问题描述
我一直在阅读和，但是应该何时使用它们并不完全清楚。
I've been reading the API documentation of Instance<T> and Provider<T>, but it isn't completely clear when they should be used.
以下方法之间的区别是什么？
What's the difference between the following approaches?
@Inject
MyBean bean;
@Inject
Instance<MyBean> bean;
@Inject
Provider<MyBean> bean;
推荐答案
提供商< T> 是一个JSR-330接口，由CDI接口扩展 Instance< T> 。
Provider<T> is a JSR-330 interface which is extended by the CDI interface Instance<T>.
注入 MyBean ，当没有匹配的bean或多个匹配的bean时，你的应用程序将在启动期间抛出异常。
Injecting MyBean, your application will throw an exception during startup when there is no matching bean or more than one matching bean.
注入实例< MyBean> ，将bean解析委托给应用程序：您可以遍历所有候选bean并且 select（）你想要的那个或者叫 isUnsatisfied（）并决定当没有匹配的bean时该怎么办。
Injecting Instance<MyBean>, bean resolution is delegated to the application: you can iterate over all candidate beans and select() the one you want or call isUnsatisfied() and decide what to do when there is no matching bean.
对于 @Dependent 范围的bean，调用 Instance.get（）将为每个bean创建一个新实例调用，当你不再需要它时，你应该为每个这样的实例调用 Instance.destroy（t）。
For beans with @Dependent scope, calling Instance.get() will create a new instance for each invocation, and you should invoke Instance.destroy(t) for each such instance when you no longer need it.
提供商只有 get（）方法，但没有 des troy（）或 select（）并且不支持迭代。在CDI环境中，对于 Provider< T> 解决的任何用例，最好使用 Instance< T> 代替。
Provider just has the get() method, but no destroy() or select() and does not support iteration. In a CDI environment, for any use case addressed by Provider<T>, you had better use Instance<T> instead.
这篇关于何时应该实例< T>和提供者< T>用于在CDI注入豆类？的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！