@@ResponseBody 过滤POJO属性

一、问题的提出。 

项目使用Spring MVC框架，并用jackson库处理JSON和POJO的转换。在POJO转化成JSON时，希望动态的过滤掉对象的某些属性。所谓动态，是指的运行时，不同的controler方法可以针对同一POJO过滤掉不同的属性。 

以下是一个Controler方法的定义，使用@ResponseBody把获得的对象列表写入响应的输出流（当然，必须配置jackson的MappingJacksonHttpMessageConverter，来完成对象的序列化）

1 2 3 4 5 6 7 8

@RequestMapping(params = "method=getAllBmForList") @ResponseBody public List<DepartGenInfo> getAllBmForList(HttpServletRequest request,         HttpServletResponse response) throws Exception {           BmDto dto = bmglService.getAllBm();     return dto.getBmList(); }

POJO定义

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

public class DepartGenInfo implements java.io.Serializable {        private String depid;      private String name;      private Company company;        //getter...      //setter... }   public class Company  {        private String comid;      private String name; <pre name="code" class="java">      //getter...      //setter... }

我希望在getAllBmForList返回时，过滤掉DepartGenInfo的name属性，以及company的comid属性。 

jackson支持@JsonIgnore和@JsonIgnoreProperties注解，但是无法实现动态过滤。jackson给出了几种动态过滤的办法，我选择使用annotation mixin 


•JSON View 
•JSON Filter 
•Annotation Mixin 
二、使用annotation mixin动态过滤

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

@RequestMapping(params = "method=getAllBmForList") public void getAllBmForList(HttpServletRequest request,         HttpServletResponse response) throws Exception {           BmDto dto = bmglService.getAllBm();          ObjectMapper mapper = new ObjectMapper();     SerializationConfig serializationConfig = mapper.getSerializationConfig();     serializationConfig.addMixInAnnotations(DepartGenInfo.class,       DepartGenInfoFilter.class);       serializationConfig.addMixInAnnotations(Company.class,       CompanyFilter.class);           mapper.writeValue(response.getOutputStream(),dto.getBmList());     return; }

DepartGenInfoFilter的定义如下：

@JsonIgnoreProperties(value={"name"}) //希望动态过滤掉的属性

public interface DepartGenInfoFilter {

}<br><span>//CompanyFilter的定义如下： </span>

 

这个实现方法看起来非常不简洁，需要在动态过滤的时候写不少代码，而且也改变了@ResponseBody的运行方式，失去了REST风格，因此考虑到使用AOP来进行处理。 

二、最终解决方案 

先看下我想达到的目标，通过自定义注解的方式来控制动态过滤。

@XunerJsonFilters(value={@XunerJsonFilter(mixin=DepartGenInfoFilter.class, target=DepartGenInfo.class)

            ,@XunerJsonFilter(mixin=CompanyFilter.class, target=Company.class)})

    @RequestMapping(params = "method=getAllBmForList")

    @ResponseBody

    public List getAllBmForList(HttpServletRequest request,

            HttpServletResponse response) throws Exception {

         

        BmDto dto = bmglService.getAllBm();

return dto.getBmList();

    }

@XunerJsonFilters和@XunerJsonFilter是我定义的注解。@XunerJsonFilters是@XunerJsonFilter的集合，@XunerJsonFilter定义了混合的模板以及目标类。

1 2 3 4 5 6 7 8 9

@Retention(RetentionPolicy.RUNTIME) public @interface XunerJsonFilters {     XunerJsonFilter[] value(); } @Retention(RetentionPolicy.RUNTIME) public @interface XunerJsonFilter {   Class<?> mixin() default Object.class;   Class<?> target() default Object.class; }

当然，只是定义注解并没有什么意义。重要的是如何根据自定义的注解进行处理。我定义了一个AOP Advice如下：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

public class XunerJsonFilterAdvice {       public Object doAround(ProceedingJoinPoint pjp) throws Throwable {         MethodSignature msig = (MethodSignature) pjp.getSignature();         XunerJsonFilter annotation = msig.getMethod().getAnnotation(                 XunerJsonFilter.class);         XunerJsonFilters annotations = msig.getMethod().getAnnotation(                 XunerJsonFilters.class);           if (annotation == null && annotations == null) {             return pjp.proceed();         }           ObjectMapper mapper = new ObjectMapper();         if (annotation != null) {             Class<?> mixin = annotation.mixin();             Class<?> target = annotation.target();                           if (target != null) {                 mapper.getSerializationConfig().addMixInAnnotations(target,                         mixin);             } else {                 mapper.getSerializationConfig().addMixInAnnotations(                         msig.getMethod().getReturnType(), mixin);             }         }                   if (annotations != null) {             XunerJsonFilter[] filters= annotations.value();             for(XunerJsonFilter filter :filters){                 Class<?> mixin = filter.mixin();                 Class<?> target = filter.target();                                   if (target != null) {                     mapper.getSerializationConfig().addMixInAnnotations(target,                             mixin);                 } else {                     mapper.getSerializationConfig().addMixInAnnotations(                             msig.getMethod().getReturnType(), mixin);                 }             }                       }                     try {             mapper.writeValue(WebContext.getInstance().getResponse()                     .getOutputStream(), pjp.proceed());         } catch (Exception ex) {             throw new RuntimeException(ex);         }         return null;     }   }

其中pointcut的expression能够匹配到目标类的方法。 

在doAround方法中，需要获得当前引用的HttpResponse对象，因此使用以下方法解决： 

创建一个WebContext工具类：

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

public class WebContext {       private static ThreadLocal<WebContext> tlv = new ThreadLocal<WebContext>();     private HttpServletRequest request;     private HttpServletResponse response;     private ServletContext servletContext;       protected WebContext() {     }       public HttpServletRequest getRequest() {         return request;     }       public void setRequest(HttpServletRequest request) {         this.request = request;     }       public HttpServletResponse getResponse() {         return response;     }       public void setResponse(HttpServletResponse response) {         this.response = response;     }       public ServletContext getServletContext() {         return servletContext;     }       public void setServletContext(ServletContext servletContext) {         this.servletContext = servletContext;     }       private WebContext(HttpServletRequest request,             HttpServletResponse response, ServletContext servletContext) {         this.request = request;         this.response = response;         this.servletContext = servletContext;     }       public static WebContext getInstance() {         return tlv.get();     }       public static void create(HttpServletRequest request,             HttpServletResponse response, ServletContext servletContext) {         WebContext wc = new WebContext(request, response, servletContext);         tlv.set(wc);     }       public static void clear() {         tlv.set(null);     } }

别忘了在web.xml中增加这个filter。 




OK，It is all。 




四、总结 

设计的一些要点： 

1、要便于程序员使用。程序员根据业务逻辑需要过滤字段时，只需要定义个"Filter“，然后使用注解引入该Filter。 

2、引入AOP来保持原来的REST风格。对于项目遗留的代码，不需要进行大幅度的修改，只需要增加注解来增加对过滤字段的支持。 

仍需解决的问题： 

按照目前的设计，定义的Filter不支持继承，每一种动态字段的业务需求就会产生一个Filter类，当类数量很多时，不便于管理。 




五、参考资料 

http://www.cowtowncoder.com/blog/archives/cat_json.html 

http://www.jroller.com/RickHigh/entry/filtering_json_feeds_from_spring