I want to solve the following problem and need advice, what the best solution is.
I have a bundle A in which a service interface X is defined. A bundle B provides a service implementation of X and contributes the implementation to the tool. A and B use Google Guice and Peaberry to configure the setup of the objects.
There are two possibilities I can use to contribute the service implementation:
Using an eclipse extension:
In this solution I can use the GuiceExtensionFactory mechanism of Peaberry to create the service implementation using Guice and therefore can inject stuff needed by the implementation. The disadvantage here is that in the bundle defining the extension point, I need the boilerplate code for the resolution of the extensions because there is to my knowledge no way to get the extensions injected into the class which uses the extensions.
This looks like this:
<extension point="A.service.X">
<xservice
...
class="org.ops4j.peaberry.eclipse.GuiceExtensionFactory:B.XImpl"
.../>
</extension>
<extension
point="org.ops4j.peaberry.eclipse.modules">
<module
class="B.XModule">
</module>
</extension>
but I need the boilerplate code like this:
private List<X> getRegisteredX() {
final List<X> ximpls = new ArrayList<>();
for (final IConfigurationElement e : Platform.getExtensionRegistry().getConfigurationElementsFor( X_EXTENSION_POINT_ID)) {
try {
final Object object = e.createExecutableExtension("class"); //$NON-NLS-1$
if (object instanceof X) {
ximpls.add((X) object);
}
} catch (final CoreException ex) {
// Log
}
}
return ximpls;
}
Using an OSGI service:
My main problem here is to ensure that the service is registered. I want the bundle loaded lazily, so at least an access to one of the classes of the bundle is required. Registering the service programmatically using Peaberry has an issue, because nobody ever asks for a class of the bundle. The solution is to provide the service as a declarative service, but I do not know a way to create the service implementation in a way, that I can use Guice to inject required objects.
So I have some questions:
Is there something I do not know so far that implements the code needed to read the extensions at an extension point generically and allows to inject the extensions to the class using the extensions?
Is there a way to ensure that the service is provided even if it is added using the standard Peaberry mechanism, i.e., the bundle is activated when the service is requested?
Is there a way like the GuiceExtensionFactory for declarative services, so that the creation of the service implementation can be done by the injector of the bundle?
Something that look like:
<?xml version="1.0" encoding="UTF-8"?>
<scr:component xmlns:scr="http://www.osgi.org/xmlns/scr/v1.1.0" name="Ximpl">
<implementation class="some.generic.guiceaware.ServiceFactory:B.Ximpl"/>
<service>
<provide interface="A.X"/>
</service>
</scr:component>
Summarized, I want a service implementation generated by Guice and I want to get the service implementations simply injected into the classes using the service without extensive boilerplate code. Has anybody a solution for that?
Sorry, to ask, but I searched the web for quite a while and so far I did not find a solution.
Thanks and best regards,
Lars
I found a solution, but since I did not find it without a lot of trying out and thinking I thought I share it here. From the options I mentioned in my posting, my solution uses the first one, that is Eclipse extension points and extensions. In order to use Guice in the context of extension points there are two aspects to consider:
Providing an extension that is created by an Guice injector
This is explained very well here: https://code.google.com/p/peaberry/wiki/GuiceExtensionFactory. There is one remark to make from my side. The creation of the extension object is done in an injector inside of the GuiceExtensionFactory, so it is an own context, which needs to be configured by the module given as additional extension to the factory. This can become an issue, if you have other needs that require creating the injector in the bundle on your own.
Defining an extension point so that the extensions are simply injected into the classes which use the extensions.
First thing to do is to define the extension point schema file as normally. It should contain the reference of an interface that has to be implemented by the extensions.
The id of the extension point has to be connected to the interface which is provided by the extensions and which is injected by guice/peaberry. Therefore peaberry provides an annotation to be used to annotate the interface:
import org.ops4j.peaberry.eclipse.ExtensionBean;
#ExtensionBean("injected.extension.point.id")
public interface InjectedInterface {
...
}
On some web pages you also find the information that if the id is equal to the qualified name of the interface, it can be found directly without the annotation but I did not try this out.
In order to enable the injection, you have to do two things to configure the Guice injector creation.
First the EclipseRegistry object of Peaberry has to be set as ServiceRegistry. Second the binding of the extension implementations to a provided service has to be done.
