Would it be possible to provide "return-from" in uLisp?


Yet another proposal: Is it feasible to implement “return-from” in uLisp? This CommonLisp feature seems to make life a lot easier since it eliminates the need to construct functions/blocks that always “culminate” in the one and only single return. I’m not sure, though, if this request shows I haven’t really understood or embraced Lisp yet… currently, I’m having a really hard time to port a function that uses several different “exits” according to changing circumstances within the function block.

What would you like to see in uLisp in 2024?

I’m not a fan of return-from, and you can usually achieve the same effect more elegantly using cond.

For example, suppose you have three algorithms for solving a problem: special-case, which is fast but only works on some cases, quick-solution, which works on most cases, and brute-force, which is slow but always works. They each return t if they succeed in solving the problem. Then in Common Lisp one might write:

(defun solve-it (problem)
  (when (special-case problem) (print "Special case OK!") (return-from solve-it))
  (when (quick-solution problem) (print "Quick solution OK!") (return-from solve-it))
  (when (brute-force problem) (print "Brute force OK!") (return-from solve-it))
  (print "Something went wrong"))

However, it’s better to use cond:

(defun solve-it (problem)
   ((special-case problem) (print "Special case OK!"))
   ((quick-solution problem) (print "Quick solution OK!"))
   ((brute-force problem) (print "Brute force OK!"))
   (t (print "Something went wrong"))))