This scenario consists of the following steps:
Export hierarchy information from the source data to a common output format, for example, a CSV file. Each row in a CSV file typically corresponds to one node in a hierarchy.
Add columns to the CSV file with properties that map each node to a corresponding dimension value. For details, see Record properties for all dimension values.
Implement a custom data source that can crawl the exported source data. You can implement your own data sources using the CAS Extension API. For details, see the Oracle Commerce CAS Extension API Guide.
Crawl the exported source data (i.e. the CSV file). CAS creates one Endeca record for each dimension value and by default, writes the records to a Record Store instance.
Here is a diagram of this example scenario: