Using GetCapturedArguments() on a variadic function with no variadic args causes a panic

[jpopadak]

Yes, those arguments make zero sense, but it works to cause the issue we are seeing in real code.
Source:

type InterfaceName interface {
	DoSomething(context.Context, ...context.Context)
}

Test:

//go:generate pegomock generate -m github.com/package/path InterfaceName

func TestInterfaceName(t *testing.T) {
	var interfaceMock health.InterfaceName = NewMockInterfaceName()
	interfaceMock.DoSomething(context.Background()) // Called with only one param

	mockInterfaceName := interfaceMock.(*MockInterfaceName)
	mockInterfaceName.VerifyWasCalledOnce().DoSomething(matchers.AnyContextContext()).GetCapturedArguments()
}

Turns out this is due to these generated lines always expecting at least 1 value passed into the variadic function: _param1 = make([][]context.Context, len(params[1]))
The code does a length check, but does not check to see if the data inside the params > the number of arguments minus the number of variadic.

Generated code:

func (c *MockInterfaceName_DoSomething_OngoingVerification) GetAllCapturedArguments() (_param0 []context.Context, _param1 [][]context.Context) {
	params := pegomock.GetGenericMockFrom(c.mock).GetInvocationParams(c.methodInvocations)
	if len(params) > 0 {
		_param0 = make([]context.Context, len(params[0]))
		for u, param := range params[0] {
			_param0[u] = param.(context.Context)
		}
		_param1 = make([][]context.Context, len(params[1]))
		for u := range params[0] {
			_param1[u] = make([]context.Context, len(params)-1)
			for x := 1; x < len(params); x++ {
				if params[x][u] != nil {
					_param1[u][x-1] = params[x][u].(context.Context)
				}
			}
		}
	}
	return
}

[petergtz]

@jpopadak You're absolutely right, that's a bug. Thanks for troubleshooting this already. I'll see if I can fix this within the next days.

[petergtz]

@jpopadak Just pushed a change on develop. Could you have a look please? If it works, I'll merge it into master and make a new release.

[petergtz]

@jpopadak Please let me know in case you still have trouble with this. Otherwise, I'll consider this fixed and will release it to master sometime next week.

[jpopadak]

Looks like this was fixed with 2.7.0. Thanks! :)