = 100) ExceptionEvent<IOException> evt)
{
   System.out.println("Invoked first");
}

void handleIOExceptionSecond(@Handles ExceptionEvent<IOException> evt)
{
 System.out.println(Invoked second);
}

The first method is invoked first since it has a higher ordinal (100) than the second method, which has the default ordinal (0).

To summarize, here’s how DeltaSpike determines the order of handlers to invoke (until a handler marks exception as handled):

  1. Unwrap exception stack

  2. Begin processing root cause

  3. Invoke any callback methods annotated with @BeforeHandles for the closest type to the exception

  4. Find handler for the closest type to the exception

  5. If multiple handlers for same type, invoke handlers with higher ordinal first

  6. Continue above steps for each exception in stack

Exception Chain Processing

When an exception is thrown, chances are it is nested (wrapped) inside other exceptions. (If you’ve ever examined a server log, you’ll appreciate this fact). The collection of exceptions in its entirety is termed an exception chain.

The outermost exception of an exception chain (e.g., EJBException, ServletException, etc) is probably of little use to exception handlers. That’s why DeltaSpike does not simply pass the exception chain directly to the exception handlers. Instead, it intelligently unwraps the chain and treats the root exception cause as the primary exception.

The first exception handlers to be invoked by DeltaSpike are those that match the type of root cause. Thus, instead of seeing a vague EJBException, your handlers will instead see an meaningful exception such as ConstraintViolationException. This feature, alone, makes DeltaSpike’s exception handling a worthwhile tool.

DeltaSpike continues to work through the exception chain, notifying handlers of each exception in the stack, until a handler flags the exception as handled or the whole exception chain has been iterated. Once an exception is marked as handled, DeltaSpike stops processing the exception chain. If a handler instructs DeltaSpike to rethrow the exception (by invoking ExceptionEvent#throwOriginal(), DeltaSpike will rethrow the exception outside the DeltaSpike exception handling infrastructure. Otherwise, it simply returns flow control to the caller.

Consider a exception chain containing the following nested causes (from outer cause to root cause):

  • EJBException

  • PersistenceException

  • SQLGrammarException

DeltaSpike will unwrap this exception and notify handlers in the following order:

  • SQLGrammarException

  • PersistenceException

  • EJBException

If there’s a handler for PersistenceException, it will likely prevent the handlers for EJBException from being invoked, which is a good thing since what useful information can really be obtained from EJBException?

APIs for Exception Information and Flow Control

There are two APIs provided by DeltaSpike that should be familiar to application developers:

  • ExceptionEvent

  • ExceptionStackEvent

ExceptionEvent

In addition to providing information about the exception being handled, the ExceptionEvent object contains methods to control the exception handling process, such as rethrowing the exception, aborting the handler chain or unmuting the current handler. Five methods exist on the ExceptionEvent object to give flow control to the handler

  • abort() — terminate all handling immediately after this handler, does not mark the exception as handled, does not re-throw the exception.

  • throwOriginal() — continues through all handlers, but once all handlers have been called (assuming another handler does not call abort() or handled()) the initial exception passed to DeltaSpike is rethrown. Does not mark the exception as handled.

  • handled() — marks the exception as handled and terminates further handling.

  • handleAndContinue() — default. Marks the exception as handled and proceeds with the rest of the handlers.

  • skipCause() — marks the exception as handled, but proceeds to the next cause in the cause container, without calling other handlers for the current cause.

  • rethrow(Throwable) — Throw a new exception after this handler is invoked

Once a handler is invoked it is muted, meaning it will not be run again for that exception chain, unless it is explicitly marked as unmuted via the unmute() method on ExceptionEvent.

Scopes

DeltaSpike Core provides the API and SPI for several scopes. Currently all scopes are only implemented in the JSF module.

