Struts2 是 Java 中开发 Web 应用程序的著名框架之一。最近我写了很多 Struts2 教程，在这篇文章中，我列出了一些重要的 Struts2 面试问题 并附上了答案，希望能帮助你在面试中。
Struts2 面试问题
- 什么是 Struts2？
- Struts1 和 Struts2 有什么区别，或者说 Struts2 相对于 Struts1 有哪些优势？
- Struts2 的核心组件有哪些？
- 在 Struts2 中什么是拦截器？
- Struts2 拦截器实现了哪种设计模式？
- 在 Struts2 中创建 Action 类的不同方式有哪些？
- Struts2 的 Action 和拦截器是线程安全的吗？
- 在 Struts2 中哪个类是前端控制器？
- Struts2 中拦截器的好处是什么？
- 什么是 ValueStack 和 OGNL？
- 列举一些在Struts2中引入的有用的注解。
- 提供一些您使用过的重要的Struts2常量。
- 在Struts2中的动作映射中，命名空间的作用是什么？
- 哪个拦截器负责将请求参数映射到动作类Java Bean属性？
- 哪个拦截器负责i18n支持？
- 在我们的动作类中使用Action接口和ActionSupport类有什么区别，您更喜欢使用哪一个？
- 我们如何在动作类中获取Servlet API的Request、Response、HttpSession等对象？
- execAndWait拦截器的作用是什么？
- token拦截器在Struts2中的作用是什么？
- 我们如何在Struts2应用程序中集成log4j？
- 不同的Struts2标签有哪些？我们如何使用它们？
- 在Struts2中，自定义类型转换器是什么？
- 我们如何编写自己的拦截器并将其映射到动作？
- 拦截器的生命周期是什么？
- 什么是拦截器栈？
- 什么是struts-default包，它有什么好处？
- Struts2操作URI的默认后缀是什么，我们如何更改它？
- 结果页面的默认位置是什么，我们如何更改它？
- 如何在Struts2应用程序中上传文件？
- 在开发Struts2应用程序时应遵循哪些最佳实践？
- 如何处理Struts2应用程序中抛出的异常？
Struts2面试问题与答案
-
什么是Struts2？
Apache Struts2是用于在Java中构建Web应用程序的开源框架。Struts2基于OpenSymphony WebWork框架进行了很大改进。它相对于Struts1有很大的提升，使其更加灵活、易于使用和扩展。Struts2的核心组件包括Action、拦截器和结果页面。Struts2提供了许多创建Action类并通过struts.xml或注解配置它们的方法。我们可以为常见任务创建自己的拦截器。Struts2配备有许多标签，并使用OGNL表达式语言。我们可以创建自己的类型转换器来呈现结果页面。结果页面可以是JSP和FreeMarker模板。
-
Struts1和Struts2有哪些区别？或者说Struts2比Struts1更好在哪些方面？
Struts2旨在克服Struts1的缺点，并使其更加灵活、可扩展。一些明显的区别包括：
组件 Struts1 Struts2 Action类 Struts1的Action类被强制扩展一个抽象类，这使得它不可扩展。 Struts2的Action类灵活，我们可以通过实现Action接口、扩展ActionSupport类或者只是拥有execute()方法来创建它们。 线程安全性 Struts1的Action类是单例的，不是线程安全的，这使得开发者需要额外小心以避免多线程带来的任何副作用。 Struts2的Action类每次请求都会被实例化，因此没有多线程问题，使得它们是线程安全的。 Servlet API耦合 Struts1的API与Servlet API紧密耦合，请求和响应对象被传递给Action类的execute()方法。 Struts2的API与Servlet API松散耦合，自动将表单bean数据映射到Action类的java bean属性中，这是我们经常使用的。然而，如果我们需要引用Servlet API类，可以使用*Aware接口。 测试 由于Servlet API的耦合，Struts1的Action类很难测试。 Struts2的Action类像普通的java类一样，我们可以通过实例化它们并设置它们的属性来轻松地测试它们。 请求参数映射 Struts1要求我们创建ActionForm类来保存请求参数，并且我们需要在struts配置文件中对其进行配置。 Struts2请求参数映射是动态进行的，我们只需要在Action类中拥有java bean属性或者实现ModelDriven接口来提供用于映射的java bean类名即可。 标签支持 Struts1使用JSTL标签，因此受限。 Struts2使用OGNL并提供不同类型的UI、控制和数据标签。它更加灵活、易于使用。 验证 Struts1支持通过手动validate()方法进行验证 Struts2支持手动验证以及验证框架集成。 视图渲染 Struts1使用标准的JSP技术将bean值提供给JSP页面进行视图渲染。 Struts2使用ValueStack来存储请求参数和属性，我们可以使用OGNL和Struts2标签来访问它们。 模块支持 Struts1的模块设计复杂，看起来像是独立的项目 Struts2为包提供“命名空间”配置，以便轻松创建模块。 -
Struts2的核心组件有哪些？
Struts2的核心组件包括：
- Action类
- 拦截器
- 结果页面，JSP或FreeMarker模板
- ValueStack、OGNL和标签库
-
在Struts2中，拦截器是什么？
拦截器是Struts2框架的支柱。Struts2拦截器负责框架的大部分处理工作，例如将请求参数传递给Action类，使Servlet API请求、响应、会话可用于Action类，验证，i18n支持等。 ActionInvocation负责封装Action类和拦截器，并按顺序触发它们。在ActionInvocation中最重要的方法是invoke()方法，该方法跟踪拦截器链并调用下一个拦截器或动作。这是Java EE框架中责任链模式的最佳示例之一。
-
Struts2拦截器实现了哪种设计模式？
Struts2 interceptors are based on intercepting filters design pattern. The invocation of interceptors in interceptor stack closely resembles Chain of Responsibility design pattern.
