Re: Setting the nth element of a list

[pjb@informatimago.com (Pascal J. Bourguignon)]

Nordlöw <per.nordlow@gmail.com> writes:

> I really believe such a function increases the readability of the code
> that uses it, don't you?
> What if I define it as a macro (defmacro) or inline (defsubst)?
> Doesn't that give the best of both worlds+

Readability is somewhat in the eyes of the reader.

That said, I can just repeat wat I wrote: setf is a standard and well
known operator designed to assign a value to any kind of place, so it
will be understood readily by any lisp programmer.

Defining redundant and specific operators just give you more work --
if you change the type of a sequence between list and vector, you'll
have to edit all the place where you used setnthcar or setnthref
instead of (setf (elt ...) ...) -- and it will give more work to the
reader who will have to refer to your functions definitions to know
what they mean.



If you were talking about a new kind of data structure, I'd agree that
it would be ok to define a setter,  but then it would have to be
defined with the arguments in the 'correct' order:

   (defun set-my-own-data-structure (new-value data-structure index1 ...) ...)

so you could bind it to setf:

   (defsetf my-own-data-structure set-my-own-data-structure)

and right away forget about set-my-own-data-structure, but write:

   (setf (my-own-data-structure my-ds i...) new-value)


-- 
__Pascal Bourguignon__