Spring is perhaps the best of the component-based frameworks that emerged at the turn of the 21st century. It vastly improves the way that developers write and deliver infrastructure code in Java-based applications. Since its inception, Spring has been recognized as a leading framework for enterprise Java development. As an end-to-end application framework, Spring mirrors some of the Java EE capabilities , but it offers a combination of features and programming conventions you won't find elsewhere.
This article introduces Spring and its core programming philosophy and methodology: Inversion of control and dependency injection. You'll also get started with Spring annotations and a couple of hands-on coding examples.
Dependency injection and inversion of control
Spring's core idea is that instead of managing object relationships yourself, you offload them to the framework. Inversion of control (IOC) is the methodology used to manage object relationships. Dependency injection is the mechanism for implementing IOC. Since these two concepts are related but different, let's consider them more closely:
- Inversion of control (IOC) does just what its name says: it inverts the traditional hierarchy of control for fulfilling object relationships. Instead of relying on application code to define how objects relate to each other, relationships are defined by the framework. As a methodology, IOC introduces consistency and predictability to object relations, but it does require you, as the developer, to give up some fine-grained control.
- Dependency injection (DI) is a mechanism where the framework "injects" dependencies into your app. It's the practical implementation of IOC. Dependency injection hinges on polymorphism, in the sense that it allows the fulfillment of a reference type to change based on configurations in the framework. The framework injects variable references rather than having them manually fulfilled in application code.
Example #1: Spring dependency injection
Inversion of control and dependency injection are best understood by using them, so we'll start with a quick programming example.
Say you're modelling a car. If you're modeling in plain old Java, you might have an interface member on the Car
class to reference an Engine
interface, as shown in Listing 1.
Listing 1. Object relations in plain old Java
public Interface Engine() { ... }
public class Car {
private Engine engine;
public Engine getEngine() { ... }
public void setEngine(Engine engine) { ... }
}
Listing 1 contains an interface for an Engine
type, and a class for the concrete Car
type, which references the Engine
. (Note that in a real programming scenario these would be in separate files.) Now, when you're creating a Car
instance, you'd set the association as shown in Listing 2.
Listing 2. Creating a Car with the Engine interface
// ...
Car newCar = new Car();
Engine sixCylEngine = new InlineSixCylinderEngine();
newCar.setEngine(sixCylEngine );
// Do stuff with the car
Note that you create the Car
object first. You then create a new object that fulfills the Engine
interface, and assign it manually to the Car
object. That is how object associations work in plain old Java.
Modeling classes and objects in Spring
Now let's look at the same example in Spring. Here, you could do something like what's shown in Listing 3. You start with the Car
class, but in this case you add an annotation to it: @Inject
.
Listing 3. Example of using the @Inject annotation in Spring
public class Car {
@Inject
private Engine engine;
// ...
}
Using the @Inject
annotation (or @Autowired
, if you prefer) tells Spring to search the context and automatically inject an object into the reference, based on a set of rules.
Next, consider the @Component
annotation, shown in Listing 4.
Listing 4. @Component annotation
@Component
public class InlineSixCylinderEngine implements Engine{
//...
}
Annotating a class with @Component
tells Spring that it is available for fulfilling injections. In this case, the InlineSixCylEngine
would be injected because it is available and satisfies the interface requirement of the association. In Spring, this is called an "autowired" injection. (See below for more about Spring's @Autowired
annotation.)
Decoupling as a design principle
Inversion of control with dependency injection removes a source of concrete dependency from your code. Nowhere in the program is there a hard-coded reference to the Engine
implementation. This is an example of decoupling as a software design principle. Decoupling application code from implementation makes your code easier to manage and maintain. The application knows less about how its parts fit together, but it's much easier to make changes at any point in the application lifecycle.
Overview of the Spring Framework
Now that you've seen some Spring code, let's take an overview of the framework and its components. As you can see, the framework consists of four main modules, which are broken into packages. Spring gives you a fair amount of flexibility with the modules you'll use.
- Core container
- Core
- Bean
- Context
- Expression Language
- Aspect-oriented programming (AOP)
- AOP
- Aspects
- Instrumentation
- Data access and integration
- JDBC
- JPA/ORM
- JMS
- Transactions
- Web
- Web/REST
- Servlet
- Struts
Rather than cover everything here, let's get started with two of the more commonly used Spring features.
Starting up a new project: Spring Boot
We'll use Spring Boot to create an example project, which we'll use to demo Spring features. Spring Boot makes starting new projects much easier, as you'll see for yourself. To begin, take a look at the main class shown below. In Spring Boot, we can take a main class with a main()
method, and then choose to run it standalone, or package for deployment in a container like Tomcat.
Listing 5 has the outlines of our main class, which will live at the standard src/main/java/hello
location.
