Table of Contents
Installment published 21st May 1989
In the last installment, I showed you the general idea for the top-down development of a compiler. I gave you the first few steps of the process for compilers for Pascal and C, but I stopped far short of pushing it through to completion. The reason was simple: if we're going to produce a real, functional compiler for any language, I'd rather do it for KISS, the language that I've been defining in this tutorial series.
In this installment, we're going to do just that, for a subset of KISS which I've chosen to call TINY.
The process will be essentially that outlined in Installment Chapter 9, A Top View, except for one notable difference. In that installment, I suggested that you begin with a full BNF description of the language. That's fine for something like Pascal or C, for which the language definition is firm. In the case of TINY, however, we don't yet have a full description … we seem to be defining the language as we go. That's OK. In fact, it's preferable, since we can tailor the language slightly as we go, to keep the parsing easy.
So in the development that follows, we'll actually be doing a top-down development of both the language and its compiler. The BNF description will grow along with the compiler.
In this process, there will be a number of decisions to be made, each of which will influence the BNF and therefore the nature of the language. At each decision point I'll try to remember to explain the decision and the rationale behind my choice. That way, if you happen to hold a different opinion and would prefer a different option, you can choose it instead. You now have the background to do that. I guess the important thing to note is that nothing we do here is cast in concrete. When you're designing your language, you should feel free to do it your way.
Many of you may be asking at this point: Why bother starting over from scratch? We had a working subset of KISS as the outcome of Installment Chapter 7, Lexical Scanning. Why not just extend it as needed? The answer is threefold. First of all, I have been making a number of changes to further simplify the program … changes like encapsulating the code generation procedures, so that we can convert to a different target machine more easily. Second, I want you to see how the development can indeed be done from the top down as outlined in the last installment. Finally, we both need the practice. Each time I go through this exercise, I get a little better at it, and you will, also.
Many years ago there were languages called Tiny BASIC, Tiny Pascal, and Tiny C, each of which was a subset of its parent full language. Tiny BASIC, for example, had only single-character variable names and global variables. It supported only a single data type. Sound familiar? At this point we have almost all the tools we need to build a compiler like that.