java - 为什么我需要在 Controller 中使用 spring 注解？

Question

我是新手，也是 Spring 的新手。开始在项目中工作，如果我只是像这样创建 Controller 类:

package controllers;

import play.mvc.Controller;
import play.mvc.Result;

public class Users extends Controller {

    public Result login(){
        return ok(views.html.login.render());
    }

}

我遇到了这个异常:

[NoSuchBeanDefinitionException: No qualifying bean of type [controllers.Users] is defined]

但是，如果我在顶层插入 spring 的这个注释:

package controllers;

import play.mvc.Controller;
import play.mvc.Result;

@org.springframework.stereotype.Controller
public class Users extends Controller {

    public Result login(){
        return ok(views.html.login.render());
    }

}

页面已呈现。但我不知道为什么会发生这种情况。我想只在必要时使用 Spring ，充分发挥其作用。所以我想知道我在 Controller 上使用这个注释是否正确。

编辑:

import configuration.WebAppConfig;
import org.springframework.context.annotation.AnnotationConfigApplicationContext;
import org.springframework.context.annotation.Configuration;
import org.springframework.context.annotation.Import;
import play.Application;
import play.GlobalSettings;

/**
 * Application wide behaviour. We establish a Spring application context for the dependency injection system and
 * configure Spring Data.
 */
public class Global extends GlobalSettings {

    /**
     * The name of the persistence unit we will be using.
     */
    static final String DEFAULT_PERSISTENCE_UNIT = "default";

    /**
     * Declare the application context to be used - a Java annotation based application context requiring no XML.
     */
    final private AnnotationConfigApplicationContext ctx = new AnnotationConfigApplicationContext();

    /**
     * Sync the context lifecycle with Play's.
     */
    @Override
    public void onStart(final Application app) {
        super.onStart(app);

        // AnnotationConfigApplicationContext can only be refreshed once, but we do it here even though this method
        // can be called multiple times. The reason for doing during startup is so that the Play configuration is
        // entirely available to this application context.
        ctx.register(SpringDataJpaConfiguration.class);
        ctx.scan("controllers", "models");
        ctx.refresh();

        // This will construct the beans and call any construction lifecycle methods e.g. @PostConstruct
        ctx.start();

    }

    /**
     * Sync the context lifecycle with Play's.
     */
    @Override
    public void onStop(final Application app) {
        // This will call any destruction lifecycle methods and then release the beans e.g. @PreDestroy
        ctx.close();

        super.onStop(app);
    }

    /**
     * Controllers must be resolved through the application context. There is a special method of GlobalSettings that we
     * can override to resolve a given controller. This resolution is required by the Play router.
     */
    @Override
    public <A> A getControllerInstance(Class<A> aClass) {
        return ctx.getBean(aClass);
    }

    /**
     * This configuration establishes the Spring context which in our case is defined in the "other" project.
     */
    @Configuration
    @Import(WebAppConfig.class)
    public static class SpringDataJpaConfiguration {
        // At the moment this class is just a entry point for the "other project" Spring context config, AppContext
    }

}

编辑2:伙计们，现在我记得这个项目中的一些东西，我们有 2 个项目，一个完全基于 java 和 spring 开发，另一个是 webapp。我们正在通过 play 导入从第一个项目构建的 .jar 。我们这样做是为了不必维护两个不同的模型。

Answer 1

在您的类Global中，您实际上告诉 Playframework 从 Spring 应用程序上下文解析 Controller :

@Override
public <A> A getControllerInstance(Class<A> aClass) {
    return ctx.getBean(aClass);
}

在这种情况下，很明显，您必须在 Controller 上使用 Spring 注释，否则它们将不会添加到应用程序上下文中。如果您不想从应用程序上下文解析 Controller ，请删除该方法。但请记住，您将无法在 Play Controller 中注入(inject) Spring 组件，因为它们不受 Spring 管理。