Develop Reference

Is each Play framework web request handled with a new dependency injected controller instance, but then what about static controller methods?

My questions are about the lifecycle of controllers in the Play framework for Java, if the controllers are stateful instances or stateless with static methods, and how to use dependency injection in the controller code.
Is each web request handled by a new instance of a Play controller class, i.e. can a controller store state in fields such as services injected into the controller constructor?
(where in the documentation is it explained?)
Has the Play framework changed since earlier versions (and if so, at what version?) regarding if controllers are stateful instances or stateless controllers with static methods?
Where can you see code examples about how the framework injects services into a controller instance when stateful controller is used and example of how to inject services into a static controller method?
Regarding the latter, i.e. injection into a static method I suppose that would either have to be a parameter to the method which the frameworks will add, or if not possible you maybe instead will have to use a service locator from within the method e.g. instantiate a Guice module class and then use "injector.getInstance" from within the static controller method.
This subject is touched in the section "Dependency injecting controllers" at the following page:
https://www.playframework.com/documentation/2.4.x/JavaDependencyInjection
However, it does not show with code how to actually inject services into a controller instance (but probably the same way as other "components" i.e. with #Inject annotation) and certainly it does not currently show how to use DI with a static controller method.
I am confused about these things because I have not found documentation being clear about my questions, and I have also read in a Play book (from 2013) that the controller methods should be programmed as stateless and the controller methods should be static.
However, when now using activator for generating a Play application for Java with the latest Play version (2.4.6) I can see that the generated Controller method (Application.index) is NOT static.
Also, at the following documentation page, the controller method is NOT static:
https://www.playframework.com/documentation/2.4.x/JavaActions
This is confusing, and since it is VERY fundamental to understand whether or not each request is handled by a Controller instance or not (i.e. if state can be used) I think this should be better documented at the page about Controller/Actions than the current documentation (the above linked page) which is not explaining it.
The documentation about dependency injection touches the subject about static and non-static methods at the section "Dependency injecting controllers" mentioning "static routes generator" but I think it should be better explained including code examples.
If someone in the Play team is reading this question, then please add some information to the above linked pages, for example please do mention (if my understanding is correct) that in previous versions of Play the controller methods were static and for those versions you should never store state in fields, but in later versions (beginning from version x?) each request is handled by an instance of a controller and can therefore use state (e.g. constructor parameters injected by the framework).
Please also provide code examples about injection used with static controller methods and injection into stateful controller instances with one instance per request.
The section "Component lifecycle" in the dependency injection page only mentions "components" but I think it should also be explicit about the controller lifecycle and its injection, since it is such a fundamental and important knowledge to communicate clearly to all developers to avoid bugs caused by misunderstandings about being stateful or not.

Is each web request handled by a new instance of a Play controller class, i.e. can a controller store state in fields such as services injected into the controller constructor? (where in the documentation is it explained?)
As far as I can tell, controllers are by default singleton objects. This is not clearly documented, but it is implied that controller instances are reused. See the migration guide for Playframework 2.4:
The injected routes generator also supports the # operator on routes, but it has a slightly different meaning (since everything is injected), if you prefix a controller with #, instead of that controller being directly injected, a JSR 330 Provider for that controller will be injected. This can be used, for example, to eliminate circular dependency issues, or if you want a new action instantiated per request.
Also, check this commend made by James Roper (Play core committer) about if controllers are singleton or not:
Not really - if using Guice, each time the controller is injected into something, a new instance will be created by default. That said, the router is a singleton, and so by association, the controllers it invokes are singleton. But if you inject a controller somewhere else, it will be instantiated newly for that component.
This suggests that the default is to reuse controller instances when responding to requests and, if you want a new action per request, you need to use the syntax described in the migration guide. But... since I'm more inclined to prove and try things instead of just believe, I've created a simple controller to check that statement:
package controllers
import play.api._
import play.api.mvc._
class Application extends Controller {
def index = Action {
println(this)
Ok(views.html.index("Your new application is ready."))
}
}
Doing multiple requests to this action prints the same object identity for all the requests made. But, if I use the # operator on my routes, I start to get different identities for each request. So, yes, controllers are (kind of) singletons by default.
Has the Play framework changed since earlier versions (and if so, at what version?) regarding if controllers are stateful instances or stateless controllers with static methods?
By default, Play had always advocated stateless controllers, as you can see at the project homepage:
Play is based on a lightweight, stateless, web-friendly architecture.
That had not changed. So, you should not use controllers' fields/properties to keep data that changes over time/requests. Instead, just use controllers' fields/properties to keep a reference to other components/services that are also stateless.
Where can you see code examples about how the framework injects services into a controller instance when stateful controller is used and example of how to inject services into a static controller method?
Regarding code examples, Lightbend templates repository is the place to go. Here are some examples that use dependency injection at the controllers level:
https://github.com/adrianhurt/play-api-rest-seed
https://github.com/knoldus/playing-reactive-mongo
https://github.com/KyleU/boilerplay
Dependency Injection with static methods is not supported, and that is why Playframework stills offers old apis to use with static methods. The rule of thumb here is: choose between DI and static methods. Trying to use both will just bring complexity to your application.

