Set primary identities in an ad hoc dataset

Adobe Experience Platform Query Service allows you to mark dataset columns as either primary or secondary identities using constraints for the SQL ALTER TABLE command. You can use this feature to ensure that flagged fields are consistent with data privacy requirements. This command allows you to add or delete constraints for both primary and secondary identity table columns directly through SQL.

Getting started

Labelling dataset columns as primary or secondary identity requires an understanding of the ALTER TABLE SQL command and a good understanding of data privacy requirements. Before continuing with this document, please review the following documentation:

Add constraints

The ALTER TABLE command allows you to label a dataset column as a person’s identity and then use that label as a primary identity by updating the associated metadata using SQL. This is especially useful when datasets are created through SQL rather than directly from a schema through the Platform UI. The command can be used to ensure that your data operations within Platform are compliant with data usage policies.

Examples

The following example adds a constraint to the existing t1 table. The values of the id column are now marked as primary identities under the IDFA namespace. An identity namespace is a keyword that declares the type of identity data that the field represents.

ALTER TABLE t1 ADD CONSTRAINT PRIMARY IDENTITY (id) NAMESPACE 'IDFA';

The second example ensures that the id column is marked as a secondary identity.

ALTER TABLE t1 ADD CONSTRAINT IDENTITY(id) NAMESPACE 'IDFA';

Drop constraints

Constraints can also be removed from table columns using the ALTER TABLE command.

Examples

The following example removes the requirement that the c1 column be labeled a primary identity in the existing t1 table.

ALTER TABLE t1 DROP CONSTRAINT PRIMARY IDENTITY (c1) ;

As seen below, the same syntax is used to when removing an identity constraint.

ALTER TABLE t1 DROP CONSTRAINT IDENTITY (c1) ;

Show identities

Use the metadata command show identities from the command line interface to display a table with every attributes that is assigned as an identity.

> show identities;

An example of a returned table is displayed below.

 tableName | columnName | datatype | namespace | ifPrimary
-----------+------------+----------+-----------+----------
(0 rows)

XDM limitations

The following list explains important considerations for updating identities in existing datasets when using XDM.

  • To specify a column as an identity, you must also define the namespace to be preserved as metadata for the column.
  • XDM does not support specifying a column name in the namespace attribute.
  • If your schema uses the identityMap XDM field, the root or top-level identityMap object must be labeled as an identity or primary identity.

On this page