Re: remove sublist

[Emanuel Berg]

Yuri Khan wrote:

> Maybe
>
>     (cl-set-difference '("0" "a" "b" "c" "d")
>                        '("a" "b")
>                        :test 'string=)
>
> ?

Excellent, thanks!

You got the right idea, "sublist" as in the
math unordered set sense - and not "set" as in
the serpent god of the paleo-pharaohs, here!

And items in the sublist that do not appear
should be dropped, as happens.

Should the test function be hash-quoted?

Only strange thing is the order is reversed,
except for the empty list as sublist! I just
said it was in the unordered sense, however one
might just as well keep the order, especially
since even/particularly in the math unordered
set sense, {1 2 3} = {3 2 1}, right? So it
should work both ways; and it might be useful
for example if you have a list of functions,
(f1 f2 f3 f4), and it is a priority list so if
f1 is defined, run it, if not, try/run f2,
etc., here, if the user dislikes f2 and f3, but
have them defined for other purposes, s/he can
remove them from the list without changing
the priority!

However how is that supposed to be done?
What about

    (defun drop-sublist (l sl)
      (if sl
          (reverse (cl-set-difference l sl :test #'string-equal))
        l))
    ;; (drop-sublist '("0" "a" "b" "c" "d") '("a" "b" "x"))
    ;; (drop-sublist '("0" "a" "b" "c" "d") '(x))
    ;; (drop-sublist '("0" "a" "b" "c" "d") '())

?

It sort of goes against grain to reverse to
*preserve* the order, but I suppose this is
what makes programming an engineering and not
a scientific endeavor...

-- 
underground experts united
http://user.it.uu.se/~embe8573