The injector creation has to be done in this way:
import org.osgi.framework.BundleContext;
import com.google.inject.Guice;
import com.google.inject.Injector;
import org.ops4j.peaberry.eclipse.EclipseRegistry;
import static org.ops4j.peaberry.Peaberry.*;
void initializer() {
Injector injector = Guice.createInjector(osgiModule(context, EclipseRegistry.eclipseRegistry()), new Module() {
binder.bind(iterable(InjectedInterface.class)).toProvider(service(InjectedInterface.class).multiple());
});
}
The extension implementations can then simply be injected like this:
private Iterable<InjectedInterface> registeredExtensions;
#Inject
void setSolvers(final Iterable<InjectedInterface> extensions) {
registeredExtensions = extensions;
}
With the described way it is possible to have injected extensions which are implemented by classes using Guice to get dependencies injected.
I did not find a solution to use osgi services so far, but perhaps there is someone who has an idea.
Best regards,
Lars
Related
I designed a new project using Guice/Gin so I could make our code more modular and swap-able especially when testing.
However, I am not able to find out how to make this work in practice. I was under the impression that I could just create a new Gin/Guice module in my test and install my 'base' module, overloading any bindings that I want to replace with specific testing implementations.
I don't want to have to use GWTTestCase and load my entire module, because it is very slow and unecissary for the types of granular testing I need to do.
I have tried using Jukito (http://code.google.com/p/jukito/), gwt-test-utils (http://code.google.com/p/gwt-test-utils/wiki/HowToUseWithGIN) and also some resources on doing this with guice (http://fabiostrozzi.eu/2011/03/27/junit-tests-easy-guice/).
None of these methods are yielding any results.
I think the Guice approach might work, if I defined a mirror guice module for my Gin module. However I really don't want to have to manage both of these. I really just want to test my GIN module like I would assume people test with Guice.
I feel like this should be really simple, can anyone point me to examples that work?
Update
Another way of looking at this question is:
How do I get the examples on the Jukito site (http://code.google.com/p/jukito/) work when the classes I am injecting are in an exernal Gin module?
**Update - In reference to Thomas Boyer's answer **
Thanks for the hint Tom, I was not able to find examples of using the adapter but I tried augmenting the Jukito examples to use the GinModuleAdapter anyway:
#RunWith(JukitoRunner.class)
public class MyGinTest {
public static class Module extends JukitoModule {
protected void configureTest() {
install(new GinModuleAdapter(new ClientModule()));
}
}
#Test
#Inject
public void testAdd(SyncedDOMModel mod){
assertNotNull(mod);
}
}
When I tried to run this test I recieved this exception:
java.lang.AssertionError: should never be actually called
at com.google.gwt.inject.rebind.adapter.GwtDotCreateProvider.get(GwtDotCreateProvider.java:43)
at com.google.inject.internal.InternalFactoryToProviderAdapter.get(InternalFactoryToProviderAdapter.java:40)
at com.google.inject.internal.ProviderToInternalFactoryAdapter$1.call(ProviderToInternalFactoryAdapter.java:46)
at com.google.inject.internal.InjectorImpl.callInContext(InjectorImpl.java:1031)
at com.google.inject.internal.ProviderToInternalFactoryAdapter.get(ProviderToInternalFactoryAdapter.java:40)
at com.google.inject.Scopes$1$1.get(Scopes.java:65)
at com.google.inject.internal.InternalFactoryToProviderAdapter.get(InternalFactoryToProviderAdapter.java:40)
at com.google.inject.internal.InternalInjectorCreator$1.call(InternalInjectorCreator.java:204)
at com.google.inject.internal.InternalInjectorCreator$1.call(InternalInjectorCreator.java:198)
at com.google.inject.internal.InjectorImpl.callInContext(InjectorImpl.java:1024)
at com.google.inject.internal.InternalInjectorCreator.loadEagerSingletons(InternalInjectorCreator.java:198)
at com.google.inject.internal.InternalInjectorCreator.injectDynamically(InternalInjectorCreator.java:179)
at com.google.inject.internal.InternalInjectorCreator.build(InternalInjectorCreator.java:109)
at com.google.inject.Guice.createInjector(Guice.java:95)
at com.google.inject.Guice.createInjector(Guice.java:72)
at com.google.inject.Guice.createInjector(Guice.java:62)
at org.jukito.JukitoRunner.ensureInjector(JukitoRunner.java:118)
at org.jukito.JukitoRunner.computeTestMethods(JukitoRunner.java:177)
at org.jukito.JukitoRunner.validateInstanceMethods(JukitoRunner.java:276)
at org.junit.runners.BlockJUnit4ClassRunner.collectInitializationErrors(BlockJUnit4ClassRunner.java:102)
at org.junit.runners.ParentRunner.validate(ParentRunner.java:344)
at org.junit.runners.ParentRunner.<init>(ParentRunner.java:74)