Ok, thank you marcospereira.
I have now also confirmed that you indeed get different instances (different toString values which can be printed/logged in a controller method) of the controller for each request.
For those who are interested, the solution (to get different instances of controller class for each request) is to use for example the following:
GET / #controllers.Application.index()
instead of the following:
GET / controllers.Application.index()
in the file "conf/routes"
AND to also use the following:
routesGenerator := InjectedRoutesGenerator
instead of the following:
routesGenerator := StaticRoutesGenerator
in the file "build.sbt"
Regarding the statement that Play has a "stateless" architecture:
Maybe I am wrong, but as far as I understand the terminology, the "stateless" means that the web server does not store any state between requests?
The word "stateless" does not mean that a controller instance can not use fields, e.g. injected into the constructor.
If an injected object is stored as a field in a controller, then that field is a "state" of the controller.
Therefore, even if you use "InjectedRoutesGenerator" and the "#" prefix to get "stateful" controller instances, that injected "state" is only stored within one request, so you can still say that the framework itself is "stateless" since the server does not store any state between multiple requests.
Please do correct me if I have misunderstood something about Play being stateless.

Why is Spring's ApplicationContext.getBean with Interface considered bad? [duplicate]

I asked a general Spring question: Auto-cast Spring Beans and had multiple people respond that calling Spring's ApplicationContext.getBean() should be avoided as much as possible. Why is that?
How else should I gain access to the beans I configured Spring to create?
I'm using Spring in a non-web application and had planned on accessing a shared ApplicationContext object as described by LiorH.
Amendment
I accept the answer below, but here's an alternate take by Martin Fowler who discusses the merits of Dependency Injection vs. using a Service Locator (which is essentially the same as calling a wrapped ApplicationContext.getBean()).
In part, Fowler states, "With service locator the application class asks for it [the service] explicitly by a message to the locator. With injection there is no explicit request, the service appears in the application class - hence the inversion of control.
Inversion of control is a common feature of frameworks, but it's something that comes at a price. It tends to be hard to understand and leads to problems when you are trying to debug. So on the whole I prefer to avoid it [Inversion of Control] unless I need it. This isn't to say it's a bad thing, just that I think it needs to justify itself over the more straightforward alternative."

