scispace - formally typeset
Search or ask a question

Showing papers by "Peter Buneman published in 1986"


Journal ArticleDOI
15 Jun 1986
TL;DR: This paper examines whether it is necessary to the together type, extent and persistence in order to model inheritance and suggests that they may be separated to provide more general database programming languages.
Abstract: In order to represent inheritance, several recent designs for database programming languages have made use of class construct, which can be thought of as a restricted data type with an associated set of instances. Moreover, these classes are persistent they survive from one program invocation to another. This paper examines whether it is necessary to the together type, extent and persistence in order to model inheritance and suggests that they may be separated to provide more general database programming languages. In particular we shall see that it is possible to assign a generic data type to a function that extracts all the objects of a given type in the database so that the class hierarchy can be derived from the type hierarchy. We shall also examine object-level inheritance and its relationship to data types for relational databases. A final section examines how the various forms of persistence interact with inheritance at both object and type level.

59 citations


Book ChapterDOI
08 Sep 1986
TL;DR: It is shown that the notion of inheritance leads to a natural representation of the operators of the relational algebra and that some of the basic properties of relational database theory, such as Armstrong's axioms, can be derived from some very simple domain theoretic relationships.
Abstract: We have tried to show that the notion of inheritance leads to a natural representation of the operators of the relational algebra and that some of the basic properties of relational database theory, such as Armstrong's axioms, can be derived from some very simple domain theoretic relationships. If these ideas have any value one would expect to be able to represent other notions in database theory, such as multi-valued dependencies and the universal relation assumption within the same framework. However, given the apparent connection with Scott's “Information Systems”, a more pressing need is to work out a proper denotational semantics for relational databases.

15 citations