使用具有多个内部对象的 GSON 解析 JSON。答案

【问题标题】：Parsing JSON using GSON with multiple inner objects.使用具有多个内部对象的 GSON 解析 JSON。
【发布时间】：2014-05-20 11:14:04
【问题描述】：

我想使用 GSON 解析以下 JSON 响应。如您所见，响应中有很多内部 json 对象。我需要为所有内部对象创建 POJO 类还是有其他选择？

{
    "nodeProperties": [
        {
            "properties": {
                "timeStamp": {
                    "value": 1400475483694,
                    "name": "connectedSince"
                },
                "macAddress": {
                    "value": "00:00:00:00:00:03"
                },
                "tables": {
                    "value": -1
                },
                "capabilities": {
                    "value": 199
                },
                "tier": {
                    "value": 1
                },
                "supportedFlowActions": {
                    "value": "[Controller, Drop, Enqueue, HwPath, Output, PopVlan, SetDlDst, SetDlSrc, SetNwDst, SetNwSrc, SetNwTos, SetTpDst, SetTpSrc, SetVlanId, SetVlanPcp, SwPath]"
                },
                "buffers": {
                    "value": 256
                },
                "description": {
                    "value": "NEXT_NEWSwitch3"
                },
                "forwarding": {
                    "value": 0
                }
            },
            "node": {
                "id": "00:00:00:00:00:00:00:03",
                "type": "OF"
            }
        },
        {
            "properties": {
                "timeStamp": {
                    "value": 1400475481261,
                    "name": "connectedSince"
                },
                "macAddress": {
                    "value": "00:00:00:00:00:02"
                },
                "tables": {
                    "value": -1
                },
                "capabilities": {
                    "value": 199
                },
                "tier": {
                    "value": 1
                },
                "supportedFlowActions": {
                    "value": "[Controller, Drop, Enqueue, HwPath, Output, PopVlan, SetDlDst, SetDlSrc, SetNwDst, SetNwSrc, SetNwTos, SetTpDst, SetTpSrc, SetVlanId, SetVlanPcp, SwPath]"
                },
                "buffers": {
                    "value": 256
                },
                "description": {
                    "value": "None"
                },
                "forwarding": {
                    "value": 0
                }
            },
            "node": {
                "id": "00:00:00:00:00:00:00:02",
                "type": "OF"
            }
        },
        {
            "properties": {
                "timeStamp": {
                    "value": 1400475478695,
                    "name": "connectedSince"
                },
                "macAddress": {
                    "value": "00:00:00:00:00:01"
                },
                "tables": {
                    "value": -1
                },
                "capabilities": {
                    "value": 199
                },
                "supportedFlowActions": {
                    "value": "[Controller, Drop, Enqueue, HwPath, Output, PopVlan, SetDlDst, SetDlSrc, SetNwDst, SetNwSrc, SetNwTos, SetTpDst, SetTpSrc, SetVlanId, SetVlanPcp, SwPath]"
                },
                "buffers": {
                    "value": 256
                },
                "description": {
                    "value": "None"
                },
                "forwarding": {
                    "value": 0
                }
            },
            "node": {
                "id": "00:00:00:00:00:00:00:01",
                "type": "OF"
            }
        }
    ]
}

【问题讨论】：

gson 库使在 java 中将 JSON 对象转换为类变得简单易行。如果您希望将这些对象表示为类，那么是的您需要创建 POJO 类。如果您不一定要这样做（即要将 JSON 转换为 Map 或其他东西），则不应使用 gson
创建 JSON 的人并不太聪明，但是将其解析为完全反映 JSON 结构的 Maps 和 Lists 是一件简单的事情。不过，一般来说，GSON 和 Jackson 并不是很好的选择。
@HotLicks 你能解释一下为什么你认为 JSON 结构不好？
而不是“forwarding : { value : 0 }”等，应该是“forwarding : 0”（当然要加上适当的引号）。额外的对象层什么也没增加。

标签： java json rest gson opendaylight

【解决方案1】：

只需将 JSON 字符串转换为 Map。

示例代码：

FileReader in = new FileReader(new File("resources/json3.txt"));
BufferedReader br = new BufferedReader(in);

