On an

advanced cluster

, hierarchical data types such as array, map, and struct are assigned those native types. For example, a map field in an Amazon S3 source might have the native data type "map (string_integer)." You cannot edit the metadata for array, map, or struct fields.

On the Data Integration Server, Data Integration flattens complex hierarchical data types into native string datatypes with precision up to 4000 characters. Some native data types come from the connector, and others come from the parser that

Data Integration

uses when it reads the source data. Parser data types are prefixed with the format type. For example, in an Amazon S3 source with the Avro format, a map field that comes from the parser has the native data type avro_string. You can change the native data type for the connector and parser fields.

To change the native data type, edit the metadata for the source, and select the appropriate data type in the

Native Type

column.

When you change the native data type, you cannot change a non-parser data type to a parser data type. For example, in an Amazon S3 source,

Data Integration

sets the native data type for the FileName field to string. You can change the native data type to nstring but not to avro_string. Similarly, you cannot change a parser data type to a non-parser data type.

For more information about editing native data types in complex file sources, see the help for the appropriate connector.