Our Approach

The Semantic Framework simplifies and brings elegance to the complex and very often messy process of data integration application design. It employs reusable data objects referred to as Semantic Types that enable you to define, in business terms, how your data is represented within expressor Studio.

How Types Work

Semantic Type artifacts hold information such as field names and data types, which are automatically created and mapped to a schema artifact, which is a reusable piece of metadata that describes the external data layout when an external data source is read into expressor Studio. The auto-generated Semantic Type inherits the characteristics of the schema by default and enables you to modify it before it is utilized in your ETL application.

Semantic Types Deconstructed

Local Composite Type – A local type is an artifact that is auto-generated when you read a new data source into expressor Studio. A local type represents an internal data structure that is mapped to a schema artifact, which defines the external data structure of the data source or target. Local composite types are local to the schema of the source data and cannot be referenced by any other schema or artifact within expressor Studio.

Shared Composite Type – A shared type can be a local type that has been promoted to shared type status, or it can be a generic structure that can be defined within the Studio application. The benefit of this type of internal data structure is that it can be reused across many external data sources and targets without affecting any of the internal ETL logic written against it.

Shared Atomic Type – An atomic type is the fundamental building block of Semantic Types and represents a single attribute that can be utilized in one or more composite types.

Benefits

Semantic Types are reusable, abstracted, internal data structures that can be mapped to multiple data sources or targets. They enable external data types to automatically adapt to the internal data types defined in the Semantic Type, which enables the Semantic Framework to insulate expressor ETL applications from any schema change that might occur in a data source or target.

Semantic Types represent a reusable abstract internal data structure that can be user defined or automatically generated from a data source schema. You can further extend the capabilities of Semantic Types by adding constraints to each attribute of a Semantic Type.

How Constraints Work

Constraints are defined on individual attributes of a composite type. Each attribute carries the following information as part of its definition:

• The name of the attribute should be descriptive for human consumption.
• A primitive data type is assigned to the attribute (supported data types: string, integer, decimal, double, date-time, byte).
• Constraints are sets of rules that the data represented by the attribute must conform to based on data type (example constraints: minimum value, maximum value, minimum, regular expressions, allowed values).

Once a constraint is defined, you can specify the corrective action that should be taken if the rule is violated. By default all corrective actions are set to escalate errors to the dataflow, and you are able to reject, skip, or redirect data and take appropriate actions. You are also able to correct data right within the constraint definition by providing default values or other available actions based on the attribute data type.

Benefits

Constraints are applied to atomic types or attributes of composite types and enable the constraint information to be carried forward wherever a given attribute is used. This streamlines the development process by decoupling data cleansing logic from any actual transformation that needs to occur to the data. All constraint-related logic is centralized and managed within the Semantic Type artifact enabling easier maintenance and artifact reusability.

What makes the Semantic Framework powerful is the fact that it is there if you need it and transparent if you don’t. If you decide to take advantage of its powerful capabilities, you can further simplify the typically highly-complex data mapping process.

The Semantic Framework approach replaces the point-to-point mapping done in traditional ETL tools with a unique, streamlined mapping structure achieved through the use of Semantic Types. By mapping source and target datasets to business data objects, expressor automatically reconciles data type discrepancies.

expressor Studio avoids labor-intensive point-to-point mapping and allows you to leverage Semantic Types to simplify data mapping.

Benefits

The expressor simplified mapping approach enables you to reuse all of your Semantic Type artifacts as many times as necessary—depending on the needs of your ETL application. You can also take full advantage of your Semantic Type definitions as well as all of your constraint definitions. This approach ensures that all changes to your Semantic Type definitions are automatically propagated through your data integration applications.