Abstract
dc:description.abstractThe basic premise of this thesis is that extensible languages afford the user considerable power and flexibility. We argue that this flexibility can, and should, be provided in a secure and error-resistant manner, but that this objective is not realised in existing extensible languages. This thesis first investigates the nature of security in programming languages, building up a simple and informal theory of the design of secure languages, and relating this theory to the notions of structured programming and .transparency. We use this theory to build a conceptual model for a secure extensible language and its physical realisation. We show that existing extensible languages fail to meet the ideals of this model in total, and proceed to design an alternative and secure system which builds upon, but attempts to avoid the pitfalls of existing systems. We base this system on a string processing language (Snip) which is itself extensible. The remainder of this thesis discusses the design and implementation (based on an abstract machine, SAM) of this language.
Degree
thesis:*- Level dc:type.qualificationlevel
- PhD
- Grantor dc:publisher.institution
- University of Glasgow
- Year dc:date.issued
- 1975
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Morton, Richard Peter McE
Rights
- Language dc:language
- en