Browse Prior Art Database

Data Base System for Program Source using a Stored Dictionary

IP.com Disclosure Number: IPCOM000080099D
Original Publication Date: 1973-Oct-01
Included in the Prior Art Database: 2005-Feb-26
Document File: 2 page(s) / 14K

Publishing Venue

IBM

Related People

Sharman, GCH: AUTHOR

Abstract

A system for holding on direct-access storage the source of large programs written in a high-level language such as PL/I is described. The system is based on the design of a segmented dictionary which is also stored on direct access. The dictionary is used by an access method which provides a method of modifying the program source, and also by a specially designed language compiler. The system avoids some of the interface problems commonly experienced in implementing large programs, by providing a record of the use of each variable. The system is suitable for interactive use with a terminal system.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Data Base System for Program Source using a Stored Dictionary

A system for holding on direct-access storage the source of large programs written in a high-level language such as PL/I is described. The system is based on the design of a segmented dictionary which is also stored on direct access. The dictionary is used by an access method which provides a method of modifying the program source, and also by a specially designed language compiler. The system avoids some of the interface problems commonly experienced in implementing large programs, by providing a record of the use of each variable. The system is suitable for interactive use with a terminal system.

Large programs are usually held on direct-access storage for reasons of physical convenience, and developed by applying successive updates to this source until a final state is achieved. In some cases the source is held in a single data set. This has the advantage that the entire program can be compiled in a single operation, during which any errors in the interfaces between procedures are detected. However, it necessitates close cooperation between programmers which can only be achieved in small teams.

Where large teams must be employed, it is usual to break up the source into individual procedures and store each procedure in a separate data set. A directory may be used to locate these data sets. This arrangement allows a number of programmers to access different parts of the source concurrently, which is especially necessary where a terminal system is used for making updates. However, it has the severe disadvantage that each separately stored procedure becomes an external procedure and must be separately compiled. The interfaces between such procedures must be separately declared in each procedure. Mistakes in these declarations cannot be detected by compilation and are a common cause of program errors. The problem is aggravated when there is a large number of procedures.

The system described comprises a data base for holding high-level language program source. The data base contains one data set per procedure and also a dictionary data set. The data base is created and updated by special purpose utility programs, or by a terminal system. Either of these is referred to as the "access method".

The dictionary serves a two-fold purpose. It acts as a catalogue, enabling a given procedure to be physically located and logically identified as a part of the program. It also serves as the compiler dictionary, specifying the environment of the procedure for compilation. The existence of this dictionary allows the access method t...