| G. Belle, C.B. Jay, and E. Moggi. Functorial ML. Submitted to PLILP 96, 1996. |
....shape, is given a meaning separate from the data that is stored in it [6, 8] For example, the shape of a matrix is given by its size, and the shape of tree is given by the corresponding unlabelled tree. It has lead to new paradigms for both parallel programming [7, 9] and functional programming [1]. These developments are analogous to those arising from our ability to handle data in isolation, which lead to parametric polymorphism, or generic data types. Such shape data divisions do not arise in standard OO program methodologies, such as OOAD [2] or MOSES [5] or formal type systems, e.g. ....
G. Belle, C.B. Jay, and E. Moggi. Functorial ML. Submitted to PLILP 96, 1996.
....is a tuple of three sizes. There are two broad streams of applications of shape theory. One is the development of shape polymorphic programming, in which programs, such as those that map a function across all the data in a structure, can be applied to inputs of many different shapes (see e.g. [BJM96]) The other is shape analysis, in which shape information is collected and exploited prior to handling the data (see e.g. JCE96] These two approaches apply in a variety of different computational paradigms [Jay96b] Shape analysis can be considered a form of partial evaluation (e.g. JGP93] ....
....a shape checker and analyser for an existing language, probably Nesl. Another direction is to extend the collection of types to which shape analysis applies, to include trees and other inductive types. Also, the shape type constructor # can be viewed as a functor in the sense of Functorial ML [BJM96] and the shape term constructor as a polymorphic combinator. ....
G. Belle, C.B. Jay, and E. Moggi. Functorial ML. Submitted to PLILP 96, 1996.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC