有时在解析json数据中的字段key是动态可变的时候，由于Gson是使用静态注解的方式来设置实体对象的，因此我们很难直接对返回的类型来判断。但Gson在解析过程中如果不知道解析的字段，就会将所有变量存储在一个Map中，我们只要实例化这个map就能动态地取出key和value了。先给出一段jsondata，这是天气预报的数据，其中day_20151002这种key是随日期而变化的，在实体类中就不能当做静态变量来处理，我们就通过map来取出其映射对象。{
“resultcode”: “200”,
“reason”: “successed!”,
“result”: {
“sk”: {
“temp”: “24”,
“wind_direction”: “东北风”,
“wind_strength”: “2级”,
“humidity”: “28%”,
“time”: “17:38”
},
“today”: {
“temperature”: “15℃~26℃”,
“weather”: “多云转晴”,
“wind”: “东北风微风”,
“week”: “星期日”,
“city”: “桂林”,
“date_y”: “2015年10月11日”,
“dressing_index”: “舒适”,
“dressing_advice”: “建议着长袖T恤、衬衫加单裤等服装。年老体弱者宜着针织长袖衬衫、马甲和长裤。”,
“uv_index”: “弱”,
“comfort_index”: “”,
“wash_index”: “较适宜”,
“travel_index”: “较适宜”,
“exercise_index”: “较适宜”,
“drying_index”: “”
},
“future”: {
“day_20151011”: {
“temperature”: “15℃~26℃”,
“weather”: “多云转晴”,
“wind”: “东北风微风”,
“week”: “星期日”,
“date”: “20151011”
},
“day_20151012”: {
“temperature”: “16℃~27℃”,
“weather”: “晴转多云”,
“wind”: “微风”,
“week”: “星期一”,
“date”: “20151012”
},
“day_20151013”: {
“temperature”: “16℃~26℃”,
“weather”: “多云转晴”,
,
“wind”: “微风”,
“week”: “星期二”,
“date”: “20151013”
},
“day_20151014”: {
“temperature”: “17℃~27℃”,
“weather”: “晴”,
“wind”: “北风微风”,
“week”: “星期三”,
“date”: “20151014”
},
“day_20151015”: {
“temperature”: “17℃~28℃”,
“weather”: “晴”,
“wind”: “北风微风”,
“week”: “星期四”,
“date”: “20151015”
},
“day_20151016”: {
“temperature”: “17℃~30℃”,
“weather”: “晴”,
“wind”: “北风微风”,
“week”: “星期五”,
“date”: “20151016”
},
“day_20151017”: {
“temperature”: “17℃~30℃”,
“weather”: “晴”,
“wind”: “北风微风”,
“week”: “星期六”,
“date”: “20151017”
}
}
},
“error_code”: 0
}12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182相关的实体类如下：public class FutureDay {
private String temperature;
private String weather;
private String wind;
private String week;
private String date;
}1234567public class Result {
private Sk sk;
private Today today;
private Map<String,FutureDay> future;
}12345public class Sk {
private String temp;
private String wind_direction;
private String wind_strength;
private String humidity;
private String time;
}1234567public class Today {
private String temperature;
private String weather;
private String week;
private String city;
private String date_y;
private String dressing_index;
private String dressing_advice;
private String uv_index;
private String comfort_index;
private String wash_index;
private String travel_index;
private String exercise_index;
private String drying_index;
}
12345678910111213141516public class Response {
private String resultcode;
private String reason;
private String error_code;
private Result result;
}123456具体解析过程如下代码所示：import java.io.BufferedReader;
import java.io.FileReader;
import java.util.Map;
import weather.*;
import com.google.gson.Gson;
public class GsonParseDynamicKey {
public static void main( String args []){
String jsondata = readJsonFile();//从文件中读取出json字符串，并打印出来
Gson gson = new Gson();
System.out.println(“Start Gson parse jsondata”);
Response response = gson.fromJson(jsondata, Response.class);
System.out.println(response.toString());
System.out.println(response.getResult().getSk().toString());
System.out.println(response.getResult().getToday().toString());
Map<String, FutureDay> future = response.getResult().getFuture(); //对动态的key，来创建map，间接从中取出实体类futrue。
System.out.println("Keyset method"); //这里取出value的方法有两种keySet() entrySet().都给出了遍历的方法
for (String key:future.keySet()){ //遍历取出key，再遍历map取出value。
System.out.println("key:"+key);
System.out.println(future.get(key).toString());
}
System.out.println("Entryset method");
for (Map.Entry<String,FutureDay> pair:future.entrySet()){//遍历取出键值对，调用getkey()，getvalue()取出key和value。
System.out.println("key:"+pair.getKey());
System.out.println(pair.getValue().toString());
}
}123456789101112131415161718192021222324252627282930这里顺便一提遍历Map的两种方法keySet()，entrySet()的差别。
keySet()方法返回的是key的集合set，entrySet()返回的是键值对的集合set。虽然两者从set遍历取出元素的方法是一样的，但是根据这个元素取出value的效率有些不同。前者取出的元素是key，还要去原map中遍历取出value。
后者取出的元素是键值对，直接调用getkey()，getvalue()方法就能快速取出key和value。显然在map中存在大量键值对时，使用entrySet()来取出value的效率更高。
本文参考：http://blog.csdn.net/Chaosminds/article/details/49049455