Para11eter Splitting in a Higher Order Functional Progra1111ing Language

This project is a part of a larger one consisting of constructing a Miranda to 'C' compiler. This compiler is to be constructed by first implementing a Miranda-interpreter in Scheme, then using a partial evaluator to generate a Miranda to Scheme compiler written in Scheme, and finally generating 'C'-code from the Scheme-code. Vve will develop a theory for doing parameter splitting in a higher order functional programming language based on a theory for doing parameter splitting in a first order functional programming language [Romanenko 90]. A reason for doing parameter splitting in connection with program specialization IS given the abstract of [Romanenko 90]. It states: Experiments on generating compilers by specializing specializers with respect to interpreters have shown that the compilers thus obtained have a natural structure only if the specializer does vm·iable splitting. Variable splitting can result in a residual program using several variables to represent the values of a single variable of the original program. In the case of functional programming, variable splitting is done be raising the a.rities of functions. The language to be transformed is to be a subset of Scheme. \Ve intend to implement a. parameter splitting program based on the theory we develop. The program will be implemented to work in a.n environment with Anders Bondorf's Similix-2 [Bondorf 90]. Instead of only handling the constructor cons and the selectors car and cdr we want the parameter splitting program to be able to handle (abstract) datastructures like Similix-2 does. The theory we are going to develop will of course reflect this. We will describe the theory in a paper. This is done in the hope that the paper might be published some time. The paper will be the main result of this project, but we will nevertheless describe the developed programs in a report.