Pathmaker Programming Guide

Defining Data for a Pathmaker Project
Preparing for Pathmaker Application Development
067868 Tandem Computers Incorporated 2–39
For each unique index that is efficient, one unique access path is created on all the
columns of the index.
For each nonunique index that is efficient, one nonunique access path is created on
all columns of the index.
No default access paths are created for shorthand views.
Note The columns included in the default access paths are based upon the SQL indexes at the time a
Pathmaker table object is added to a Pathmaker catalog. The Pathmaker product does not track
subsequent changes to a table’s indexes. You are responsible for keeping access path definitions
consistent with the corresponding indexes. To do so, you can manually update the access path definition
or you can delete and re-add the tables that have changed.
Adding Access Paths. You can add new access paths, delete existing access paths, or add
or delete columns from existing access paths. Access paths need not correspond to
existing indexes on a table, even though the initial default paths do so.
The processing requirements of your application will determine if you need to change
an existing access path or create new access paths. Here are a few suggestions for
determining the access paths an application requires:
Identify each SELECT operation on a single table. Is there a unique access path for
each case? If not, you can perform either of these actions:
Add a unique index to the appropriate table using SQLCI. Then either:
Delete and re-add the SQL table object associated with that table on the
SQL Table Registration screen and use the resulting default access path for
the index you added. (This approach will delete any access paths that you
have added manually for this table.)
Add a new access path composed of the columns of the new unique index
you added and set the default field to Y. (This approach will not delete
access paths that you have added manually for this table.)
Use the Table Access Paths screen to define a new access path for the table,
using columns that, together, constitute a unique key. You must enter Y in the
Unique field and in the Default field for this new access path. (If the columns
you have indicated for this access path do not constitute a unique key, the
SELECT statement will produce a run time error.)
Identify each FETCH operation on a single table. Is there an access path for each
case? (Although you can use an access path that the Pathmaker product does not
consider efficient for a FETCH operation, be aware that doing so could adversely
affect the performance of your application.)
To create an efficient access path for a FETCH operation, you use SQLCI to add an
index that meets the Pathmaker criteria for efficiency as described in the
subsection “Access Path Attributes” to the appropriate table. Then either: