Include the extractions and relationships json in output

[himynamesdave]

These were previously generated for the note objects #35

We should track these in output/ as data--UUID.json

With a "extractions" and "relationships" section;

{
  "extractions": [
    {
      "ai": {
        "extractions": {
          "extraction_0": {
            "type": "<extraction_key>",
            "value": "<extracted_value>",
            "original_text": "<original_text>",
            "start_index": [
              "start_index_n"
            ]
          },
          "extraction_n": {
            "type": "<extraction_key>",
            "value": "<extracted_value>",
            "original_text": "<original_text>",
            "start_index": [
              "start_index_n"
            ]
          }
        }
      },
      "lookup": {
        "extractions": {
          "extraction_n": {
            "type": "<extraction_key>",
            "value": "<extracted_value>",
            "original_text": "<original_text>",
            "start_index": [
              "start_index_n"
            ]
          }
        }
      },
      "pattern": {
        "extractions": {
          "extraction_n": {
            "type": "<extraction_key>",
            "value": "<extracted_value>",
            "original_text": "<original_text>",
            "start_index": [
              "start_index_n"
            ]
          }
        }
      }
    }
  ],
  "relationships": [
    {
      "source_ref": "<source extraction id>",
      "target_ref": "<target extraction id>",
      "relationship_type": "<valid relationship type>"
    },
    {
      "source_ref": "<source extraction id>",
      "target_ref": "<target extraction id>",
      "relationship_type": "<valid relationship type>"
    }
  ]
}

This will allow us to hook other tools into this file to identify the location of files (or text used for extraction), for example.