jsonlib中jsonConfig的配置

- 日理万妓 2022-08-06 05:11 157阅读 0赞

json-lib的普及率应该还是比较广的，不过在Spring中倒是倾向于jackson，无奈新项目做到一般才知道jackson，所以还是依旧使用了json-lib~

# 改变某种类型默认值的序列化值 #

在序列化的时候，比较常见的问题是数据库为null的int型字段，被序列化后就赋值成0了，这在有些时候不是我们想看到的结果，那怎么办呢？

json-lib提供了JsonConfig对象供我们来自定义一些序列化时的操作：

jsonConfig.registerDefaultValueProcessor(Integer.class,
    				new DefaultValueProcessor() {
    					@Override
    					public Object getDefaultValue(Class type) {
    						return null;
    					}
    				});

也就是说，对于需要 序列化的对象的Integer的字段来说，只要值为null（也就是Integer的默认值），那么在序列化成json的时候，该key的value则也会为null

我们可以通过以下代码测试一下：

public static void main(String[] args) {
    		JsonConfig jsonConfig = new JsonConfig();
    		jsonConfig.registerDefaultValueProcessor(Integer.class,
    				new DefaultValueProcessor() {
    					@SuppressWarnings("unchecked")
    					@Override
    					public Object getDefaultValue(Class type) {
    						return "";
    					}
    				});
    
    		Employee employee = new Employee();
    		employee.setName("zhu");
    		employee.setRole(1);
    		System.out.println(JSONObject.fromObject(employee));
    		System.out.println(JSONObject.fromObject(employee, jsonConfig));
    	}

输出为：

{"depno":"","empno":"","groupno":0,"name":"zhu","online":0,"phone":"","role":1}
    {"depno":"","empno":"","groupno":null,"name":"zhu","online":null,"phone":"","role":1}

#     #

# 改变某种类型的序列化策略 #

考虑在序列化的时候我们可能并不想原模原样的照着 对象 的值来序列化，也许我们想要玩点花样？比如加个密啊，或者加个格式化的策略

就拿Date类型来说吧，如果说按照默认的序列化策略，看看会是什么结果：

\{"birthday":\{"date":24,"day":3,"hours":11,"minutes":33,"month":11,"seconds":4,"time":1419391984356,"timezoneOffset":-480,"year":114\}\}

我想大多数情况下你应该不会是想要这种结果吧？格式化成“yyyy-MM-dd”这种格式也许是最常见的

那么我们可以通过为Date类型的字段配置一个我们自己的Processor：

首先，我们需要实现JsonValueProcessor接口，这个接口有两个方法需要实现：

<pre name="code" class="java">public Object processObjectValue(String paramString,Object paramObject, JsonConfig paramJsonConfig);
    
    public Object processArrayValue(Object paramObject,JsonConfig paramJsonConfig);

当对象的某个属性是该类型时，那么调用processObjectValue(...)方法来序列化该属性

而当对象的某个属性是该类型的数组或List时，那么调用processArrayValue(...)方法来序列化该属性

知道了这个，下面我们就来实现序列化中Date类型的格式化吧，这里对于Date的数组类型就不处理了，直接返回null：

public static void main(String[] args) {
    		JsonConfig jsonConfig = new JsonConfig();
    		jsonConfig.registerJsonValueProcessor(Date.class,
    				new JsonValueProcessor() {
    
    					/**
    					 * paramString -> 参数名 paramObject -> 参数值
    					 */
    					@Override
    					public Object processObjectValue(String paramString,
    							Object paramObject, JsonConfig paramJsonConfig) {
    						if (paramObject == null) {
    							return null;
    						}
    						String ret = null;
    						try {
    							SimpleDateFormat format = new SimpleDateFormat(
    									"yyyy-MM-dd HH:mm:ss");
    							ret = format.format((Date) paramObject);
    						} catch (Exception e) {
    							SimpleDateFormat format = new SimpleDateFormat(
    									"yyyy-MM-dd");
    							ret = format.format((Date) paramObject);
    						}
    						return ret;
    					}
    
    					@Override
    					public Object processArrayValue(Object paramObject,
    							JsonConfig paramJsonConfig) {
    						return null;
    					}
    				});
    		Employee employee = new Employee();
    		employee.setName("zhu");
    		employee.setRole(1);
    		employee.setBirthday(new Date());
    		System.out.println("no jsonConfig:" + JSONObject.fromObject(employee));
    		System.out.println("with jsonConfig:" + JSONObject.fromObject(employee, jsonConfig));
    	}

输出结果是：

no jsonConfig:{"birthday":{"date":24,"day":3,"hours":13,"minutes":46,"month":11,"seconds":31,"time":1419399991491,"timezoneOffset":-480,"year":114},"depno":"","empno":"","groupno":0,"name":"zhu","online":0,"phone":"","role":1}
    with jsonConfig:{"birthday":"2014-12-24 13:46:31","depno":"","empno":"","groupno":0,"name":"zhu","online":0,"phone":"","role":1}

#     #

# 过滤需要序列化的属性 #

过滤某些不必要序列化的属性有两种方式： 1、给定一个过滤条件，若满足则被过滤掉2、直接指定需要过滤的属性名

**先来看看第一种：**

public static void main(String[] args) {
    		JsonConfig jsonConfig = new JsonConfig();
    		jsonConfig.setJsonPropertyFilter(new PropertyFilter() {
    			@Override
    			public boolean apply(Object source, String name, Object value) {
    				return value == null;
    			}
    		});
    		Employee employee = new Employee();
    		employee.setName("zhu");
    		employee.setRole(1);
    		System.out.println("no jsonConfig:" + JSONObject.fromObject(employee));
    		System.out.println("with jsonConfig:" + JSONObject.fromObject(employee, jsonConfig));
    	}

输出结果是：

no jsonConfig:{"birthday":null,"depno":"","empno":"","groupno":0,"name":"zhu","online":0,"phone":"","role":1}
    with jsonConfig:{"name":"zhu","role":1}

**第二种过滤方式**，直接根据属性的名称：

public static void main(String[] args) {
    		JsonConfig jsonConfig = new JsonConfig();
    		jsonConfig.setExcludes(new String[] { "depno", "empno", "groupno",
    				"online" });
    		Employee employee = new Employee();
    		employee.setName("zhu");
    		employee.setRole(1);
    		System.out.println("no jsonConfig:" + JSONObject.fromObject(employee));
    		System.out.println("with jsonConfig:"
    				+ JSONObject.fromObject(employee, jsonConfig));
    	}

可以看到，这里通过jsonConfig.setExcludes(new String\[\]\{"double", "boolean"\});将double和boolean这两个属性

输出结果：

no jsonConfig:{"birthday":null,"depno":"","empno":"","groupno":0,"name":"zhu","online":0,"phone":"","role":1}
    with jsonConfig:{"birthday":null,"name":"zhu","phone":"","role":1}