@@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