{
  "$defs": {
    "SourceType": {
      "description": "Supported ingestion source types.",
      "enum": [
        "csv",
        "json",
        "jsonl",
        "parquet",
        "sql",
        "mongo",
        "bigquery",
        "snowflake",
        "pdf",
        "html",
        "transcript",
        "text"
      ],
      "title": "SourceType",
      "type": "string"
    }
  },
  "description": "An unstructured text document to be chunked and embedded.\n\nAttributes:\n    id: Deterministic document ID.\n    text: Full extracted text.\n    metadata: Arbitrary provenance/metadata (filename, page, title, ...).\n    source_type: The originating source type.\n    uri: The originating URI/path.\n    mime: Best-effort MIME type.",
  "properties": {
    "id": {
      "title": "Id",
      "type": "string"
    },
    "text": {
      "title": "Text",
      "type": "string"
    },
    "metadata": {
      "additionalProperties": true,
      "title": "Metadata",
      "type": "object"
    },
    "source_type": {
      "anyOf": [
        {
          "$ref": "#/$defs/SourceType"
        },
        {
          "type": "null"
        }
      ],
      "default": null
    },
    "uri": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Uri"
    },
    "mime": {
      "anyOf": [
        {
          "type": "string"
        },
        {
          "type": "null"
        }
      ],
      "default": null,
      "title": "Mime"
    }
  },
  "required": [
    "id",
    "text"
  ],
  "title": "Document",
  "type": "object"
}