unofficial mirror of emacs-devel@gnu.org 
 help / color / mirror / code / Atom feed
From: Ken Raeburn <raeburn@raeburn.org>
To: John Wiegley <jwiegley@gmail.com>
Cc: Emacs development discussions <emacs-devel@gnu.org>
Subject: Re: Any interest in a function like this to add to subr.el?
Date: Wed, 19 Oct 2016 01:01:36 -0400	[thread overview]
Message-ID: <3D8D2EC8-5492-4A3B-8766-E27C17C8537A@raeburn.org> (raw)
In-Reply-To: <m27f954af9.fsf@newartisans.com>


> On Oct 18, 2016, at 17:25, John Wiegley <jwiegley@gmail.com> wrote:
> 
>>>>>> "DG" == Dmitry Gutov <dgutov@yandex.ru> writes:
> 
> GD> Then I agree that we want a function like that. We'd want a destructive
> GD> version of it as well, though.
> 
> Yes, good idea!  How about something like this:
> 
> (defun sort-on* (seq predicate accessor)
>  "Sort SEQ using PREDICATE applied to values returned by ACCESSOR.
> This is a destructive version of `sort-on', which attempts to
> reuse storage as much as possible."
>  (let ((seq2 seq))
>    (while seq2
>      (setcar seq2 (cons (funcall accessor (car seq2)) (car seq2)))
>      (setq seq2 (cdr seq2))))

Is there a shorthand function for this bit?  I.e., mutate every element of the list according to some supplied function, but reusing the same list storage.  If so, the body of sort-on* becomes about as short as your original sort-on.

Logically it’d be a variant of mapcar but “mapcar*” is something else…

>  (setq seq (sort* seq #'(lambda (x y) (funcall predicate (car x) (car y)))))

Is there any benefit to sort* instead of sort here?  As far as I can see, sort* just calls sort after (1) checking whether the list we just iterated over is a list, and (2) checking whether we supplied a “:key” keyword, which we didn’t.

Some other observations:

“sort” and “sort*” will work on vectors, too, returning the now-sorted vector; your sort-on, using mapcar, will make it into a list, and sort-on* will fail trying to apply car and cdr to vectors.  Were these intended to work on lists only?

As far as your doc strings go, you say sort-on will apply the accessor “at most once per element in the list”, but it’s going to be exactly once per element, isn’t it?

The “cdr” pass before returning could walk the intermediate sorted list in the sort-on case too, making it more space-efficient.  Since the inner “mapcar” call has allocated us some unshared storage, we can just alter it in place and return it to the caller.

Ken


  reply	other threads:[~2016-10-19  5:01 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-10-18 17:53 Any interest in a function like this to add to subr.el? John Wiegley
2016-10-18 20:13 ` Dmitry Gutov
2016-10-18 20:34   ` John Wiegley
2016-10-18 21:09     ` Dmitry Gutov
2016-10-18 21:25       ` John Wiegley
2016-10-19  5:01         ` Ken Raeburn [this message]
2016-10-19  6:02           ` John Wiegley
2016-10-19 14:51         ` Stefan Monnier
2016-10-19 14:48       ` Stefan Monnier
2016-10-19 23:12         ` Dmitry Gutov
2016-10-20  1:48           ` Stefan Monnier
2016-10-19 13:46 ` Andy Moreton
2016-10-19 14:52   ` Stefan Monnier

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.gnu.org/software/emacs/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3D8D2EC8-5492-4A3B-8766-E27C17C8537A@raeburn.org \
    --to=raeburn@raeburn.org \
    --cc=emacs-devel@gnu.org \
    --cc=jwiegley@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://git.savannah.gnu.org/cgit/emacs.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).