Rule based programming in C
Charles Paul Kosta · Proceedings of the 17th conference on ACM Annual Computer Science Conference · 1989
Recently there has been a trend to describe computer functionality using rule based paradigms. This is done to conform to Expert Systems shells and to assist rapid prototyping tools. To facilitate and continue this trend onto platforms that do not as yet have stable expert systems shells (transputers, hypercube, others) it would be useful to have a rule based paradigm that could be compiled and executed using a C compiler.It is possible to develop a small collection of '#define' statements that will look and act much like a ruled based system; yet, be converted to actual C compilable structures. The most useful feature is that both Forward and Backward chaining principles can be combined into the rule descriptions. Much of the functionality of an 'OPS--like' language, and all of the features of a Prolog-like language can be generated in this manner. Additionally, the '#define' statements can be modified to allow for tracing and debugging features without having to change any of the rule base.The present instantiation in C does lack some features; such as, dynamic loading of new rules and the absence of any object--oriented data structures. However, these can be added - to form useful rule based programming in C.