Listing 5. Main class with Spring Boot
package hello;
import org.springframework.boot.SpringApplication;
import org.springframework.boot.autoconfigure.SpringBootApplication;
@SpringBootApplication
public class Application {
public static void main(String[] args) {
SpringApplication.run(Application.class, args);
}
}
Note two things about the above code: First, all of the work is abstracted into the framework. The main class boots up the app, but it doesn't know anything about how the app works or delivers its functionality. Second, the SpringApplication.run()
does the actual job of booting the app and passing in the Application
class itself. Again, the work the app does is not apparent here.
The @SpringBootApplication
annotation wraps up a few standard annotations and tells Spring to look at the package where the main class exists for components. In our previous example, with the car and engine, this would allow Spring to find all classes annotated with @Component
and @Inject
. The process itself, called component scanning, is highly customizable.
You can build the app with the standard mvn clean install
, and you can run it with the Spring Boot goal (mvn spring-boot:run
). Before doing that, let's look at this application's pom.xml
file.
Listing 6. Starter pom.xml
<groupId>com.javaworld</groupId>
<artifactId>what-is-spring</artifactId>
<version>1.0.0</version>
<parent>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-parent</artifactId>
<version>2.1.3.RELEASE</version>
</parent>
<dependencies>
</dependencies>
<properties>
<java.version>1.8</java.version>
</properties>
<build>
<plugins>
<plugin>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-maven-plugin</artifactId>
</plugin>
</plugins>
</build>
Note two important features in the above code:
- The
parent
element relies on thespring-boot-starter-parent
project. This parent project defines a number of useful defaults, such as the default compiler level of JDK 1.8. For the most part, you can just trust that it knows what it's doing. As an example, you can omit the version number for many common dependencies, andSpringBootParent
will set the versions to be compatible. When you bump up the parent's version number, the dependency versions and defaults will also change. - The
spring-boot-maven-plugin
allows for the executable JAR/WAR packaging and in-placerun
(via themvn spring-boot:run
command).
Adding Spring Web as a dependency
So far, we've been able to use spring-boot
to limit how much work we put in to get an app up and running. Now let's add a dependency and see how quickly we can get something in a browser.
Listing 7. Adding Spring Web to a project
<dependency>
<groupId>org.springframework.boot</groupId>
<artifactId>spring-boot-starter-web</artifactId>
</dependency>
Now that we've got a basic project setup, we're ready for our two examples.
Example #2: Building RESTful endpoints with Spring Web
We've used spring-boot-starter-web
to bring in several dependencies that are useful for building web applications. Next we'll create a route handler for a URL path. Spring's web support is part of the Spring MVC (Model-View-Controller) module, but don't let that worry you: Spring Web has full and effective support for building RESTful endpoints, as well.
The class whose job it is to field URL requests is known as a controller, as shown in Listing 8.
Listing 8. Spring MVC REST controller
package hello;
import org.springframework.stereotype.Controller;
import org.springframework.ui.Model;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.bind.annotation.RequestMethod;
import org.springframework.web.bind.annotation.ResponseBody;
import org.springframework.web.bind.annotation.RequestParam;
@Controller
public class GreetingController {
@RequestMapping(value = "/hi", method = RequestMethod.GET)
public String hi(@RequestParam(name="name", required=false, defaultValue="JavaWorld") String name, Model model) {
return "Hello " + name;
}
}
The @Controller annotation
The @Controller
annotation identifies a class as a controller. A class marked as a controller is also automatically identified as a component class, which makes it a candidate for auto-wiring. Wherever this controller is needed, it will be plugged into the framework. In this case, we'll plug it into the MVC system to handle requests.
The controller is a specialized kind of component. It supports the @RequestMapping
and @ResponseBody
annotations that you see on the hi()
method. These annotations tell the framework how to map URL requests to the app.
At this point, you can run the app with mvn spring-boot:run
. When you hit the /hi
URL, you'll get a response like "Hello, JavaWorld."
Notice how Spring has taken the basics of autowiring components, and delivered a whole web framework. With Spring, you don't have to explicitly connect anything together!
The @Request annotations
The @RequestMapping
allows you to define a handler for a URL path. Options include defining the HTTP method you want, which is what we've done in this case. Leaving RequestMethod
off would instruct the program to handle all HTTP method types.
The @RequestParam
argument annotation allows us to map the request parameters directly into the method signature, including requiring certain params and defining default values as we've done here. We can even map a request body to a class with the @RequestBody
argument annotation.
REST and JSON response
If you are creating a REST endpoint and you want to return JSON from the method, you can annotate the method with @ResponseBody
. The response will then be automatically packaged as JSON. In this case you'll return an object from the method.
Example #3: Spring with JDBC
Now let's do something more interesting with our request handler: let's return some data from a database. For the purpose of this example, we'll use the H2 database. Thankfully, Spring Boot supports the in-memory H2 DB out of the box.