FME considers a Teradata dataset to be a database containing a collection of relational tables, and a table to be an FME feature type with each row corresponding to one FME feature.
Table schemas must be defined in the FME mapping file before they can be read and can consist of both geometric and primary column types, however, only one geometry column is read per table while others are discarded.
The basic reading process involves opening a connection to the database, querying metadata, and querying data. The data is read by submitting SQL queries and parsing the returned result sets.
If null geometries are read, they are treated as non-geometry features and the attributes are preserved.