COBOL report program generation by macro processor

Stevens, Sheila (1981). COBOL report program generation by macro processor. PhD thesis The Open University.

DOI: https://doi.org/10.21954/ou.ro.00010108

Abstract

The research described is directed towards the development of an interactive COBOL program generator. This aims to simplify computer usage and contribute to the more effective use of computer resources, by alleviating the burdensome, time-consuming and error-prone activity of programming.

The work is based on PG/1, an existing macro processor, and in particular considers the generation of report programs.

The problems of program generation by macro processor are first explored using Filetab, an existing report specification language,' and avoiding the interactive aspect. This establishes the feasibility of generating COBOL programs but identifies shortcomings in the macro processor facilities. Enhancements to the macro processor are defined and the PG/2 version is created
for use.

The development of an interactive self-instructional COBOL report program generator for the casual user is chosen as the area for deeper investigation.

From a consideration of data bases and query languages it is concluded that a relational view of the data forms the most natural basis for a computer-dominated, non-procedural report-specifying dialogue with a casual user.

A design study for an interactive self-teaching COBOL report program generator^ based on the PG/2 macro processor and offering some of the benefits of a Relational Data base Management System, is described. The validation of the user's responses is a major undertaking and leads to a large system. Details of the macros implementing selected features are presented.

It is concluded that program generation appears efficient but may be of even greater use to a professional programmer than to the non-specialist user. The PG/2 macro processor, although not an ideal tool for COBOL program generation, serves to identify the characteristics of an appropriate tool. There is potential for extending and adapting the techniques used to other applications.