We survey fundamental concepts in inverse programming and present the Universal Resolving Algorithm (URA), an algorithm for inverse computation in a first-order, functional programming language. We discuss the principles behind the algorithm, including a three-step approach based on the notion of a perfect process tree, and demonstrate our implementation with several examples. We explain the idea of a semantics modifier for inverse computation which allows us to perform inverse computation in other programming languages via interpreters.On leave from DIKU,

No abstract

No abstract

No abstract

An approach for systematically modifying the semantics of programming languages by semantics modifiers is described. Semantics modifiers are a class of programs that allow the development of general and reusable "semantics components". Language independence is achieved through the interpretive approach: an interpreter serves as a mediator between the new language and the language for which the non-standard semantics was implemented. Inverse computation, equivalence transformation and neighborhood analysis are shown to be semantics modifiers. Experiments with these modifiers show the computational feasibility of this approach. Seven modifier projections are given which allow the efficient implementation of non-standard interpreters and non-standard compilers by program specialization or other powerful program transformation methods.