I mentioned this in a comment on the other question, but the whole idea of Inversion of Control is to have none of your classes know or care how they get the objects they depend on. This makes it easy to change what type of implementation of a given dependency you use at any time. It also makes the classes easy to test, as you can provide mock implementations of dependencies. Finally, it makes the classes simpler and more focused on their core responsibility.
Calling ApplicationContext.getBean() is not Inversion of Control! While it's still easy to change what implemenation is configured for the given bean name, the class now relies directly on Spring to provide that dependency and can't get it any other way. You can't just make your own mock implementation in a test class and pass that to it yourself. This basically defeats Spring's purpose as a dependency injection container.
Everywhere you want to say:
MyClass myClass = applicationContext.getBean("myClass");
you should instead, for example, declare a method:
public void setMyClass(MyClass myClass) {
this.myClass = myClass;
}
And then in your configuration:
<bean id="myClass" class="MyClass">...</bean>
<bean id="myOtherClass" class="MyOtherClass">
<property name="myClass" ref="myClass"/>
</bean>
Spring will then automatically inject myClass into myOtherClass.
Declare everything in this way, and at the root of it all have something like:
<bean id="myApplication" class="MyApplication">
<property name="myCentralClass" ref="myCentralClass"/>
<property name="myOtherCentralClass" ref="myOtherCentralClass"/>
</bean>
MyApplication is the most central class, and depends at least indirectly on every other service in your program. When bootstrapping, in your main method, you can call applicationContext.getBean("myApplication") but you should not need to call getBean() anywhere else!

Reasons to prefer Service Locator over Inversion of Control (IoC) are:
Service Locator is much, much easier for other people to following in your code. IoC is 'magic' but maintenance programmers must understand your convoluted Spring configurations and all the myriad of locations to figure out how you wired your objects.
IoC is terrible for debugging configuration problems. In certain classes of applications the application will not start when misconfigured and you may not get a chance to step through what is going on with a debugger.
IoC is primarily XML based (Annotations improve things but there is still a lot of XML out there). That means developers can't work on your program unless they know all the magic tags defined by Spring. It is not good enough to know Java anymore. This hinders less experience programmers (ie. it is actually poor design to use a more complicated solution when a simpler solution, such as Service Locator, will fulfill the same requirements). Plus, support for diagnosing XML problems is far weaker than support for Java problems.
Dependency injection is more suited to larger programs. Most of the time the additional complexity is not worth it.
Often Spring is used in case you "might want to change the implementation later". There are other ways of achieving this without the complexity of Spring IoC.
For web applications (Java EE WARs) the Spring context is effectively bound at compile time (unless you want operators to grub around the context in the exploded war). You can make Spring use property files, but with servlets property files will need to be at a pre-determined location, which means you can't deploy multiple servlets of the same time on the same box. You can use Spring with JNDI to change properties at servlet startup time, but if you are using JNDI for administrator-modifiable parameters the need for Spring itself lessens (since JNDI is effectively a Service Locator).
With Spring you can lose program Control if Spring is dispatching to your methods. This is convenient and works for many types of applications, but not all. You may need to control program flow when you need to create tasks (threads etc) during initialization or need modifiable resources that Spring didn't know about when the content was bound to your WAR.
Spring is very good for transaction management and has some advantages. It is just that IoC can be over-engineering in many situations and introduce unwarranted complexity for maintainers. Do not automatically use IoC without thinking of ways of not using it first.

It's true that including the class in application-context.xml avoids the need to use getBean. However, even that is actually unnecessary. If you are writing a standalone application and you DON'T want to include your driver class in application-context.xml, you can use the following code to have Spring autowire the driver's dependencies:
public class AutowireThisDriver {
private MySpringBean mySpringBean;
public static void main(String[] args) {
AutowireThisDriver atd = new AutowireThisDriver(); //get instance
ClassPathXmlApplicationContext ctx = new ClassPathXmlApplicationContext(
"/WEB-INF/applicationContext.xml"); //get Spring context
//the magic: auto-wire the instance with all its dependencies:
ctx.getAutowireCapableBeanFactory().autowireBeanProperties(atd,
AutowireCapableBeanFactory.AUTOWIRE_BY_TYPE, true);
// code that uses mySpringBean ...
mySpringBean.doStuff() // no need to instantiate - thanks to Spring
}
public void setMySpringBean(MySpringBean bean) {
this.mySpringBean = bean;
}
}
I've needed to do this a couple of times when I have some sort of standalone class that needs to use some aspect of my app (eg for testing) but I don't want to include it in application-context because it is not actually part of the app. Note also that this avoids the need to look up the bean using a String name, which I've always thought was ugly.

