On Friday, November 8, 2024 4:27:40 PM MST Dom DiSc via Digitalmars-d-learn wrote:
> I _very_ often use this pattern:
>
> ```
> fun(ref int[] a)
> {
>     assert(a.length && a.length<=100);
>     int[100] b;
>     b[0 .. a.length-1] = a[];
>     b[a.length .. 100] = 5;
> }
> ```
>
> I consider this perfectly safe, but DScanner gives warnings for
> this, no matter if a check is present or not.
> What's the recommended fix for this? I have no idea what else to
> use.

Well, I don't know what it's warning about, but the code is wrong, because the lengths of the slices don't match in

    b[0 .. a.length-1] = a[];

You'll get a RangeError being thrown when you run the code.

- Jonathan M Davis

On Friday, 8 November 2024 at 23:27:40 UTC, Dom DiSc wrote:
> ...

My version:

import std.stdio, std.array, std.range, std.algorithm;

void foo(ref int[] a, int filler){
   int[5] b;
   auto len = min(b.length,a.length);
   b[0..len] = a[0..len];
   b[len..5] = filler;
   writeln("a = ", a);
   writeln("b = ", b);
}

void main(){
   int[] a;

   a = iota(2).array;
   foo(a, 77);

   writeln("");

   a = iota(11).array;
   foo(a, 77);
}

output:

a = [0, 1]
b = [0, 1, 77, 77, 77]

a = [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
b = [0, 1, 2, 3, 4]

Matheus.

On Saturday, 9 November 2024 at 04:02:46 UTC, Jonathan M Davis wrote:
[...]
>     b[0 .. a.length-1] = a[];
>
> You'll get a RangeError being thrown when you run the code.
[...]

Therefore a fix might be to not deminish the length:
    b[ 0 .. a.length]= a[];

-manfred

No, no. My problem is: I want to access the last element of an array and assign it to the element of same index in another array (so I can't use $), and I checked that the source array has at least one element.

The questions are:
- Why is it considered dangerous to use the expression a.length-1 ?
- What else is proposed to implement the desired effect without warning?