volume.json

{
  "$schema": "http://json-schema.org/schema#",
  "title": "Databrary Volume",
  "description": "Basic organizational unit for data.",
  "definitions": {
    "date": {
      "type": "string",
      "pattern": "^[12][0-9]{3}-[01][0-9]-[0-3][0-9]$"
    },
    "numeric": {
      "type": "string",
      "pattern": "^[0-9]*(\\.[0-9]*)?$"
    },
    "citation": {
      "description": "Publications/products corresponding to study volumes.",
      "type": "object",
      "properties": {
        "head": {"type": "string"},
        "title": {"type": "string"},
        "url": {"type": "string",
          "title": "URL or DOI",
          "format": "uri"
        },
        "authors": {"type": "array",
          "items": {"type": "string"}
        },
        "year": {"type": "integer",
          "minimum": 1900,
          "maximum": 2900
        }
      },
      "required": ["head"],
      "additionalProperties": false
    },
    "link": {
      "description": "Link to externals resources/references.",
      "type": "object",
      "properties": {
        "head": {"type": "string"},
        "url": {"type": "string",
          "title": "URL or DOI",
          "format": "uri"
        }
      },
      "required": ["head", "url"],
      "additionalProperties": false
    },
    "container": {
      "title": "session or materials",
      "description": "Organizational unit within volume containing related files (with common annotations), often corresponding to an individual data session (single visit/acquisition/participant/group/day).",
      "type": "object",
      "properties": {
        "id": {"type": "integer"},
        "key": {"type": "string"},
        "name": {"type": "string"},
        "date": {"$ref": "#/definitions/date"},
        "top": {"type": "boolean",
          "default": false
        },
        "release": {
          "description": "Sharing/release permissions granted by participants on (portions of) contained data.",
          "anyOf": [
            {"$ref": "#/definitions/release"}
          ],
          "default": null
        },
        "records": {
          "description": "Attachment of records to slots.",
          "type": "array",
          "items": {"$ref": "#/definitions/record"}
        },
        "assets": {
          "description": "Attachment point of assets, which, in the case of timeseries data, should match asset.duration.",
          "type": "array",
          "items": {"$ref": "#/definitions/asset"}
        }
      },
      "required": ["key"],
      "additionalProperties": false
    },
    "release": {
      "title": "sharing release level",
      "description": "Levels at which participants or researchers may choose to share data.",
      "enum": [null, "PRIVATE", "SHARED", "EXCERPTS", "PUBLIC"]
    },
    "offset": {
      "description": "Point in time in milliseconds or MM:SS.sss, used as endpoint for segment.  Null represents indeterminate value, usually representing +-Infinity as an endpoint.",
      "type": ["number", "string", "null"],
      "pattern": "^[-+]?([0-9]+:){0,2}[0-9]+(\\.[0-9]*)?$"
    },
    "segment": {
      "description": "Explicit range of time (within a container): an empty segment, a singleton point in time, or a start and end point.",
      "type": "array",
      "items": {"$ref": "#/definitions/offset"},
      "maxItems": 2,
      "additionalItems": false,
      "default": [null, null]
    },
    "position": {
      "default": [null, null]
    },
    "record": {
      "description": "Sets of metadata measurements organized into or applying to a single cohesive unit. These are assumed to be unique within a volume, such that two identical records are considered the same object.",
      "type": "object",
      "properties": {
        "id": {"type": "integer"},
        "key": {"type": "string"},
        "category": {
          "description": "Types of records that are relevant for data organization.",
          "enum": ["participant", "pilot", "exclusion", "condition", "group", "task", "context"]
        },
        "positions": {"type": "array",
	  "items": {"$ref": "#/definitions/segment"}
	},
        "ID": {"type": "string"},
        "name": {"type": "string"},
        "reason": {"type": "string",
          "enum": ["Did not meet inclusion criteria","Procedural/experimenter error","Withdrew/fussy/tired","Outlier"]
        },
        "summary": {"type": "string"},
        "description": {"type": "string"},
        "birthdate": {"$ref": "#/definitions/date"},
        "gender": {"type": "string",
          "enum": ["Female","Male"]
        },
        "race": {"type": "string",
          "enum": ["American Indian or Alaska Native","Asian","Native Hawaiian or Other Pacific Islander","Black or African American","White","More than one","Unknown or not reported"]
        },
        "ethnicity": {"type": "string",
          "enum": ["Not Hispanic or Latino","Hispanic or Latino"]
        },
        "disability": {"type": "string"},
        "language": {"type": "string"},
        "setting": {"type": "string",
          "enum": ["Lab","Home","Classroom","Outdoor","Clinic"]
        },
        "country": {"type": "string"},
        "state": {"type": "string",
          "enum": ["AL","AK","AZ","AR","CA","CO","CT","DE","DC","FL","GA","HI","ID","IL","IN","IA","KS","KY","LA","ME","MT","NE","NV",
            "NH","NJ","NM","NY","NC","ND","OH","OK","OR","MD","MA","MI","MN","MS","MO","PA","RI","SC","SD","TN","TX","UT","VT","VA","WA",
            "WV","WI","WY"]
        },
        "info": {"type": "string"},
        "gestational age": {"$ref": "#/definitions/numeric"},
        "pregnancy term": {"type": "string",
          "enum": ["Full term","Preterm"]
        },
        "birth weight": {"$ref": "#/definitions/numeric"}
      },
      "required": ["key", "category"],
      "additionalProperties": false
    },
    "asset": {
      "title": "file",
      "description": "Assets reflecting files in primary storage.",
      "type": "object",
      "properties": {
        "id": {"type": "integer"},
        "release": {"$ref": "#/definitions/release"},
        "name": {"type": "string"},
        "position": {"anyOf": [
          {"$ref": "#/definitions/offset",
            "description": "Start point only (end point inferred from length)."
          },
          {"$ref": "#/definitions/segment"},
          {"type": "string",
            "enum": ["auto"],
            "description": "Automatically postioned immediately following previous positioned asset."
          }
        ]},
        "file": {"type": "string",
          "description": "Relative path to file under '/nyu/stage/$directory'."
        },
        "clip": {"$ref": "#/definitions/segment",
          "description": "Portion of source file to extract as this asset. Note that segment is relative to clip start."
        },
        "options": {"type": "array",
          "items": {"type": "string"},
          "description": "Additional transcode options to pass to ffmpeg."
        },
        "replace": {"type": ["string", "integer"],
          "description": "Original file staging path or asset id to replace."
        }
      },
      "additionalProperties": false
    }
  },
  "type": "object",
  "properties": {
    "name": {"type": "string"},
    "containers": {"type": "array",
      "items": {"$ref": "#/definitions/container"},
      "description": "The first container is required and must have 'top' set.",
      "minItems": 1
    },
    "directory": {"type": "string",
      "title": "Asset base directory",
      "description": "Staging directory (normally within /nyu/stage) within which to find all asset files.",
      "default": ""
    }
  },
  "required": ["name", "containers"],
  "additionalProperties": false
}