Struts2 provide different ways to create action classes.
1. By implementing Action interface
2. Using Struts2 @Action annotation
3. By extending ActionSupport class
4. Any normal java class with execute() method returning String can be configured as Action class.
Struts2 Action classes are thread safe because an object is instantiated for every request to handle it. Struts2 interceptors are singleton classes and a new thread is created to handle the request, so it's not thread safe and we need to implement them carefully to avoid any issues with shared data.
`org.apache.struts2.dispatcher.ng.filter.StrutsPrepareAndExecuteFilter` is the Front Controller class in Struts2 and every request processing starts from this class. Earlier versions of Struts2 uses `org.apache.struts2.dispatcher.FilterDispatcher` as Front Controller class.
Some of the benefits of interceptors are:
- Interceptor plays a crucial role in achieving high level of separation of concerns.
- Struts2 interceptors are configurable, we can configure it for any action we want.
- We can create our own custom interceptors to perform some common tasks such as request params logging, authentication etc. This helps us in taking care of common tasks at a single location, achieving low maintenance cost.
- We can create interceptors stack to use with different actions.
ValueStack is the storage area where the application data is stored by Struts2 for processing the client requests. The data is stored in `ActionContext` objects that use ThreadLocal to have values specific to the particular request thread. Object-Graph Navigation Language (OGNL) is a powerful Expression Language that is used to manipulate data stored on the ValueStack. As you can see in architecture diagram, both interceptors and result pages can access data stored on ValueStack using OGNL.
Some of the important annotations introduced in Struts2 are:
1. @Action to create action class
2. @Actions to configure single class for multiple actions
3. @Namespace and @Namespaces for creating different modules
4. @Result for result pages
5. @ResultPath for configuring result pages location
Some of the Struts2 constants that I have used are:
1. **struts.devMode** to run our application in development mode. This mode does reload properties files and provides extra logging and debugging feature. It's very useful while developing our application but we should turn it off while moving our code to production.
2. **struts.convention.result.path** to configure the location of result pages. By default Struts2 look for result pages at {WEBAPP-ROOT}/{Namespace}/ and we can change the location with this constant.
3. **struts.custom.i18n.resources** to define global resource bundle for i18n support.
4. **struts.action.extension** to configure the URL suffix to for Struts2 application. Default suffix is .action but sometimes we might want to change it to .do or something else.
We can configure above constants in the struts.xml file like below.
```
<constant name="struts.devMode" value="true"></constant>
<constant name="struts.action.extension" value="action,do"></constant>
<constant name="struts.custom.i18n.resources" value="global"></constant>
<constant name="struts.convention.result.path" value="/"></constant>
```
Struts2 namespace configuration allows us to create modules easily. We can use namespace to separate our action classes based on their functionality, for example admin, user, customer etc.
`com.opensymphony.xwork2.interceptor.ParametersInterceptor` interceptor is responsible for mapping request parameters to the Action class java bean properties. This interceptor is configured in struts-default package with name "params". This interceptor is part of basicStack and defaultStack interceptors stack.
`com.opensymphony.xwork2.interceptor.I18nInterceptor` interceptor is responsible for i18n support in Struts2 applications. This interceptor is configured in struts-default package with name "i18n" and it's part of i18nStack and defaultStack.
