In our project we have a difficult model that contains a lot of arrays and nested fields. This model describes interaction with other systems so its filling can be very different. It is something like logs. We save this data in Hadoop using external table and orc files. It is a very big struct field. Apache ORC Tools is used by us for creating orc files and making orc schema. We don't have problem with writing, but when we read data we get a lot of "null" values. I understand that reasons of it are difficult model and different filled fields. But output in beeline or Hue is unreadable:
{""id"":""mqqqnc"",""value"":[{""Ipv4Value"":null,""StringValue"":
{""value"":""99""}}]},{""id"":""eee"",""value"":
[{""Ipv4Value"":null,""StringValue"":{""value"":""254103""}}]},
{""id"":""uli"",""value"":[{""Ipv4Value"":null,""StringValue"":
{""value"":""asfsfsd""}}
This example is a little part of our struct field but we have three null field.
Is it possible to not show fields with null value?