In addition to documentation data (Predict objects), XRef data is written for implemented objects to document what has actually been implemented for an application. Active retrieval functions evaluate both documentation data and XRef data to determine
if objects documented in the dictionary are not yet implemented,
if implemented members are not yet documented or
if entry points of implemented members are not yet documented or
if documentation data differs from the implementation.
Inconsistent use of field definitions can be found, along with answers to question such as "Which programs refer to file ABC?".
Active retrieval functions evaluate XRef data for objects in all libraries.
This document covers the following topics:
XRef data can be created in any of the following cases:
By cataloging a Natural object with the parameter with XREF set to Y.
By processing a 3GL program with Adabas Native SQL or the Predict Preprocessor.
By generating a DB2 request module with the Natural for DB2 function Create DBRM.
By creating a Predict system program object of type E (external object) and language Z (system program) documenting a program that is only available as object code. The implementation pointer for a system program must be specified explicitly. One entry point (with the ID of the program object) is created by Predict, additional entry points have to be specified manually.
An implemented member is a program object for which XRef data has been created with any of the options listed above.
For a general description of how XRef data is created and used, see the section Overview of Predict in the Introduction to Predict documentation.
See the sections LIST XREF for Natural and LIST XREF for Third Generation Languages for a more detailed description of the structure and contents of XRef data, and functions to retrieve information from XRef data.