www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.learn - No trace of cumulativeFold except in the source.

reply e-y-e <yurtqbqn grr.la> writes:
Recently I needed to use a cumulative sum function, so I looked 
in phobos' documentation for 'cumulative' but found nothing 
useful. Then I looked in the forums for it and found nothing 
useful. But when I searched phobos for it I found cumulativeFold 
in std.algorithm.iteration: 
https://github.com/dlang/phobos/blob/master/std/algorithm/iteration.d#L3127. So
I tried this:

auto cumulativeSum(Range)(Range r)
{
     import std.algorithm.iteration : cumulativeFold;

     return r.cumulativeFold!((a, b) => a +b);
}

but when I run it I get 'Error: module std.algorithm.iteration 
import 'cumulativeFold' not found'. Anyone can reproduce this? 
looking at the source I can't see how it could possibly occur but 
it is certainly a bug right?
Oct 23 2016
parent reply Jonathan M Davis via Digitalmars-d-learn writes:
On Sunday, October 23, 2016 07:46:19 e-y-e via Digitalmars-d-learn wrote:

 Recently I needed to use a cumulative sum function, so I looked
 in phobos' documentation for 'cumulative' but found nothing
 useful. Then I looked in the forums for it and found nothing
 useful. But when I searched phobos for it I found cumulativeFold
 in std.algorithm.iteration:
 https://github.com/dlang/phobos/blob/master/std/algorithm/iteration.d#L312
 7. So I tried this:

 auto cumulativeSum(Range)(Range r)
 {
      import std.algorithm.iteration : cumulativeFold;

      return r.cumulativeFold!((a, b) => a +b);
 }

 but when I run it I get 'Error: module std.algorithm.iteration
 import 'cumulativeFold' not found'. Anyone can reproduce this?
 looking at the source I can't see how it could possibly occur but
 it is certainly a bug right?


It's not a bug. It's just too new. You looked at the master branch on
github, whereas what you're probably using on your computer is 2.071.2,
which does not have cumulativeFold, because it was added at some point after
2.071.2.

- Jonathan M Davis
Oct 23 2016
parent reply e-y-e <yurtqbqn grr.la> writes:
On Sunday, 23 October 2016 at 09:11:08 UTC, Jonathan M Davis 
wrote:

 On Sunday, October 23, 2016 07:46:19 e-y-e via 
 Digitalmars-d-learn wrote:

 ...


 It's not a bug. It's just too new. You looked at the master 
 branch on github, whereas what you're probably using on your 
 computer is 2.071.2, which does not have cumulativeFold, 
 because it was added at some point after 2.071.2.

 - Jonathan M Davis


Ok, that checks out. I looked at the commit and it was in April 
so I just assumed it would be in the release by now.
Oct 23 2016
parent reply Jonathan M Davis via Digitalmars-d-learn writes:
On Sunday, October 23, 2016 10:10:40 e-y-e via Digitalmars-d-learn wrote:

 On Sunday, 23 October 2016 at 09:11:08 UTC, Jonathan M Davis

 wrote:

 On Sunday, October 23, 2016 07:46:19 e-y-e via

 Digitalmars-d-learn wrote:

 ...


 It's not a bug. It's just too new. You looked at the master
 branch on github, whereas what you're probably using on your
 computer is 2.071.2, which does not have cumulativeFold,
 because it was added at some point after 2.071.2.

 - Jonathan M Davis


 Ok, that checks out. I looked at the commit and it was in April
 so I just assumed it would be in the release by now.


Per

http://dlang.org/changelog/2.071.0.html

2.071.0 came out at the beginning of April, and 2.072 has been slow in
coming, so we've only had point releases since then, and a new function
would not be added with a point release. But yes, per the current release
scheme, normally, there would have been a new major release by now (it is
finally in beta though).

- Jonathan M Davis
Oct 23 2016
parent reply e-y-e <yurtqbqn grr.la> writes:
On Sunday, 23 October 2016 at 10:19:07 UTC, Jonathan M Davis 
wrote:

 On Sunday, October 23, 2016 10:10:40 e-y-e via 
 Digitalmars-d-learn wrote:

 ...


 Per

 http://dlang.org/changelog/2.071.0.html

 2.071.0 came out at the beginning of April, and 2.072 has been 
 slow in coming, so we've only had point releases since then, 
 and a new function would not be added with a point release. But 
 yes, per the current release scheme, normally, there would have 
 been a new major release by now (it is finally in beta though).

 - Jonathan M Davis


On this topic, do you think a 'cumulativeSum' specialisation 
based on the 'sum' specialisation be welcome in phobos? Here's a 
quick prototype, obviously not library standard but the basic 
logic is there: 
https://gist.github.com/anonymous/4fb79b4aba79b59348273288993740cb
Oct 23 2016
parent Jonathan M Davis via Digitalmars-d-learn writes:
On Sunday, October 23, 2016 19:20:43 e-y-e via Digitalmars-d-learn wrote:

 On this topic, do you think a 'cumulativeSum' specialisation
 based on the 'sum' specialisation be welcome in phobos? Here's a
 quick prototype, obviously not library standard but the basic
 logic is there:
 https://gist.github.com/anonymous/4fb79b4aba79b59348273288993740cb


If you think that it would be generally useful and worth having in Phobos,
by all means, get it up to snuff for inclusion in Phobos and create a PR for
it.

- Jonathan M Davis
Oct 27 2016