at org.junit.runners.BlockJUnit4ClassRunner.<init>(BlockJUnit4ClassRunner.java:55)
at org.jukito.JukitoRunner.<init>(JukitoRunner.java:72)
My gin module is part of a GWTP project, and looks like this:
public class ClientModule extends AbstractPresenterModule {
#Override
protected void configure() {
install(new DefaultModule(ClientPlaceManager.class));
bindPresenter(MainPagePresenter.class, MainPagePresenter.MyView.class,
MainPageView.class, MainPagePresenter.MyProxy.class);
bindConstant().annotatedWith(DefaultPlace.class).to(NameTokens.main);
bindPresenterWidget(MapTreePresenter.class,
MapTreePresenter.MyView.class, MapTreeView.class);
bindPresenterWidget(MapTreeItemPresenter.class,
MapTreeItemPresenter.MyView.class, MapTreeItemView.class);
bind(ResourcePool.class).to(DefferredResourcePool.class);
bind(WebSocket.class).to(WebSocketImpl.class);
}
}
As you can somewhat see, the class I am injecting in my test SyncedDOMModel, uses a WebSocket which I bind in my module. When I am testing, I don't want to use a real websocket and server. So I want to overload that binding in my test, with a class that basically emulates the whole thing. It's easier to just inject a different implementation of the WebSocket in this case rather than use mocking.
If it helps, this is a basic outline of the SyncedDOMMOdel class:
public class SyncedDOMMOdel {
....
#Inject
public SyncedDOMModel(WebSocket socket){
this.socket = socket;
}
....
}
You can use the GinModuleAdapter to use any GinModule as a Guice Module.
Obviously, you won't benefit from GIN's specific features: default to GWT.create() when something has no particular binding (this includes interfaces and abstract classes, which would throw in Guice), and automatically search for a RemoteService interface when an interface whose name ends Async has no specific binding.
And you won't be able to use anything that depends on JSNI or deferred binding (GWT.create()), as in any non-GWTTestCase unit test.
Recently I was looking at some source code provided by community leaders in their open source implementations. One these projects made use of IOC. Here is sample hypothetical code:
public class Class1
{
private ISomeInterface _someObject;
public Class1(ISomeInterface someObject)
{
_someObject = someObject;
}
// some more code and then
var someOtherObject = new SomeOtherObject();
}
My question is not about what the IOCs are for and how to use them in technical terms but rather what are the guidelines regarding object creation. All that effort and then this line using "new" operator. I don't quite understand. Which object should be created by IOC and for which ones it is permissible to be created via the new operator?
As a general rule of thumb, if something is providing a service which may want to be replaced either for testing or to use a different implementation (e.g. different authentication services) then inject the dependency. If it's something like a collection, or a simple data object which isn't providing behaviour which you'd ever want to vary, then it's fine to instantiate it within the class.
Usually you use IoC because:
A dependency that can change in the future
To code against interfaces, not concrete types
To enable mocking these dependencies in Unit Testing scenarios
You could avoid using IoC in the case where you don't control the dependency, for example an StringBuilder is always going to be an StringBuilder and have a defined behavior, and you usually don't really need to mock that; while you might want to mock an HttpRequestBase, because it's an external dependency on having an internet connection, for example, which is a problem during unit tests (longer execution times, and it's something out of your control).
The same happens for database access repositories and so on.
I start loving OSGi services more and more and want to realize a lot more of my components as services. Now I'm looking for best-practice, especially for UI components.
For Listener-relations I use the whiteboard-pattern, which IMHO opinion is the best approach. However if I want more than just notifications, I can think of three possible solutions.
Imagine the following scenario:
interface IDatabaseService {
EntityManager getEntityManager();
}
[1] Whiteboard Pattern - with self setting service
I would create a new service interface:
interface IDatabaseServiceConsumer {
setDatabaseService(IDatabaseService service);
}
and create a declarative IDatabaseService component with a bindConsumer method like this
protected void bindConsumer(IDatabaseServiceConsumer consumer) {
consumer.setDatabaseService(this);
}
protected void unbindConsumer(IDatabaseServiceConsumer consumer) {
consumer.setDatabaseService(null);
}
This approach assumes that there's only one IDatabaseService.