@WindowScoped

@ViewAccessScoped

@GroupedConversationScoped

Creating a Custom CDI Scope

To create a custom CDI scope to match your needs, complete the following steps:

  1. Create an Annotation with annotated with @javax.inject.Scope;

    Example
    @Scope
    @Retention(RetentionPolicy.RUNTIME)
    @Target({ElementType.TYPE,ElementType.METHOD,ElementType.FIELD})
    public @interface ACustomScope {}
  2. Create an Extension to add the scope and a context for it.

    Example
    public class ACustomScopeExtension implements Extension, Serializable {
    
        public void addACustomScope(@Observes final BeforeBeanDiscovery event) {
            event.addScope(ACustomScope.class, true, false);
        }
    
        public void registerACustomScopeContext(@Observes final AfterBeanDiscovery event) {
            event.addContext(new ACustomScopeContext());
        }
    }
  3. Implement a javax.enterprise.context.spi.Context interface to hold the javax.enterprise.inject.spi.Bean instances according to your needs.

    Example
    public class ACustomScopeContext implements Context, Serializable {
    
      // Get the scope type of the context object.
        public Class<? extends Annotation> getScope() {
            return ACustomScope.class;
        }
    
        // Return an existing instance of certain contextual type or create a new instance by calling
        // javax.enterprise.context.spi.Contextual.create(CreationalContext) and return the new instance.
        public <T> T get(Contextual<T> contextual, CreationalContext<T> creationalContext) {
            Bean bean = (Bean) contextual;
            // you can store the bean somewhere
            if (somewhere.containsKey(bean.getName())) {
                return (T) somewhere.get(bean.getName());
            } else {
                T t = (T) bean.create(creationalContext);
                somewhere.put(bean.getName(), t);
                return t;
            }
        }
    
        // Return an existing instance of a certain contextual type or a null value.
        public <T> T get(Contextual<T> contextual) {
            Bean bean = (Bean) contextual;
            // you can store the bean somewhere
            if (somewhere.containsKey(bean.getName())) {
                return (T) somewhere.get(bean.getName());
            } else {
                return null;
            }
        }
    
      // Determines if the context object is active.
        public boolean isActive() {
            return true;
        }
    
    }

Deactivatable

DeltaSpike allows you to deactivate its own pre-configured parts (like Extensions, event-broadcasters,…​). Therefore DeltaSpike offers org.apache.deltaspike.core.spi.activation.ClassDeactivator and org.apache.deltaspike.core.spi.activation.Deactivatable.

A ClassDeactivator allows to specify deactivated classes (if they implement Deactivatable) which can’t be deactivated/customized via std. CDI mechanisms (like the veto-method or alternative/specialized CDI-beans). This might be the case e.g. for CDI Extensions because CDI mechanisms are not available at startup time.

Use it mainly to deactivate specific parts explicitly (blacklist approach), if there is an issue with such parts (and waiting for the next release isn’t an option).

You just need to implement your ClassDeactivator.

The ClassDeactivator should be resolved by any ConfigSource using the key org.apache.deltaspike.core.spi.activation.ClassDeactivator. For example, if we need to provide our own version of the SecurityExtension, we can disable the SecurityExtension provided by DeltaSpike with the following ClassDeactivator:

Disable a specific extension
public class CustomClassDeactivator implements ClassDeactivator
{
    private static final long serialVersionUID = 1L;

    @Override
    public Boolean isActivated(Class<? extends Deactivatable> targetClass)
    {
        if (targetClass.equals(SecurityExtension.class))
        {
            return Boolean.FALSE;
        }
        return null; //no result for the given class
    }
}

Now, we can use the file /META-INF/apache-deltaspike.properties (or any other ConfigSource) with the following key/value:

org.apache.deltaspike.core.spi.activation.ClassDeactivator=org.test.CustomClassDeactivator

SecurityExtension still gets started, because it isn’t possible to veto it, however, it isn’t processing beans (once deactivated) and therefore it’s e.g. possible to extend and customize the default implementation provided by DeltaSpike.

The following listing shows how to enable only a minimal set of extensions. Technically that’s possible, however, it isn’t suggested to use such an approach, because you might disable mechanisms need later on (in your project).