We can implement Action interface to create our action classes. This interface has a single method execute() that we need to implement. The only benefit of using this interface is that it contains some constants that we can use for result pages, these constants are SUCCESS, ERROR, NONE, INPUT and LOGIN. ActionSupport class is the default implementation of Action interface and it also implements interfaces related to Validation and i18n support. ActionSupport class implements Action, Validateable, ValidationAware, TextProvider and LocaleProvider interfaces. We can override the validate() method of ActionSupport class to include field level validation login in our action classes. Depending on the requirements, we can use any of the approaches to creating Struts 2 action classes, my favorite is ActionSupport class because it helps in writing validation and i18n logic easily in action classes.
Struts2 action classes don't provide direct access to Servlet API components such as Request, Response, and Session. However, sometimes we need these access in action classes such as checking HTTP method or setting cookies in response. That's why Struts2 API provides a bunch of \*Aware interfaces that we can implement to access these objects. Struts2 API uses dependency injection to inject Servlet API components in action classes. Some of the important Aware interfaces are SessionAware, ApplicationAware, ServletRequestAware, and ServletResponseAware. You can read more about them in How to get [Servlet API Session in Struts2 Action Classes](/community/tutorials/get-servlet-session-request-response-context-attributes-struts-2-action) tutorial.
Struts2 provides execAndWait interceptor for long running action classes. We can use this interceptor to return an intermediate response page to the client and once the processing is finished, final response is returned to the client. This interceptor is defined in the struts-default package and implementation is present in `ExecuteAndWaitInterceptor` class. Check out [Struts2 execAndWait interceptor example](/community/tutorials/struts2-execandwait-interceptor-example-for-long-running-actions) to learn more about this interceptor and how to use it.
One of the major problems with web applications is the double form submission. If not taken care, double form submission could result in charging double amount to customer or updating database values twice. We can use a token interceptor to solve the double form submission problem. This interceptor is defined in the struts-default package but it's not part of any interceptor stack, so we need to include it manually in our action classes. Read more at [Struts2 token interceptor](/community/tutorials/struts2-token-interceptor-example) example.
Struts2 provides easy integration of log4j API for logging purpose, all we need to have is log4j configuration file in the WEB-INF/classes directory. You can check out the sample project at [Struts2 Log4j integration](/community/tutorials/struts2-and-log4j-integration-example-project).
Struts2 provides a lot of custom tags that we can use in result pages to create views for client request. These tags are broadly divided into three categories- Data tags, Control tags and UI tags. We can use these tags by adding these in JSP pages using taglib directive.
```
<%@ taglib uri="/struts-tags" prefix="s" %>
```
Some of the important Data tags are property, set, push, bean, action, include, i18n and text tag. Read more at [Struts2 Data Tags](/community/tutorials/struts-2-data-tags-example-tutorial). Control tags are used for manipulation and navigation of data from a collection. Some of the important Control tags are if-elseif-else, iterator, append, merge, sort, subset and generator tag. Read more at [Struts2 Control Tags](/community/tutorials/struts-2-control-tags-example-tutorial). Struts2 UI tags are used to generate HTML markup language, binding HTML form data to action classes properties, type conversion, validation, and i18n support. Some of the important UI tags are form, textfield, password, textarea, checkbox, select, radio and submit tag. Read more about them at [Struts2 UI Tags](/community/tutorials/struts-2-ui-tags-form-checkbox-radio-select-submit).
Struts2 support OGNL expression language and it performs two important tasks in Struts 2 – data transfer and type conversion. OGNL is flexible and we can easily extend it to create our own custom converter class. Creating and configuring custom type converter class is very easy. The first step is to fix the input format for the custom class. The second step is to implement the converter class. Type converter classes should implement `com.opensymphony.xwork2.conversion.TypeConverter` interface. Since in web application, we always get the request in form of String and send the response in the form of String, Struts 2 API provides a default implementation of TypeConverter interface, StrutsTypeConverter. StrutsTypeConverter contains two abstract methods – convertFromString to convert String to Object and convertToString to convert Object to String. For implementation details, read [Struts2 OGNL Example Tutorial](/community/tutorials/struts2-ognl).