[Update] Usage would look like this:
class MyUIClass ... {
private IDatabaseService dbService;
Consumer c = new IDatabaseServiceConsumer() {
setDatabaseService(IDatabaseService service) {
dbService = service;
}
}
Activator.registerService(IDatabaseServiceConsumer.class,c,null);
...
}
[2] Make my class a service
Image a class like
public class DatabaseEntryViewer extends TableViewer
Now, I just add bind/unbind methods for my IDatabaseService and add a component.xml and add my DatabaseEntryViewer. This approach assumes, that there is a non-argument constructor and I create the UI components via a OSGi-Service-Factory.
[3] Classic way: ServiceTracker
The classic way to register a static ServiceTracker in my Activator and access it. The class which uses the tracker must handle the dynamic.
Currently I'm favoring the first one, as this approach doesn't complicated object creation and saves the Activator from endless, static ServiceTrackers.
I have to agree with #Neil Bartlett, your option 1 is backward. You are in effect using an Observer/Observable pattern.
Number 2 is not going to work, since the way UI objects lifecycles are managed in RCP won't allow you to do what you want. The widget will have to be created as part of the initialization of some sort of view container (ViewPart, Dialog, ...). This view part is typically configured and managed via the Workbench/plugin mechanism. You should work with this, not against it.
Number 3 would be a simple option, not necessarily the best, but simple.
If you use Spring DM, then you can easily accomplish number 2. It provides a means to inject your service beans into your UI Views, Pages, etc. You use a spring factory to create your views (as defined in your plugin.xml), which is configured via a Spring configuration, which is capable of injecting your services into the bean.
You may also be able to combine the technique used by the SpringExtensionFactory class along with DI to accomplish the same thing, without introducing another piece of technology. I haven't tried it myself so I cannot comment on the difficulty, although it is what I would try to do to bridge the gap between RCP and OSGi if I wasn't already using Spring DM.
I am new to OSGi and came across several examples about OSGi services.
For example:
import org.osgi.framework.*;
import org.osgi.service.log.*;
public class MyActivator implements BundleActivator {
public void start(BundleContext context) throws Exception {
ServiceReference logRef =
context.getServiceReference(LogService.class.getName());
}
}
My question is, why do you use
getServiceReference(LogService.class.getName())
instead of
getServiceReference("LogService")
If you use LogService.class.getName() you have to import the Interface. This also means that you have to import the package org.osgi.services.log in your MANIFEST.MF.
Isn't that completely counterproductive if you want to reduce dependencies to push loose coupling? As far as I know one advantage of services is that the service consumer doesn't have to know the service publisher. But if you have to import one specific Interface you clearly have to know who's providing it. By only using a string like "LogService" you would not have to know that the Interface is provided by org.osgi.services.log.LogService.
What am I missing here?
Looks like you've confused implementation and interface
Using the actual interface for the name (and importing the interface , which you'll end up doing anyway) reenforces the interface contract that services are designed around. You don't care about the implemenation of a LogService but you do care about the interface. Every LogService will need to implement the same interface, hence your use of the interface to get the service. For all you know the LogService is really a wrapper around SLF4J provided by some other bundle. All you see is the interface. That's the loose coupling you're looking for. You don't have to ship the interface with every implementation. Leave the interface it's own bundle and have multiple implementations of that interface.
Side note: ServiceTracker is usually easier to use, give it a try!
Added benefits: Using the interface get the class name avoids spelling mistakes, excessive string literals, and makes refactoring much easier.
After you've gotten the ServiceReference, your next couple lines will likely involve this:
Object logSvc = content.getService(logRef)
// What can you do with logSvc now?!? It's an object, mostly useless
// Cast to the interface ... YES! Now you need to import it!
LogSerivce logger = (LogService)logSvc;
logger.log(LogService.LOG_INFO, "Interfaces are a contract between implementation and consumer/user");
If you use the LogService, you're coupled to it anyway. If you write middleware you likely get the name parameterized through some XML file or via an API. And yes, "LogService" will fail terribly, you need to use the fully qualified name: "org.osgi.service.log.LogService". Main reason to use the LogService.class.getName() pattern is to get correct renaming when you refactor your code and minimize spelling errors. The next OSGi API will very likely have:
ServiceReference<S> getServiceReference(Class<S> type)
calls to increase type safety.
