This is a read-only archive of lispforum.com. The forum was locked to new users and posts and is preserved here as static HTML from a database snapshot taken on 2019-09-07.

correct behaviour of nconc

3 posts · 4006 views

Martin Kalbfuß

#1 Fri Jul 24, 2015 12:44 pm

Hi guys,
I'm currently implementing my own lisp interpreter. At the moment I working on the function nconc. I'm not sure how it should behave for the following cases:

(nconc 1)

(nconc nil 1)

The gnu common lisp compiler returns 1 for both cases. Shouldn't the return value be a list. Isn't throwing an exception the better behaviour? Or is this behaviour forced by the standard?

regards,
Martin Kalbfuß

Goheeca

#2 Sat Jul 25, 2015 9:31 am

Re: correct behaviour of nconc

By the description section in CLHS you apply 2nd and 3rd rules.

[2] (nconc nil . lists) == (nconc . lists)
[3] (nconc list) => list

For your first code you apply the 3rd rule and

The last list may be any object.

That's:

(nconc 1) => 1

For your second code you apply the 2nd rule then the 3rd one:

(nconc nil 1) == (nconc 1) ;;; is equal to (nconc nil . (1)) == (nconc . (1))
(nconc 1) => 1

cl-2dsyntax is my attempt to create a Python-like reader. My mirror of CLHS (and the dark themed version). Temporary mirrors of aferomentioned: CLHS and a dark version.

edgar-rft

#3 Sat Jul 25, 2015 5:21 pm

Re: correct behaviour of nconc

While Goheecha's explanation is correct, an easier-to-understand explanation is that NCONC is the destructive version of APPEND, which is defined as:

CLHS APPEND wrote:... The last argument is not copied; it becomes the CDR of the final dotted pair of the concatenation of the preceding lists, or is returned directly if there are no preceding non-empty lists.

- edgar