One of the coolest benefits of using something like Spring is that you don't have to wire your objects together. Zeus's head splits open and your classes appear, fully formed with all of their dependencies created and wired-in, as needed. It's magical and fantastic.
The more you say ClassINeed classINeed = (ClassINeed)ApplicationContext.getBean("classINeed");, the less magic you're getting. Less code is almost always better. If your class really needed a ClassINeed bean, why didn't you just wire it in?
That said, something obviously needs to create the first object. There's nothing wrong with your main method acquiring a bean or two via getBean(), but you should avoid it because whenever you're using it, you're not really using all of the magic of Spring.

The motivation is to write code that doesn't depend explicitly on Spring. That way, if you choose to switch containers, you don't have to rewrite any code.
Think of the container as something is invisible to your code, magically providing for its needs, without being asked.
Dependency injection is a counterpoint to the "service locator" pattern. If you are going to lookup dependencies by name, you might as well get rid of the DI container and use something like JNDI.

Using #Autowired or ApplicationContext.getBean() is really the same thing. In both ways you get the bean that is configured in your context and in both ways your code depends on spring.
The only thing you should avoid is instantiating your ApplicationContext. Do this only once! In other words, a line like
ApplicationContext context = new ClassPathXmlApplicationContext("AppContext.xml");
should only be used once in your application.

One of Spring premises is avoid coupling. Define and use Interfaces, DI, AOP and avoid using ApplicationContext.getBean() :-)

One of the reasons is testability. Say you have this class:
interface HttpLoader {
String load(String url);
}
interface StringOutput {
void print(String txt);
}
#Component
class MyBean {
#Autowired
MyBean(HttpLoader loader, StringOutput out) {
out.print(loader.load("http://stackoverflow.com"));
}
}
How can you test this bean? E.g. like this:
class MyBeanTest {
public void creatingMyBean_writesStackoverflowPageToOutput() {
// setup
String stackOverflowHtml = "dummy";
StringBuilder result = new StringBuilder();
// execution
new MyBean(Collections.singletonMap("https://stackoverflow.com", stackOverflowHtml)::get, result::append);
// evaluation
assertEquals(result.toString(), stackOverflowHtml);
}
}
Easy, right?
While you still depend on Spring (due to the annotations) you can remove you dependency on spring without changing any code (only the annotation definitions) and the test developer does not need to know anything about how spring works (maybe he should anyway, but it allows to review and test the code separately from what spring does).
It is still possible to do the same when using the ApplicationContext. However then you need to mock ApplicationContext which is a huge interface. You either need a dummy implementation or you can use a mocking framework such as Mockito:
#Component
class MyBean {
#Autowired
MyBean(ApplicationContext context) {
HttpLoader loader = context.getBean(HttpLoader.class);
StringOutput out = context.getBean(StringOutput.class);
out.print(loader.load("http://stackoverflow.com"));
}
}
class MyBeanTest {
public void creatingMyBean_writesStackoverflowPageToOutput() {
// setup
String stackOverflowHtml = "dummy";
StringBuilder result = new StringBuilder();
ApplicationContext context = Mockito.mock(ApplicationContext.class);
Mockito.when(context.getBean(HttpLoader.class))
.thenReturn(Collections.singletonMap("https://stackoverflow.com", stackOverflowHtml)::get);
Mockito.when(context.getBean(StringOutput.class)).thenReturn(result::append);
// execution
new MyBean(context);
// evaluation
assertEquals(result.toString(), stackOverflowHtml);
}
}
This is quite a possibility, but I think most people would agree that the first option is more elegant and makes the test simpler.
The only option that is really a problem is this one:
#Component
class MyBean {
#Autowired
MyBean(StringOutput out) {
out.print(new HttpLoader().load("http://stackoverflow.com"));
}
}
Testing this requires huge efforts or your bean is going to attempt to connect to stackoverflow on each test. And as soon as you have a network failure (or the admins at stackoverflow block you due to excessive access rate) you will have randomly failing tests.
So as a conclusion I would not say that using the ApplicationContext directly is automatically wrong and should be avoided at all costs. However if there are better options (and there are in most cases), then use the better options.

