Unused function parameters

[Andrew Lytvynov]

Hello, gophers.

What is the reason behind allowing unused function parameters (as opposed to disallowing unused variables)?

Currently, this should work:

func dosomething(param string) {

    fmt.Println("param not used")

}

It feels like this should be treated the same way as unused variables are (compilation error). At least I can't imagine any case where it would break valid idiomatic code.

Thoughts?

[Andrew Gerrand]

Unused variables are always a programming error, whereas it is common to write a function that doesn't use all of its arguments.

One could leave those arguments unnamed (using _), but then that might confuse with functions like

func foo(_ string, _ int) // what's this supposed to do?

The names, even if they're unused, provide important documentation.

Andrew

--
You received this message because you are subscribed to the Google Groups "golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Andrew Lytvynov]

That's pretty much my proposal: only allow them unnamed (_).

I can imagine only 2 cases where they are used:

backward-compatibility for libs

implementing some interfaces

I can't see how any of those cases would be confusing with unnamed parameters.

I'd be glad to check any example of func which would be confusing without parameters' names, though.

[Ian Lance Taylor]

On Tue, Apr 30, 2013 at 6:34 AM, Andrew Gerrand <a...@golang.org> wrote:
> Unused variables are always a programming error, whereas it is common to
> write a function that doesn't use all of its arguments.

Think about methods in particular, that must take a certain set of
arguments in order to satisfy an interface.

Ian

[Andrew Gerrand]

It's a subtle issue. Not one that I feel strongly about, either.

In any case, we can't make the change for Go 1.x:

  http://golang.org/doc/go1compat.html

Andrew

[Andrew Lytvynov]

I agree on it being rather subtle, and i do realize it won't happen any time soon.

Just wanted to hear some opinions.

Thanks Andrew, Ian.

[Aaron France]

FWIW,

I think it *should* happen if only to be as strict as possibly you can be (since a lot of the other cases of "unused X" are covered) but I'm not sure of the gains.

--

[Alexei Sholik]

Erlang's approach is to treat "_names" similarly to "_" names. That is, it will argue about unused name, but not about unused _name. This make it possible to preserve the documenting qualities of the parameter names while also documenting the intent of not using the parameter deliberately.

--
Best regards
Alexei Sholik

[Andrew Lytvynov]

I don't think this is the best solution for such small issue. Too much effort for very little gain.