Bulk Upload Your Metadata

If you are looking to upload many metadata records at once, Nexus can support that. There are two options:

1. Paste raw JSON into the request body here:

{{your.nexus.url}}/api/v1/projects/{projectId}/datasources/{dataSourceId}/metadata/CreateMetadata

2. Provide a JSON file here:

{{your.nexus.url}}/api/v1/projects/{projectId}/datasources/{dataSourceId}/metadata/CreateMetadataFromFile

Note: Bulk uploading metadata will be enabled in a future release. For deployed instances, replace {your_nexus_url} with the URL of your Nexus deployment (e.g., https://deeplynx.inl.gov).

Build a raw JSON payload

If opting for raw JSON (option 1), in the body of your request, you will need a JSON that can be processed by our backend. In the JSON object, there should be arrays for the types of metadata you want to upload. You can upload classes, relationships, records, edges, and tags.

Keep in mind if pursuing file upload (option 2), the JSON in said file also needs to be formatted correctly, similarly to option 1.

Classes

If you provide a name of a class that already exists in the project, Nexus will update the existing class to contain the provided description and uuid (if any).

Relationships

If a relationship with the same name already exists in the project, Nexus will update the existing relationship to contain the description and uuid you provide (if any).

Records

Any referenced tags and classes that do not exist will be created by Nexus. If there is already a record with the same original_id in the datasource, Nexus will update the existing record to contain the additional information you provided.

Edges

Any referenced relationships that do not exist will be created by Nexus. Any Edges with the same origin_id and destination_id within the project will be updated to include the relationship that you provided (if any).

Tags

Example

Here is an example of a valid JSON:

{
  "classes": [
    {
      "name": "Reactor Core"
    },
    {
      "name": "Core material",
      "description": "Primary substrate or base material used in the manufacturing process",
      "uuid": "a3f5e8c2-4b7d-4e9a-b1c3-5d8f2a6e9b4c"
    }
  ],
  "relationships": [
    {
      "name": "relates_to"
    },
    {
      "name": "depends_on",
      "description": "Indicates that one entity requires or relies upon another entity to function properly or be completed",
      "uuid": "7c2d9f3e-6a1b-4c8e-9d5a-3f7b8e2c4a1d"
    }
  ],
  "records": [
    {
      "original_id": "1",
      "name": "Reactor Core",
      "class_name": "Reactor Core",
      "properties": {
        "model results name": [
          "k_crit",
          "k_min",
          "k_max"
        ],
        "height": 200.0,
        "diameter": 224.0,
        "total core power": 5
      },
      "description": "Core Description"
    },
    {
      "original_id": "2",
      "name": "Core material",
      "class_name": "Core material",
      "properties": {
        "description": "Graphite"
      },
      "description": "Graphite",
      "uri": "uri_to_data",
      "tags": [
        "eco-friendly"
      ]
    }
  ],
  "edges": [
    {
      "origin_oid": "2",
      "relationship_name": "relates_to",
      "destination_oid": "1"
    }
  ],
  "tags": [
    {
      "name": "dangerous"
    },
    {
      "name": "safe"
    },
    {
      "name": "expensive"
    }
  ]
}

Why Natural Keys?

You may have noticed that we are using natural keys (origin_oid, class_name, etc.) instead of IDs. This is to simplify the generation of JSON so you do not need to know their IDs in the Nexus database beforehand. Nexus will check to see if the natural key references metadata in the same JSON object, then if it doesn’t find one, it checks the database for any records that have those natural keys. If a record with the natural key already exists in the database, Nexus will update the existing record with the fields that were provided in the JSON.