Deep-merge array items in example JSONs before generating schemas

[alperenbolat]

my json file is ...

{  
  "sabit": {
     "lokasyon":{
            "lat":35.35719841230264,
            "lng":18.569335341453552
         },
      "gemi":{  
         "irtifa":30,
         "yalpa":40,
         "dikeyYalpa":50,
         "pruva":60,
         "kabarma":70
      }
   },
   "arayuzler":{  
      "frekans":80,
      "secilenler":[  
         {  
            "tip":"Konsol",
            "isim":"Konsol",
            "silinemez":true
         },
         {  
            "tip":"Dosya",
            "isim":"Dosya",
            "ozellikler":{  
               "dosya":"noktalar.txt"
            },
            "silinemez":true
         },
         {  
            "isim":"ethernet1",
            "ozellikler":{  
               "ip":"192.168.1.1",
               "port":"3000",
               "protokol": "TCP"
            },
            "tip":"Ethernet",
            "id":1440658104985
         },
         {  
            "isim":"raw2",
            "ozellikler":{  
               "macAddr":"08:00:26:27:31:20"
            },
            "tip":"Raw Ethernet",
            "id":1440658326143
         },
         {  
            "isim":"seri1",
            "ozellikler":{  
               "port":"3000"
            },
            "tip":"Seri Kanal",
            "id":1440658229857
         }
      ]
   }
}

it is generate result is 'özellkler' :
public class Ozellikler {

@JsonProperty("port")
private String port;

/**
 * 
 * @return
 *     The port
 */
@JsonProperty("port")
public String getPort() {
    return port;
}

but json file is contain : port , mac address , ip , protocol , dosya
ı am not resolve this problem
I would appreciate it if you could help
thanks...

[thachhoang]

A simpler example:

[
    {
        "scalar" : 99,
        "list" : [
            { "a" : 1 }
        ],
        "obj" : { "name" : "k" }
    },
    {
        "scalar" : "what",
        "list" : [
            { "b" : 1 }
        ],
        "obj" : { "index" : 8 }
    }
]

In the root array, we're ignoring the first object entirely, overwriting the keys with values from the second object.

We can merge values for the same keys, so the schema is more accurate:

[
    {
        "scalar" : "what", # pick last, because different types
        "list" : [
            # will recursive merge these
            { "a" : 1 },
            { "b" : 1 }
        ],
        "obj" : { "name" : "k", "index" : 8 }
    }
]