CODEWITHSUNDEEP
jsonjq
ScottTheFalcon
ScottTheFalcon

Reputation: 23

Processing JSON with jq - handling array index/name into output

I'm trying to use jq to parse a JSON file for me. I want to get a value from a definition header into the output data in place of an index. A simplified example:

{
  "header": {
    "type": {
      "0": {
             "name": "Cats"
           },
      "3": {
             "name": "Dogs"
           }
    }
  },
  "data": [
    {
      "time": "2019-01-01T02:00:00Z",
      "reading": { 
                   "0": {"value": 90, "note": "start" },
                   "3": {"value": 100 }
                 } 
    }
  ]
}

Using a jq command like jq '.data[] | {time: .time, data: .reading[]}' gives me:

  "time": "2019-01-01T02:00:00Z",
  "data": {
    "value": 90,
    "note": "start"
  }
}
{
  "time": "2019-01-01T02:00:00Z",
  "data": {
    "value": 100
  }
}

I need to get "Cats" or "Dogs" into the result, heading towards an SQL insert.

Something like:

{
  "time": "2019-01-01T02:00:00Z",
  "data": {
    "type: "Cats",      <- line added
    "value": 90,
    "note": "start"
  }
}
...

Or better yet:

{
  "time": "2019-01-01T02:00:00Z",
  "Cats": {                    <- label set to "Cats" instead of "data"
    "value": 90,
    "note": "start"
  }
}
...

Is there a way I can get - what I see as the array index "0" or "3" - to be added as "Cats" or "Dogs"?

Upvotes: 2

Views: 321

Answers (1)

peak
peak

Reputation: 116760

Using the built-in function, INDEX, for creating a dictionary allows a straightforward solution as follows:

(.header.type
 | INDEX(to_entries[]; .key)
 | map_values(.value.name)) as $dict
| .data[] 
| (.reading | keys_unsorted[]) as $k
| {time} + { ($dict[$k]) : .reading[$k] }

Output

{
  "time": "2019-01-01T02:00:00Z",
  "Cats": {
    "value": 90,
    "note": "start"
  }
}
{
  "time": "2019-01-01T02:00:00Z",
  "Dogs": {
    "value": 100
  }
}

Upvotes: 1

Related Questions