Type properties and formats
- Last UpdatedAug 19, 2026
- 3 minute read
Type properties define the fields exposed by a type and how values for those fields are stored and interpreted. In addition to specifying the property data type, optional keywords can be used to further describe property behavior.
The following table describes the keywords that can be used when defining properties within a type.
|
Keyword |
Description |
|---|---|
|
type |
Defines the data type of the property, such as a string, integer, number, or a reference type. It must resolve to a data type supported by the destination system. You can specify a type value directly or a reftypeid referencing an existing type definition. OMF does not enforce a fixed list of supported property types; validation and interpretation are performed by the receiving endpoint. Properties must have an associated type. |
|
format |
Optionally refines how a property’s value is interpreted or stored for a given base type. The meaning and supported values for format are determined by the receiving system. Use this keyword only if required by the destination. Note: This keyword does not influence indexing behavior in CONNECT data services. |
|
isindex |
Indicates that the property is used as an index for streaming data. At least one indexed property is required for streaming data types. Multiple index properties may be defined, and their order determines the compound index. |
|
uom |
Optionally specifies the unit of measure for a numeric property. The unit defined on the type may be overridden at the container level using propertyoverrides. |
|
name |
Optional display name for the property. Value defined on the type can be overridden at the container level using propertyoverrides. |
|
description |
Optional description for the property. Value defined on the type can be overridden at the container level using propertyoverrides. |
|
metadata |
Optional custom metadata associated with the property definition. |
Nullable Type Properties
A property may allow null values depending on its type definition and destination system behavior. Nullability is interpreted by the receiving system and is not enforced by the OMF specification
Type Reuse
OMF supports reusing type definitions through the use of reftypeid at the property level. Type reuse is supported for referenced types that share the same classification as the containing type, or for types that do not specify a classification.
When reftypeid is used on a property, the referenced type’s properties are composed under that property as a nested structure. This allows common property groups to be defined once and reused across multiple type definitions, promoting consistency and reducing schema duplication.
In the example below, a reusable type named LocationProperties is defined and then referenced by the TankV2 type using reftypeid. The TankV2 type includes a Location property that contains the Latitude and Longitude sub‑properties defined by LocationProperties.
The referenced type LocationProperties does not define a classification and does not define an index property. As a result, it cannot be used as a standalone type for instance data, and is intended only for reuse as a composed property within other types.
{
"id":"LocationProperties",
"type":"object",
"properties": {
"Latitude":{ "type": "number", "format": "float32" },
"Longitude":{ "type": "number", "format": "float32" }
}
}, {
"id":"TankV2",
"type":"object",
"classification":"entity",
"properties": {
"TankName": { "type": "string", "isname": true, "isindex": true },
"Serial": { "type": "string" },
"Model": { "type": "string" },
"Location": { "reftypeid":"LocationProperties" }
}
}
Type Qualifiers
OMF allows properties to include additional metadata and constraint information, such as units of measure and numeric limits. These attributes describe the intended meaning of a property and may be interpreted or enforced by the destination system. Certain property attributes may be overridden at the container level using propertyoverrides.
Data Quality
OMF supports explicit modeling of data quality through the qualityschema property. Properties marked with a qualityschema indicate that they carry quality information and define the quality encoding used. This allows quality to be interpreted consistently across different source systems and enables multiple properties to contribute to quality context when required. Interpretation and enforcement of data quality are handled by the destination system.
Interpolation values
The following interpolation values may be specified to describe the intended handling of time-series data. Support and behavior are destination-specific, and unsupported values may be ignored.
-
continuous
-
discrete
-
stepwisecontinuousleading
-
stepwisecontinuousfollowing