Seamless Use of SQL

Home | Previous Page | Next Page Access Methods > User-defined Access Methods >

Seamless Use of SQL

With the aid of a user-defined secondary access method, an SQL statement can use one or more indexes.

Further, with the aid of a user-defined secondary access method, indexes can provide access to the following extended data:

User-defined types
Data inside a smart large object
External data sources
Nonrelational data

In addition, with the aid of a user-defined secondary access method, an index can contain any of the following key types:

Return values from a user-defined function
Approximate values such as stem words for a full-text search
Attributes of data such as length
Relative position to other data in a hierarchy or area of space

The end user can use SQL to access both IBM Informix data and virtual index data. A virtual index requires a user-defined access method to make the data in the index accessible to Dynamic Server. In Figure 1, a single application processes IBM Informix data as well as virtual data in an external location and smart-large-object storage.

Figure 1. An Application Using a Secondary Access Method

Flowchart shows that a client program can access an Informix table via an SQL engine and a built-in secondary access method through a B-tree index. Alternatively, the client program can access an Informix table via an SQL engine and a user-defined secondary access method through a virtual index stored either in an extspace or an sbspace.