XDM field type constraints

Last update: April 2, 2025

CREATED FOR:

  • Developer

In Experience Data Model (XDM) schemas, a field’s type constrains what kind of data the field can contain. This document provides an overview of each core field type, including the other serialization formats they can be mapped to and how to define your own field types in the API in order to enforce different constraints.

Getting started

Before using this guide, please review the basics of schema composition for an introduction to XDM schemas, classes, and schema field groups.

If you plan on defining your own field types in the API, it is strongly recommended that you start with the Schema Registry developer guide to learn how to create field groups and data types to include your custom fields in. If you are using the Experience Platform UI to create your schemas, see the guide on defining fields in the UI to learn how implement constraints on fields that you define within custom field groups and data types.

Base structure and examples

XDM is built on top of JSON Schema, and therefore XDM fields inherit a similar syntax when defining their type. Understanding how different field types are represented in JSON Schema can help indicate the base constraints of each type. Custom fields names are case-insensitive and must have different names at the same level in your schema.

NOTE
See the API fundamentals guide for more information on JSON Schema and other underlying technologies in Experience Platform APIs.

The following table outlines how each XDM type is represented in JSON Schema, along with an example value that conforms to the type:

XDM typeJSON SchemaExample
String 
{"type": "string"}
"Platinum"
Number 
{"type": "number"}
12925.49
Long 
{
  "type": "integer",
  "maximum": 9007199254740991,
  "minimum": -9007199254740991
}
1478108935
Integer 
{
  "type": "integer",
  "maximum": 2147483648,
  "minimum": -2147483648
}
24906290
Short 
{
  "type": "integer",
  "maximum": 32768,
  "minimum": -32768
}
15781
Byte 
{
  "type": "integer",
  "maximum": 128,
  "minimum": -128
}
90
Date* 
{
  "type": "string",
  "format": "date"
}
"2019-05-15"
DateTime* 
{
  "type": "string",
  "format": "date-time"
}
"2019-05-15T20:20:39+00:00"
Boolean 
{"type": "boolean"}
true

*All date-formatted strings must conform to the ISO 8601 standard (RFC 3339, section 5.6).

Mapping XDM types to other formats

The sections below describe how each XDM type maps to other common serialization formats:

NOTE
Among the standard XDM types listed in the tables below, the Map type is also included. Maps are used in standard schemas when data is represented as keys that map to certain values, or where keys cannot reasonably be included in a static schema and must be treated as data values.
Many standard XDM components use map types, and you can also define custom map fields if desired. The map type’s inclusion in the tables below is intended to help you determine how to map your existing data to XDM if it is currently stored in any of the formats listed below.

Parquet, Spark SQL, and Java

XDM typeParquetSpark SQLJava
StringType: BYTE_ARRAY
Annotation: UTF8		StringTypejava.lang.String
NumberType: DOUBLELongTypejava.lang.Double
LongType: INT64LongTypejava.lang.Long
IntegerType: INT32
Annotation: INT_32		IntegerTypejava.lang.Integer
ShortType: INT32
Annotation: INT_16		ShortTypejava.lang.Short
ByteType: INT32
Annotation: INT_8		ByteTypejava.lang.Short
DateType: INT32
Annotation: DATE		DateTypejava.util.Date
DateTimeType: INT64
Annotation: TIMESTAMP_MILLIS		TimestampTypejava.util.Date
BooleanType: BOOLEANBooleanTypejava.lang.Boolean
MapMAP-annotated group

(<key-type> must be STRING)		MapType

(keyType must be StringType)		java.util.Map