Hi @Paulito ,
The easiest way I can think of is to do this in two Processors:
1- JoltTransformJson: Allows you to transform your json by simplifying it into an array or records where each record has list of fieldname:fieldvalue. To achieve this you need to provide the following jolt spec in the "Jolt Specification" property of the processor:
[
{
"operation": "shift",
"spec": {
"Result": {
"ResultData": {
"DATARECORD": {
"*": {
"DATAFIELD": {
"*": {
"FIELDVALUE": "[&3].@(1,FIELDNAME)"
}
}
}
}
}
}
}
}
]
Basically the spec above will give you the following json based on the provided input:
[ {
"name" : "JSMITH",
"namedesc" : "John Smith",
"hireddate" : "01-JAN-2010"
}, {
"name" : "RSTONE",
"namedesc" : "Robert Stone",
"hireddate" : "01-JAN-2011"
} ]
2- QueryRecord Processor: to allow you to select the fields you are interested in for the given API as follows. The query is just like sql query and you can either specify wildcard (*) for all fields or just list particular fields as follows:
The out put of the QueryRecord will look like this:
[
{
"name": "JSMITH",
"namedesc": "John Smith"
},
{
"name": "RSTONE",
"namedesc": "Robert Stone"
}
]
Of course you can make this dynamic for each API by providing both the Jolt spec and the Query as flowfile attributes since both allow expression language (EL) in the value field.
You can also achieve the same result by using just the Jolt Transformation processor but the spec would be more complex and I dont want to overwhelm you in case you are new to it, but if you are interested in that let me know and I will be happy to provide you the spec.
If this helps please accept solution.
Thanks
