Data manipulation language requirements for database management systems

In the July 1972 issue of Datamation , Schubert (1972) discusses the database management capabilities of the April 1971 report of the CODASYL Data Base Task Group (DBTG) (1971). The April 1971 report has been accepted by the CODASYL Programming Language Committee for inclusion of the COBOL data mani...

Full description

Bibliographic Details
Published in:The Computer Journal
Main Authors: Parsons, R. G., Dale, A. G., Yurkanan, C. V.
Format: Text
Language:English
Published: Oxford University Press 1974
Subjects:
DML
Online Access:http://comjnl.oxfordjournals.org/cgi/content/short/17/2/99
https://doi.org/10.1093/comjnl/17.2.99
Description
Summary:In the July 1972 issue of Datamation , Schubert (1972) discusses the database management capabilities of the April 1971 report of the CODASYL Data Base Task Group (DBTG) (1971). The April 1971 report has been accepted by the CODASYL Programming Language Committee for inclusion of the COBOL data manipulation language (DML) and its associated subschema into JOB COBOL. A new CODASYL committee has been established to develop the DBTG data description language (DDL) into a common data description language suitable for a complete designation of all the elements of a database which can then be assessed by extensions to existing programming languages and possibly a high-level self-contained language. It is the purpose of this paper to outline extensions to the proposed DML to supply a language capability ( a ) that is specifically designed to clarify and facilitate the writing of algorithms for processing hierarchical data structures, and ( b ) that provides facilities for manipulating structural data. We will show that the process of accessing the database on the bases of arbitrary Boolean functions of retrieval terms should not be left to the capabilities of the host language of the DML but should be made a part of the DML. In addition, the capability of the DML, as presently defined, to access a database on the basis of the values of data items is too restrictive and extensions to the present capabilities are proposed.