Possible but not suggested
public class WhitelistFilter implements ClassDeactivator
{
    private List<Class<?>> limitedExtensions =
      new ArrayList<Class<?>>()
    {{
        add(ConfigurationExtension.class);
        add(PartialBeanBindingExtension.class);
        add(RepositoryExtension.class);
    }};

    @Override
    public Boolean isActivated(
      Class<? extends Deactivatable> deactivatableClass)
    {
        return !Extension.class.isAssignableFrom(deactivatableClass) ||
            limitedExtensions.contains(deactivatableClass);
    }
}

Instead it’s better to disable the part you really like to deactivate (see CustomClassDeactivator).

Deactivate Deactivatable-Classes via Config

The default implementation of ClassDeactivator allows to deactivate classes by adding config-entries to one of your config-sources (like META-INF\apache-deltaspike.properties). The following example shows how it would look like e.g. in case of the SecurityExtension:

deactivate.org.apache.deltaspike.security.impl.extension.SecurityExtension=true

Asynchronous Operations

DeltaSpike provides support for executing code in an asynchronous manner. The behavior is implemented as three different interceptors for your beans.

  • @Futureable - Designed for bean methods that return Future’s of some form. The method call will automatically be submitted to an ExecutorService

  • @Locked - Ability to prevent concurrent access to a method based on its usage of reads/writes.

  • @Throttled - Ability to limit how frequently a method can be invoked.

Utilities

DeltaSpike provides many utility classes (no constructor / static methods) that can be useful for your project.

Below you can find an information about these classes.

AnnotationUtils

Utilities for working with annotations on methods and classes.

  • #findAnnotation — obtains an Annotation instance of a given annotation Class from the given Annotation[], recursing any possible stereotype tree along the way

  • #extractAnnotationFromMethod — uses findAnnotation to obtain an Annotation from a given Method

  • #extractAnnotationFromMethodOrClass — uses findAnnotation to obtain an Annotation on either the given Method or the given Class, in that order

  • #getQualifierHashCode — computes the hashCode of a qualifier annotation, taking into account only the "binding" members (not annotated @Nonbinding)

ArraysUtils

A collection of utilities for working with Arrays

  • #asSet — Create a set from an array. If the array contains duplicate objects, the last object in the array will be placed in resultant set.

BeanUtils

A set of utility methods for working with beans.

  • #getQualifiers — Extract the qualifiers from a set of annotations.

  • #extractAnnotation — Extract the annotations.

  • #createInjectionPoints — Given a method, and the bean on which the method is declared, create a collection of injection points representing the parameters of the method.

ContextUtils

A set of utility methods for working with contexts.

  • #isContextActive — Checks if the context for the scope annotation is active.

ClassDeactivationUtils

Helper methods for ClassDeactivator

  • #isActivated — Evaluates if the given Deactivatable is active.

To add a custom ClassDeactivator add org.apache.deltaspike.core.spi.activation.ClassDeactivator=my.CustomClassDeactivator to META-INF\apache-deltaspike.properties. Or configure it via a custom ConfigSource.

ExceptionUtils

Helper methods to deal with Exceptions

  • #throwAsRuntimeException — helper which allows to use a trick to throw a catched checked exception without a wrapping exception.

  • #changeAndThrowException — helper which allows to use a trick to throw a cached checked exception without a wrapping exception.

PropertyFileUtils

Helper methods for Property files

  • #resolvePropertyFiles — Allows to lookup for resource bundle files.

  • #loadProperties — Load a Properties file from the given URL.

  • #getResourceBundle — Return the ResourceBundle for the current default Locale.

ProxyUtils

Helper for CDI proxies

  • #getUnproxiedClass — Return class of the real implementation.

  • #isProxiedClass — Analyses if the given class is a generated proxy class.

StringUtils

A collection of utilities for working with Strings.

  • #isEmpty — return true if the String is null or empty ( string.trim().isEmpty() )