Control Structures

We now have one final touch before we can call our Lisp dialect a programming language: we must implement conditions and function definitions. Once we get this done we can write a program that will, for example, compute the Fibonacci sequence - a sure sign of a language approaching Turing completeness! Fortunately doing this is rather simple - all we have to do is add two more functions to our interpreter. We can implement conditional statements like this:

blaiseIf [condExpr, expr1, expr2] =
    do eval_cond <- eval condExpr
       if (0 `notEqual` eval_cond) then eval expr1
                                   else eval expr2
    where notEqual val1 (BlaiseInt val2) = val1 /= val2

Function definitions can be done in a similar manner:

blaiseFnArgs = ["args", "..."]
blaiseFn = do
    [(BlaiseList args), (BlaiseList body)] <- getSymbols blaiseFnArgs
    let newFn = do evalBody <- mapM eval body
                   return $ last evalBody
    return $ BlaiseFn newFn (map (\(BlaiseSymbol arg)->arg) args)

That's it! We can finally define fib to compute a Fibonacci sequence like so [3]:

(set fib (fn (n)
   (if (eq n 0) 1
       (if (eq n 1) 1
           (+ (fib (- n 1)) (fib (- n 2)))))))

We can now call fib as if it were any other function!

(fib 4)
(fib 10)

[3] Because the interpreter doesn't support newlines you have to type the fib function definition on one line for it to work.