We can implement `com.opensymphony.xwork2.interceptor.Interceptor` interface to create our own interceptor. Once the interceptor class is ready, we need to define that in struts.xml package where we want to use it. We can also create interceptor stack with our custom interceptor and defaultStack interceptors. After that we can configure it for action classes where we want to use our interceptor. One of the best example of using custom interceptor is to validate session, read more about it at [Struts2 Interceptor Tutorial](/community/tutorials/struts-2-interceptor-example).
Interceptor interface defines three methods - init(), destroy() and intercept(). init and destroy are the life cycle methods of an interceptor. Interceptors are Singleton classes and Struts2 initialize a new thread to handle each request. init() method is called when interceptor instance is created and we can initialize any resources in this method. destroy() method is called when application is shutting down and we can release any resources in this method. intercept() is the method called every time client request comes through the interceptor.
An interceptor stack helps us to group together multiple interceptors in a package for further use. struts-default package creates some of the mostly used interceptor stack - basicStack and defaultStack. We can create our own interceptor stack at the start of the package and then configure our action classes to use it.
struts-default is an abstract package that defines all the Struts2 interceptors and commonly used interceptor stack. It is advisable to extend this package while configuring our application package to avoid configuring interceptors again. This is provided to help developers by eliminating the trivial task of configuring interceptor and result pages in our application.
The default URI suffix for Struts2 action is .action, in Struts1 default suffix was .do. We can change this suffix by defining struts.action.extension constant value in our Struts2 configuration file as:
```
<constant name="struts.action.extension" value="action,do"></constant>
```
By default Struts2 looks for result pages in {WEBAPP-ROOT}/{Namespace}/ directory but sometimes we want to keep result pages in another location, we can provide struts.convention.result.path constant value in Struts2 configuration file to change the result pages location. Another way is to use @ResultPath annotation in action classes to provide the result pages location.
File Upload is one of the common tasks in a web application. That's why Struts2 provides built-in support for file upload through FileUploadInterceptor. This interceptor is configured in the struts-default package and provide options to set the maximum size of a file and file types that can be uploaded to the server. Read more about FileUpload interceptor at [Struts2 File Upload Example](/community/tutorials/struts-2-file-upload-example).
Some of the best practices while developing Struts2 application are:
1. Always try to extend struts-default package while creating your package to avoid code redundancy in configuring interceptors.
2. For common tasks across the application, such as logging request params, try to use interceptors.
3. Always keep action classes java bean properties in a separate bean for code reuse and implement ModelDriven interface.
4. If you have custom interceptor that you will use in multiple actions, create interceptor stack for that and then use it.
5. Try to divide your application in different modules with namespace configuration based on functional areas.
6. Try to use Struts2 tags in result pages for code clarify, if needed create your own type converters.
7. Use development mode for faster development, however make sure production code doesn't run in dev mode.
8. Use Struts2 i18n support for resource bundles and to support localization.
9. Struts2 provides a lot of places where you can have resource bundles but try to keep one global resource bundle and one for action class to avoid confusion.
10. struts-default package configures all the interceptors and creates different interceptor stacks. Try to use only what is needed, for example if you don't have localization requirement, you can avoid i18n interceptor.
Struts2 provides a very robust framework for exception handling. We can specify global results in packages and then map specific exceptions to these result pages. The exception mapping can be done at the global package level as well as the action level. It's a good idea to have exception result pages to provide some information to the user when some unexpected exception occurs that is not processed by the application. The sample configuration in the struts.xml file looks like below.
```
<package name="user" namespace="/" extends="struts-default">
<global-results>
<result name="exception">/exception.jsp</result>
<result name="runtime_exception">/runtime_exception.jsp</result>
<result name="error">/error.jsp</result>
</global-results>
<global-exception-mappings>
<exception-mapping exception="java.lang.Exception" result="exception"></exception-mapping>
<exception-mapping exception="java.lang.Error" result="error"></exception-mapping>
<exception-mapping exception="java.lang.RuntimeException" result="runtime_exception"></exception-mapping>
</global-exception-mappings>
<action name="myaction" class="com.journaldev.struts2.exception.MyAction">
</action>
<action name="myspecialaction" class="com.journaldev.struts2.exception.MySpecialAction">
<exception-mapping exception="java.io.IOException" result="login"></exception-mapping>
<result name="login">/error.jsp</result>
</action>
</package>
```
Read more at [Struts2 Exception Handling Example](/community/tutorials/struts2-exception-handling-example-tutorial).
这就是关于Struts2面试问题和答案的全部内容，如果你遇到了我遗漏的任何重要问题，请通过评论让我知道。
Source:
https://www.digitalocean.com/community/tutorials/struts2-interview-questions-and-answers