Re: lengths and stuff

[Eric Abrahamsen]

Jean Louis <bugs@gnu.support> writes:

> * Tomas Hlavaty <tom@logand.com> [2020-12-27 23:53]:
>> On Sun 27 Dec 2020 at 10:52, Drew Adams <drew.adams@oracle.com> wrote:
>> >> Using length in a predicate is yet completely different and most
>> >> likely bad because to answer the predicate, traversing the whole list
>> >> is wasted time and energy.
>> >
>> > (lambda (xs) (= (length xs) 25))  ; need to count elts
>> 
>> Is that a joke?
>> 
>> The predicates were proposed to help programers avoid writing such bad
>> code.  And if somebody writes such bad code, it is almost trivial to fix
>> it with search and replace:
>> 
>>    "(= (length" -> "(length="
>>    "(< (length" -> "(length<"
>>    "(> (length" -> "(length>"
>>    "(/= (length" -> "(length/="
>>    "(<= (length" -> "(length<="
>>    "(>= (length" -> "(length>="
>> 
>> Also the "symmetry" (or exhaustiveness?) here is to assist in easily
>> fixing bad code with minimum changes.
>> 
>> Nothing of course helps to those who are determined to unneccessarily
>> count _all_ the elements of lists.
>
> May I understand if that proposal is to implement it in C or in Lisp?
> Does that mean when (length< list-1 10) is implemented in C it would
> become faster than: (< (length list-1) 10) ?
>
> Or is that proposal to make Lisp functions like:
>
> (defun length< (list n)
>   (when (< (length list) n)
>   t))

They're already implemented in Emacs master, and they're in C. As I
understand it, one of the main advantages (beyond being simply faster,
possibly) is that they return early: if all you want to do is check if a
list has more than 2 elements, you don't have to calculate the whole
length, you can return t once you've found the third element.