Indexing principles for relational languages applied to PROLOG code generation

Abstract

In this paper we propose an extensible, flexible, multi-argument indexing technique for relational languages. We present a compiler producing indexing header code for a PROLOG emulator based on the Warren Abstract Machine. We will show that our technique combines positive aspects of relational database methods and other existing WAM-based indexing schemes. All the indexing concepts introduced are implemented in LISP for the relational-functional programming language RELFUN.