I started learning spring today and i have a question regarding what happens to the annotations when java files with annotations is compiled ?.
The reason i am asking this is because of the fundamental difference i see when we choose to use the xml approach vs the annotations approach , and what i think is the philosophy of spring. The way i understand is spring says that all your java classes can be simple pojo's and all the spring related config should be kept independent (Like xml file.)
In case of developing spring application using xml *.java files have no idea about spring container and are compiled in to .class without any spring related dependencies.
But now when we annotate the .java file and the file is compiled the compiled file now has all spring related dependencies hard baked in to it and no longer are your classes simple pojo's.
Is this correct ? I am not sure if i am missing some thing here.
Annotations can be considered as metadata of a class or its element (method, field, local variable...). When you put annotation, you don't implement any behaviour. You just give additional info on an element.
That way, Spring, which is in charge of instanciating its bean can collect the info with reflection (see also this site) and process it.
To conclude, your Spring beans still remain POJO and there is no difference with the XML way (...from that point of view) since Spring gets from annotations the information it would have got from XML .
I think you are right and your question is justifiable, that's the way how I think about it too.
Not only compiled code but also dependency on spring jars bother me. Once you use this annotations your resulting jar depends on spring library.
It's reasonable to store beans in model according to DDD but spring is some kind of infrastructure layer so I didn't like the dependency.
Even if you would use XML, it's useful for few placed to use attributes. E.g. #Required attribute which is useful to verify that linked bean was injected. So, I've decide to use constructor dependency injection to omit this attribute, see my article. I completely leave out the dependency on spring in the code.
You can probably find such mind hook for many annotation you want/force to use.
You can use annotations only for your configuration classes, without marking them actual bean classes. In such scenario if you not use spring you just not load configuration classes.
Related
I'm new to spring 5 and my question is why does spring use DI and IOC? I mean why do we have to write the beans in an XML (legacy) and then create it where we need it? Why don't we use one method instead that gives us that object, until we want to use this complex mechanism that happens in the spring container?
And another question is, does not reading XML slow down the program? Because we are reading from the hard disk anyway.
Note: It is true that we can use annotations, but for now I want to ask a question about reading from xml.
Spring IoC Container is the core of Spring Framework. It creates the objects, configures and assembles their dependencies, manages their entire life cycle. The Container uses Dependency Injection(DI) to manage the components that make up the application. It gets the information about the objects from a configuration file(XML) or Java Code or Java Annotations and Java POJO class. These objects are called Beans. Since the Controlling of Java objects and their lifecycle is not done by the developers, hence the name Inversion Of Control.
More on link HERE
As for your first part of the question.
why does spring use DI
To allow the developer to keep his code loose, and not entagle classes, it keeps your code clean.
In object oriented design, the amount of coupling refers to how much the design of one class depends on the design of another class. In other words, how often do changes in class A force related changes in class B? Tight coupling means the two classes often change together, loose coupling means they are mostly independent. In general, loose coupling is recommended because it's easier to test and maintain.
You may find this paper by Martin Fowler (PDF) helpful.
I mean why do we have to write the beans in an XML (legacy) and then create it where we need it
Note: We write the bean in XML and it is created when application starts when it looks at bean defintion, techinally you are never creating a bean, you are fetching only created bean from Spring Container(IOC) that Spring created for you when you started your application.
We are writing bean blueprint, or just bean, so that it can be constructed, placed in the Spring Container when the application starts, and then we have it at our disposal that we can fetch it using getBean method.
The whole point of "why", is because by default all beans are scoped as singleton, that means, when you fetch a bean, and do with it whatever you want, you do not worry about memory or anything, Spring takes care of the beans for you if they are scoped as a Singleton.
Second question:
And another question is, does not reading XML slow down the program? Because we are reading from the hard disk anyway.
There is no difference in performance between annotation or XML, it is just a different approach, I am not sure what you mean by "reading from hard disk", but one way or another you will have to configure your application, yes, many forums prefer to run away from XML, but in my honest opinion the only reason for that is because when you write a bad configuration in XML it is lot harder to find it compared to configuration in Java that will throw an exception.
XML, application.properties files require a redeployment of the application, while annotation and java configuration require recompilation of your project, so both of them have "flaws", but it is normal and quite understandable to me.
