fix(instance): index out of range (warningServerTypeDeprecated)

[frivoire]

Fixes #5682

Given my (almost-zero) golang level, this PR is "vibe-coded".
But I reviewed it and Copilot's analysis looks right to me :

The bug is a classic Go slice mistake in warningServerTypeDeprecated in helpers_types.go:69-70:
make([]string, 0, 2) creates a slice with length 0 and capacity 2. Writing to warning[0] requires the length to be at least 1 — this panics for any server with EndOfService: true.

(my instance indeed has "end_of_service": true, in the API json)

And the details of the fix:

The PR changes it to use a slice literal instead. This is Go's composite literal syntax. It does three things in one line:

1. Creates a slice
2. Initializes it with the values you provide inside { }
3. Automatically sets both length and capacity to match the number of elements

So []string{ value1 } creates a slice with length=1, capacity≥1, already containing that first element.

Can someone validate this (or at least let the CI run on it) ? 🙏
Thanks !

[codecov-commenter]

Codecov Report

❌ Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 50.66%. Comparing base (113811f) to head (a5565e6).
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
internal/namespaces/instance/v1/helpers_types.go	0.00%	2 Missing ⚠️

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5683   +/-   ##
=======================================
  Coverage   50.66%   50.66%           
=======================================
  Files         340      340           
  Lines       78134    78134           
=======================================
  Hits        39583    39583           
  Misses      37055    37055           
  Partials     1496     1496           

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[pypaut]

Hello @frivoire, thank you for your pull request! I allowed the CI to run and there is a issue with the linting.

[frivoire]

Hello Geoffrey and thanks for CI 👍

there is a issue with the linting.

About the issue: I'm not sure to understand the issue... why is this an issue ?
Is it about performance ? From what I see here: For most programs usage of prealloc will be a premature optimization., so I'm doubting a bit.
Furthermore, today's code is using prealloc (IIUC), so the CI is pushing us towards a syntax that is bug-prone (cf the issue that I'm trying to fix) for potentially a non-existing perf question, so ... is it really the right tradeoff (for a CLI tool) ?

[pypaut]

I understand your point of vue. Though this kind of optimisation seems overkill yet, the code abides by it for now and enforcing it didn't cost much so far.

Could you try replacing the warning[0] with append for example?

[remyleone]

Could you provide us a snippet of code that reproduce the issue you observed? I would like to include a test to ensure that it is fixed for good

[frivoire]

@pypaut :

Could you try replacing the warning[0] with append for example?

Oki, I've rewritten the branch with this idea => the CI should be happy :)
And I've tested it locally (by building the cli + executing again the exact broken command) => no more issue

[frivoire]

Coucou @remyleone 👋

Could you provide us a snippet of code that reproduce the issue you observed? I would like to include a test to ensure that it is fixed for good

I gave the following command in the related issue:

scw instance server get bd7de168-a479-4793-aa99-ddc7609fe315

And I'm guessing that it's because my instance has "end_of_service": true in the API json (it's a START1-XS)