// Convert JSON string into MAP object
Gson gson = new Gson();
Type type = new TypeToken<Map<String, ArrayList<Map<String, Map<String, Object>>>>>() {}.getType();
Map<String, ArrayList<Map<String, Map<String, Object>>>> map = gson.fromJson(br, type);

for (String key : map.keySet()) {
    System.out.println(key);
    for (Map<String, Map<String, Object>> value : map.get(key)) {
        for (String k : value.keySet()) {
            System.out.println(k);
            for (String k1 : value.get(k).keySet()) {
                System.out.println(k1 + ":" + value.get(k).get(k1));
            }
        }
        System.out.println("--------------");
    }
}

in.close();
br.close();

您也可以使用一些 POJO 类来实现，它们是 JSON 字符串的副本，更简单、更短。

示例代码：

class PropertiesNode {
    Properties properties;
    Node node;
    // getter & setter
}

class Node {
    String id;
    String type;
    // getter & setter
}

class Properties {
    Map<String, Object> timeStamp;
    Map<String, String> macAddress;
    Map<String, Integer> tables;
    Map<String, Integer> capabilities;
    Map<String, Integer> tier;
    Map<String, String> supportedFlowActions;
    Map<String, Integer> buffers;
    Map<String, String> description;
    Map<String, Integer> forwarding;
    // getter & setter
}

Gson gson = new Gson();
Type type = new TypeToken<Map<String, ArrayList<PropertiesNode>>>() {}.getType();
Map<String, ArrayList<PropertiesNode>> nodeProperties = gson.fromJson(br, type);
System.out.println(new GsonBuilder().setPrettyPrinting().create().toJson(nodeProperties).toString());

输出：

{
    "nodeProperties": [
      {
        "properties": {
          "timeStamp": {
            "value": 1.400475483694E12,
            "name": "connectedSince"
          },
          "macAddress": {
            "value": "00:00:00:00:00:03"
          },
          "tables": {
            "value": -1
          },
          "capabilities": {
            "value": 199
          },
          "tier": {
            "value": 1
          },
          "supportedFlowActions": {
            "value": "[Controller, Drop, Enqueue, HwPath, Output, PopVlan, SetDlDst, SetDlSrc, SetNwDst, SetNwSrc, SetNwTos, SetTpDst, SetTpSrc, SetVlanId, SetVlanPcp, SwPath]"
          },
          "buffers": {
            "value": 256
          },
          "description": {
            "value": "NEXT_NEWSwitch3"
          },
          "forwarding": {
            "value": 0
          }
        },
        "node": {
          "id": "00:00:00:00:00:00:00:03",
          "type": "OF"
        }
      },
      {
        "properties": {
          "timeStamp": {
            "value": 1.400475481261E12,
            "name": "connectedSince"
          },
          "macAddress": {
            "value": "00:00:00:00:00:02"
          },
          "tables": {
            "value": -1
          },
          "capabilities": {
            "value": 199
          },
          "tier": {
            "value": 1
          },
          "supportedFlowActions": {
            "value": "[Controller, Drop, Enqueue, HwPath, Output, PopVlan, SetDlDst, SetDlSrc, SetNwDst, SetNwSrc, SetNwTos, SetTpDst, SetTpSrc, SetVlanId, SetVlanPcp, SwPath]"
          },
          "buffers": {
            "value": 256
          },
          "description": {
            "value": "None"
          },
          "forwarding": {
            "value": 0
          }
        },
        "node": {
          "id": "00:00:00:00:00:00:00:02",
          "type": "OF"
        }
      },
      {
        "properties": {
          "timeStamp": {
            "value": 1.400475478695E12,
            "name": "connectedSince"
          },
          "macAddress": {
            "value": "00:00:00:00:00:01"
          },
          "tables": {
            "value": -1
          },
          "capabilities": {
            "value": 199
          },
          "supportedFlowActions": {
            "value": "[Controller, Drop, Enqueue, HwPath, Output, PopVlan, SetDlDst, SetDlSrc, SetNwDst, SetNwSrc, SetNwTos, SetTpDst, SetTpSrc, SetVlanId, SetVlanPcp, SwPath]"
          },
          "buffers": {
            "value": 256
          },
          "description": {
            "value": "None"
          },
          "forwarding": {
            "value": 0
          }
        },
        "node": {
          "id": "00:00:00:00:00:00:00:01",
          "type": "OF"
        }
      }
    ]
  }

【讨论】：

谢谢布拉杰。虽然这可行，但我选择在类属性中为成员变量创建 POJO，而不是映射。