Create an intelligent structure model using a representative file. Export the model to an .amodel file. After you save the file locally, you can copy it to the relevant file storage system.

For more information, see Intelligent structure models in the Data Integration help.

Create a connection to access data in files that are stored in the relevant system. You can create the following types of connections that will work with the data objects that can incorporate an intelligent structure:

Hadoop Distributed File System
Amazon S3
Microsoft Azure Blob
Microsoft Azure Data Lake Store

Create a data object with an intelligent structure model to represent the files stored as sources. You can create the following types of data objects with an intelligent structure model:
Complex file
Amazon S3
Microsoft Azure Blob
Microsoft Azure Data Lake Store
Configure the data object properties.
In the data object read operation, configure columns to project hierarchical data as a complex data type. Enable the Project Column as Complex Data Type property in the Column Projection properties.

Create a data object to write the data to target storage.
Configure the data object properties.

Create a mapping.
Add a Read transformation based on the data object with the intelligent structure model.
Based on the mapping logic, add other transformations that are supported on the Spark engine. Link the ports and configure the transformation properties based on the mapping logic.
Add a Write transformation based on the data object that passes the data to the target storage or output. Link the ports and configure the transformation properties based on the mapping logic.

Configure the following mapping run-time properties:

Select Hadoop as the validation environment and Spark as the engine.
Select Hadoop as the execution environment and select a Hadoop connection.

Validate the mapping and fix any errors.
Optionally, view the Spark engine execution plan to debug the logic.
Run the mapping.