Update JSON output to be consistent across pipelings - also handle duplicate column names

[pskrbasu]

Update Steampipe JSON output to align all cases around a single format. Cases:

• steampipe query --output json
• powerpipe query run --output json
• snapshot data format
• Pipes query API

Proposed format:

{
  "columns": [
    {
      "name": "foo",
      "data_type": "int4"
    },
    {
      "name": "foo_{random4}",
      "data_type": "int4",
      "original_name": "foo"
    }
  ],
  "rows": [
    {
      "foo": 1,
      "foo_{random4}": 2
    }
  ],
  "metadata": {
   ...
  }
}