Re: How to perform concat()s without errors on empty vectors or conversions on single-element vectors??

["Ruud H.G. van Tol" <rvtol@isolution.nl>]

On 2024-02-02 06:34, Joe Slater wrote:
> I want to convert a vector whose members may be vectors into one long 
> vector. This can be done using concat(), but I have two problems:
> 1) it fails when the vector is empty, rather than returning the empty 
> vector; and
> 2) when the vector has a single element it returns that element rather 
> than a vector.
>
> I.e., I input:
> (16:29) > v=[[1,2],[1,[2,3]],[[1,2],[3,4]],[1],[]]; 
> for(x=1,#v,print(concat(v[x])))
> [1, 2]
> [1, 2, 3]
> [1, 2, 3, 4]
> 1
>   ***   at top-level: ...[1],[]];for(x=1,#v,print(concat(v[x])))
>   *** ^--------------
>   *** concat: domain error in concat: vector = []
>   ***   Break loop: type 'break' to go back to GP prompt
> What I want to get is the following, with no errors:
> [1, 2]
> [1, 2, 3]
> [1, 2, 3, 4]
> [1]
> []
>
> I can do this with explicit testing, but it seems to me there should 
> be a more natural way.

This works for your example:

? {
  my(v=[[1,2],[1,[2,3]],[[1,2],[3,4]],[1],[]]);
  [ if("t_VEC"==type(t) && #t>1, concat(t), concat([],t)) |t<-v ]
}
% [[1, 2], [1, 2, 3], [1, 2, 3, 4], [1], []]


but likely needs to go recursive if the shape is hairier, like:

? {
  my(v=[[1,2],[1,[2,3]],[[1,2],[3,4],[5,[6,7]]],[1],[]]);
  [ if("t_VEC"==type(t) && #t>1, concat(t), concat([],t)) |t<-v ]
}
% [[1, 2], [1, 2, 3], [1, 2, 3, 4, 5, [6, 7]], [1], []]


So create some flatten(v) function?

-- Ruud