Anyway, I would never use these low level API unless you develop middleware. If you actually depend on a concrete class DS is infinitely simpler, and even more when you use it with the bnd annotations (http://enroute.osgi.org/doc/217-ds.html).
#Component
class Xyz implements SomeService {
LogService log;
#Reference
void setLog( LogService log) { this.log = log; }
public void foo() { ... someservice ... }
}
If you develop middleware you get the service classes usually without knowing the actual class, via a string or class object. The OSGi API based on strings is used in those cases because it allows us to be more lazy by not creating a class loader until the last moment in time. I think the biggest mistake we made in OSGi 12 years ago is not to include the DS concepts in the core ... :-(
You cannot use value "LogService"
as a class name to get ServiceReference, because you have to use fully qualified class name
"org.osgi.services.log.LogService".
If you import package this way:
org.osgi.services.log;resolution:=optional
and you use ServiceTracker to track services in BundleActivator.start() method I suggest to use "org.osgi.services.log.LogService" instead of LogService.class.getName() on ServiceTracker initializazion. In this case you'll not get NoClassDefFoundError/ClassNotFountException on bundle start.
As basszero mentioned you should consider to use ServiceTracker. It is fairly easy to use and also supports a much better programming pattern. You must never assume that a ServiceReference you got sometime in the past is still valid. The service the ServiceReference points to might have gone away. The ServiceTracker will automatically notify you when a service is registered or unregistered.
I am working in a project that has two main parts: a class library assembly and the main application. Both are using Castle Windsor for IoC and both manually setup their list of of components in code (to aid refactoring and prevent the need for a config file). Currently the main application has code like this:
public static void Main()
{
// Perform library IoC setup
LibraryComponent.Init();
// Perform application IoC setup
IoC.Register<IXyz, Abc>("abc");
// etc, etc, ...
// Start the application code ...
}
However the call to initialise the library doesn't seem like a good solution. What is the best way to setup a class library that uses an IoC container to decouple its internal components?
Edit:
Lusid proposed using a static method on each public component in the library that would in turn make the call to initialise. One possible way to make this a bit nicer would be to use something like PostSharp to do this in an aspect-oriented way. However I was hoping for something a bit more elegant ;-)
Lusid also proposed using the AppDomain.AssemblyLoad event to perform custom steps at load time, however I am really after a way to avoid the client assembly from requiring any setup code.
Thanks!
I'm not sure if I'm understanding exactly the problem you are trying to solve, but my first guess is that you are looking for a way to decouple the need to call the Init method from your main application.
One method I've used in the past is a static constructor on a static class in the class library:
static public class LibraryComponent {
static LibraryComponent() {
Init();
}
}
If you have multiple class libraries, and would like a quick and dirty way of evaluating all of them as they are loaded, here's a (kinda hairy) way:
[STAThread]
static void Main()
{
AppDomain.CurrentDomain.AssemblyLoad += new AssemblyLoadEventHandler(CurrentDomain_AssemblyLoad);
}
static void CurrentDomain_AssemblyLoad(object sender, AssemblyLoadEventArgs args)
{
IEnumerable<Type> types = args.LoadedAssembly.GetTypes()
.Where(t => typeof(IMyModuleInterface).IsAssignableFrom(t));
foreach (Type t in types)
{
doSomethingWithType(t);
}
}
The Where clause could be anything you want, of course. The code above would find any class deriving from IMyModuleInterface in each assembly that gets loaded into the current AppDomain, and then I can do something with it, whether it be registering dependencies, maintaining an internal list, whatever.
Might not be exactly what you are looking for, but hopefully it helps in some way.
You could have a registration module. Basically LibraryComponent.Init() function takes an IRegistrar to wire everything up.
The IRegistrar could basically have a function Register(Type interface, Type implementation). The implimentor would map that function back to their IOC container.
The downside is that you can't rely on anything specific to the container your using.
Castle Windsor actually has a concept called facilities that are basically just ways of wrapping standardised pieces of configuration. In this model, you would simply add the two facilities to the container and they would do the work.
Of course, this wouldn't really be better than calling a library routine to do the work unless you configured the facilities in a configuration file (consider binsor). If you are really allergic to configuration files, your current solution is probably the best.