The idea is that you rely on dependency injection (inversion of control, or IoC). That is, your components are configured with the components they need. These dependencies are injected (via the constructor or setters) - you don't get then yourself.
ApplicationContext.getBean() requires you to name a bean explicitly within your component. Instead, by using IoC, your configuration can determine what component will be used.
This allows you to rewire your application with different component implementations easily, or configure objects for testing in a straightforward fashion by providing mocked variants (e.g. a mocked DAO so you don't hit a database during testing)

Others have pointed to the general problem (and are valid answers), but I'll just offer one additional comment: it's not that you should NEVER do it, but rather that do it as little as possible.
Usually this means that it is done exactly once: during bootstrapping. And then it's just to access the "root" bean, through which other dependencies can be resolved. This can be reusable code, like base servlet (if developing web apps).

There is another time when using getBean makes sense. If you're reconfiguring a system that already exists, where the dependencies are not explicitly called out in spring context files. You can start the process by putting in calls to getBean, so that you don't have to wire it all up at once. This way you can slowly build up your spring configuration putting each piece in place over time and getting the bits lined up properly. The calls to getBean will eventually be replaced, but as you understand the structure of the code, or lack there of, you can start the process of wiring more and more beans and using fewer and fewer calls to getBean.

I've only found two situations where getBean() was required:
Others have mentioned using getBean() in main() to fetch the "main" bean for a standalone program.
Another use I have made of getBean() are in situations where an interactive user configuration determines the bean makeup for a particular situation. So that, for instance, part of the boot system loops through a database table using getBean() with a scope='prototype' bean definition and then setting additional properties. Presumably, there is a UI that adjusts the database table that would be friendlier than attempting to (re)write the application context XML.

however, there are still cases where you need the service locator pattern.
for example, i have a controller bean, this controller might have some default service beans, which can be dependency injected by configuration.
while there could also be many additional or new services this controller can invoke now or later, which then need the service locator to retrieve the service beans.

You should to use: ConfigurableApplicationContext instead of for ApplicationContext

struts2 conversation and validation

I am working on a struts2 project that has interdependent forms.
I found struts2-conversation, stepped through their simple-example
and understood the conversation mechanism this far (please correct me if I got something wrong):
The Controller is mapped in the struts.xml
It holds the serializable ConversationContext and the Storing-Service
The ConversationContext holds POJOs mapped on forms by naming convention
Now my question is where to put the validation?
In this structure the controller is only one extending ConversationSupport and thereby ActionSupport supplying the validate, prepare and addField- & ActionError methods.
But validating within the controller would mean to validate the whole context, which does not really serve the issue.
I tried validation through annotation within the POJOs, within the context as described above which gives me some NullPointerException as if the context wasn't flushed and I think the xml-validation approach of struts2 is just too stiff. (btw how to let the generated javascripts be minified before being served? And why is there so many options?)
Mark's conversation-interceptor approach had similar problems coming up which's workarounds I didn't really get. Maybe you can help me there.

If you would like to use annotations on your model classes, it works fine with the plugin (as do the other validation approaches).
To validate your model, add #VisitorFieldValidator to the getModel() method in your controller. In the example app, you would then also add #VisitorFieldValidator to the getContact() and getPreferences() methods. Then you can use the validation annotations on the fields you wish to validate.
The service in the example is just there as a simple example of using an injected service in a Struts2 controller and how that can integrate easily with the conversation framework, but it is not directly related or needed (and I would recommend using either Spring, Guice, or CDI for dependency injection in the real world).
The ConversationContext class is intended mostly for internal use by the framework. You should be able to avoid interacting with it by using the annotations and conventions. Unless you simply wish to be adventurous.
To use XML validation in the example app, you would have to change the package name to remove the "struts2" word in order for the Struts2 resource loading tool to load the XML.

