Building Opaque Data Types

An opaque data type is an atomic, or fundamental, data type that you define for the database. The database server has no information about the opaque data type until you provide routines that describe it. As you build an opaque data type, you need to consider the following topics:

• How the information in the opaque data type is organized
• How to store and retrieve the data type
• What the standard operations mean with respect to the opaque data type:
  • What does it mean to add two pieces of data? Is it even possible to add the data?
  • When is one data item larger than another?
  • Can you relate this data to built-in data types?
• What unique operations this data has:
  • Does this data type allow you to find a picture?
  • Can you say that one data item is inside another?

Creating an Opaque Data Type, describes the basic steps for creating an opaque data type. Writing Support Functions, describes the support functions that an opaque data type uses.

Creating an opaque type and all of the routines that are required to support it is a major task. Theoretically, you could sit down and write all of the required routines. However, it is recommended that you use the IBM Informix DataBlade Developer's Kit (DBDK) because DBDK enforces standards that facilitate migration between different versions of the database server.

A DataBlade module is a group of database objects and supporting code that manages user-defined data or adds new features. A DataBlade module can include extended data types, routines, casts, aggregates, access methods, SQL code, client code, and installation programs. DataBlade modules that support various special-purpose opaque data types are provided. To find out what DataBlade modules are available, contact your sales representative.