[Bug] Script crashes with "bad array subscript" when service_mesh input is empty

[PragalvaXFREZ]

Description

The mesheryctl.sh script crashes with a bash error when the service_mesh input is not provided, even though it's marked as optional in action.yml.

Error

mesheryctl.sh: line 61: adapters: bad array subscript

Root Cause

In mesheryctl.sh, when using a config file (the else branch starting at line 58), the script immediately accesses:

shortName=$(echo ${adapters[$service_mesh]} | cut -d ':' -f1)

If service_mesh is empty (not provided by the caller), bash throws "bad array subscript" because you cannot access an associative array with an empty key.

Steps to Reproduce

1. Call the action with profile_filename but without service_mesh:

- uses: layer5io/meshery-smp-action@master
  with:
    provider_token: ${{ secrets.PROVIDER_TOKEN }}
    platform: docker
    profile_name: test-profile
    profile_filename: test-config.yaml
    # service_mesh not provided

2. The script enters the else branch (line 58) because perf_filename is set
3. Line 61 crashes because $service_mesh is empty

Impact

This bug is causing the meshery/meshery scheduled e2e tests to fail continuously. The workflow at .github/workflows/mesheryctl-e2e.yaml calls this action without providing service_mesh, triggering this crash.

See: https://github.com/meshery/meshery/actions/workflows/mesheryctl-e2e.yaml

Proposed Fix

Wrap the adapter array access in a conditional check:

if [[ -n "$service_mesh" ]]; then
    shortName=$(echo ${adapters[$service_mesh]} | cut -d ':' -f1)
    shortName=${shortName#meshery-}
    if [[ -n "$shortName" ]]; then
        docker network connect bridge meshery_meshery-"$shortName"_1 2>/dev/null || true
        docker network connect minikube meshery_meshery-"$shortName"_1 2>/dev/null || true
    fi
fi

I have a PR ready to submit with this fix.

Environment

• Action version: master branch
• Triggered from: meshery/meshery scheduled e2e workflow

[PragalvaXFREZ]

@lekaf974 Any thoughts regarding the proposed fix?

[lekaf974]

@lekaf974 Any thoughts regarding the proposed fix?

I don't have to check this part, let me take some time to understand what ths script is doing, I'll come back to you soon