Get application components without #Autowired

How would you extract something prior 2.5 version from .xml config? It bothers me because if #Autowired is removed from my arsenal I would not really know what to do.
Say I want to use some DAO implementation.
In service class I usually write:
#Autowired
someDaoInterface generalDao;
Then I typically call
generalDao.someInterfaceMethod(someParam param);
How would I extract implementation from config in Spring 2.0 to use this method?
Is it as dumb as just: new ApplicationContext(pathToXml) and then use .getBean or there is other way?
Why do I ask for taking bean out from configuration file?
Because in Spring MVC how can you perform your logic without getting beans out from the application context.
If you have #Controller handler then you need to make calls to the service classes' methods? So they should be somehow retrieved from the context and the only way so far is using #Autowired? Then I would also want to populate Service classes as I stated in previous example with DAO classes and they also need to be retrieved from the application context, so I would be able to write logic for service classes themself. How would people do it in the past?
I see the #Autowired as the only mean of taking something out, not because it is convenient to wire automatically - I am perfectly ok with XML.

You still have option to wire it explicitely via property or constructor parameter. (Anyway, autowired is not going to work if there is ambiguity in your container )
Of course, you can use application context and getBean() in your java code, but it violates DI pattern and makes all the spring stuff useless. Purpose of DI is to decouple your business loginc from implementation details - it's not business logic it's how and where it dependencies come from. Dependencies are just there.
By using ApplicationContext.getBean() you are breaking this pattern, and introduce dependency to:
spring itself
your configuration names
After you done this, you can as well drop use of DI and spring because you just voided all the advandages DI is providing to you. (BTW, #Autowired also introduces dependency to spring, and violates DI pattern, it also implies that there is only one instance available)
Also, answer is: in ideal case there shall be no reference to spring in your code at all.
No imports, no annotations - just interfaces of collaborating entities.

Swing with Guice

I'm already using Guice for the data model of my app, and so far I'm quite happy with it. However, the GUI part is about to become a Big Ball of Mud. I find it hard to use Guice here, because Swing components and models are tightly coupled, and often force a certain initialization order.
My application consists basically of a header with a lot of filters, a central and quite complex JTree component, and a lot of actions (e.g. from a JPopup), menus, dialogs, wizards etc. The main problem is that I have a lot of coupling between the components and actions (e.g. complicated validations, tree updates...). Could you give me some advice how to structure that GUI with Guice?
I'm aware of libs like GUTS, but the documentation is really thin, I'd rather avoid to add another dependency to my project and to learn another API (e.g. I don't know the Swing Application Framework).

I'd rather suggest a proper MVC, even better Presentation Model - View - Controller. Separate your code properly and Guice will fit in naturally. For example:
View classes should have a building part which draws the static content (labels, the tree, buttons, etc) and updating code which reacts to changes in the Presentation Model. All the action listeners should invoke some code on the controller. Both the Presentation Model and the controller should be injected by Guice, like all the other dependencies.
This organization would allow for easy testing with replacing the View with some testing code which will listen to changes in the Presentation Model and invoke actions on the controller.

I would advise to look at Guts-GUI.
It is a Swing UI framework based on Guice dependency injection model.

I think the problem is probably the overall architecture. You probably want to see if you can refactor the application to be simpler and more modular. Like Boris recommended, I would also suggest using the Presentation Model pattern - search for Martin Fowler and Karsten Lentzsch - and the JGoodies library.
For the problem with Actions, see how the Swing Application Framework and Netbeans Plaform handle them. For example, if an Action is used in both a view and a menu, you might want to make it available through a global map.

We have recently started on GUICE with swing. Here's what we have done which may be of help to you.
a. When you want to inject a model into a table/tree you could inject a model provider instead and do a provider.get() to get the model in the contructor.
for example
public class Mytable extends JTable {
public Mytable(Provider<MytableModel> modelProvider) {
this.setModel(modelProvider.get());
}
}
b. You could make a model generic and make use of the same model in different tables.
c. The model could have a handle to datasource which could be injected by a factory where necessary with assisted inject. Number of examples available in stack overflow.
d. Your model could be a flexible data structure which uses model to return the relevant cell/leaf/other value.
e. When you have actions, you inject actions in your component builders and attach them to relevant components.
f. Use #Singleton annotation where you have a shared object such as model injected in 2 different objects, such as action and the component.]
g. Use custom scope when you need to use number of instance sets of objects. Custom scopes are great for this as it keeps the code really clean.
Hope the above helps...