But in the end I believe that it is a matter of preference, I know personally quite a few people that combine annotations with XML configuration and they know lot more about Spring compared to me.
So in summary, it is pain to write beans and their configuration, same as you can write a class with methods without creating an interface for it since the result will be the same, but it will help you in the long run since you do not have to worry about memory or if you destroyed that bean or if you did not.
It would be nice that you read about
1.Lazy initialization of beans
2.Eager initialization of beans
3.Singleton scope of beans
4.Prototype scope of beans
I'm providing a Spring RestController as a part of a library. Currently I have reusable jaxrs services, but I need to make Spring Boot alternatives. One RestController for example has 2 dependencies: one is a service that I could see being a bean and the other a String property.
I'm wondering what is the idiomatic way to expect to get those dependencies from users consuming the library. I had a few ideas about how it might happen, but wasn't sure what was the right or at least best practice way to do it.
Should users construct the RestController manually using the constructor (not using dependency injection)? I actually couldn't even figure out how to do this such that the Spring Boot Application knew about it and didn't see it in guides, so I was assuming this isn't the normal way to provide RestControllers. I only wondered if this was the right way to go as dependency injection being used for a third party library class's dependencies seems like it could be hard to manage.
Should both be beans, with the String property being a named bean? I have this one working, but I'm wondering if consumers of the library having to provide beans that the library's RestController expects is tricky or a bad practice.
Should the simple service be a bean and the String injected via #Value?
Is there some alternative or better way?
I want to remove spring from a little library that really doesn't need it in favour of wiring the classes together in code. I feel it should be possible to auto-generate the code to stick the library together using the same logic Spring does to do it at runtime. Does code to do this already exist?
(Of course it might be more complicated for any AOP type stuff Spring is doing - but when it's just vanilla instantiation and autowiring it should be relatively simple?)
If you have a library containing Spring beans that need to be wired together before an application can use them, does it make sense to include any sort of bean configuration file in the JAR (such as the /META-INF directory)? The idea is to give the application the option of importing this into its master Spring context configuration.
There may be more than one way to wire these beans, so I could provide a bean configuration file for each of the standard ways in which you'd typically wire them together.
Or, do I force the application to wire these up explicitly?
If it helps, the specifics of my problem involve a library I created to encapsulate our product's persistence layer. It contains Service, DAO and model beans. The DAO implementations currently use Hibernate (this probably won't change). Some of the DAO implementations need different kinds of Strategy beans injected into them (database encryption logic), depending on the type of database we are deploying on (MySQL vs SQL Server, etc). So we have potentially a few different configuration scenarios. I could also provide datasource bean configurations, relying on property substitution at the app level to inject all the particulars needed by the datasource.
Thanks for your input!
In this case, it's a good idea to provide some beans files, either as examples for documentation purposes, or as fully-fledged files ready for including into a wider context.
If your beans' wiring can get complex, then you shouldn't really leave it entirely up to the library client to figure it out.
This is more of a documentation and education task, really.
We have a very comfortable setup using JPA through Spring/Hibernate, where we attach a PersistenceUnitPostProcessor to our entity manager factory, and this post processor takes a list of project names, scans the classpath for jars that contain that name, and adds those jar files for scanning for entities to the persistence unit, this is much more convenient than specifying in persistence.xml since it can take partial names and we added facilities for detecting the different classpath configurations when we are running in a war, a unit test, an ear, etc.
Now, we are trying to replace Spring with Seam, and I cant find a facility to accomplish the same hooking mechanism. One Solution is to try and hook Seam through Spring, but this solution has other short-comings on our environment. So my question is: Can someone point me to such a facility in Seam if exists, or at least where in the code I should be looking if I am planning to patch Seam?
Thanks.
If you're running in a Java EE container like JBoss 6 (and I really recommend so), all you need is to package your beans into a jar, place a META-INF/persistence.xml inside it and place the jar into your WAR or EAR package. All #Entity annotated beans inside the jar will be processed.
For unit-testing, you could point the <jar-file> element to the generated .class output directory and Hibernate will also pick the Entities. Or even configure during runtime using Ejb3Configuration.addAnnotatedClass;
#see http://docs.jboss.org/hibernate/entitymanager/3.6/reference/en/html/configuration.html