An implementation of ECL data types

  • Authors:
  • Benjamin M. Brosgol

  • Affiliations:
  • Center for Research in Computing Technology, Harvard University, Cambridge, Mass.

  • Venue:
  • Proceedings of the international symposium on Extensible languages
  • Year:
  • 1971

Quantified Score

Hi-index 0.00

Visualization

Abstract

The data-type definition scheme in ECL is designed to furnish its users with a natural notation, in which the composition and behavior of complex objects can be readily described, and which simultaneously produces efficient underlying representations. The purpose of this paper is to discuss how these objectives are met in the implementation of ECL on Harvard's PDP-10. Section II outlines the fundamentals of data definition in ECL. For a full discussion of the ECL programming system and its language component, the reader is referred to [1], [2], or [4]; an overview of the system is presented in a companion paper by Wegbreit [5]. A complete description of ECL's data extension mechanism is given in [1] and [3]. Section III presents the main features of the implementation of the ECL data definition scheme, discusses the “mode compiler”, analyzes the internal representation chosen for data objects, and describes the efficiencies resulting from this representation. Finally, section IV deals with some of the interfaces between the data extension mechanism and the rest of the ECL system, and section V proposes some areas where further research would be useful.