labelled fallthrough?

[Manlio Perillo]

The go spec, at https://golang.org/ref/spec#ExprSwitchStmt, says:

`In a case or default clause, the last non-empty statement may be a (possibly labeled) "fallthrough" statement to indicate that control should flow from the end of this clause to the first statement of the next clause`.

This is misleading; and to increase confusion there are no example of fallthrough or "labeled" fallthrough.

Today I got the "The Go Programming Language", and there are no examples here, too (if the book index is correct).

I say it is misleading, since following the text I tried to do:

    fallthrough label

but this is not supported by the grammar.

The correct syntax if

    goto label

as it is explained at https://github.com/golang/go/wiki/Switch.

However the goto does not need to be the last statement, and it is not really a fallthrough.

I think the text should be corrected.

Thanks  Manlio

[Ian Lance Taylor]

What correction do you propose?

To clarify, a labelled fallthrough statement is something like

label:
fallthrough

That is, a fallthrough statement may be the target of a goto
statement. You are of course correct that "fallthrough label" is not
supported.

Ian

[Manlio Perillo]

Il giorno martedì 24 novembre 2015 16:10:53 UTC+1, Ian Lance Taylor ha scritto:

On Tue, Nov 24, 2015 at 7:03 AM, Manlio Perillo
<manlio....@gmail.com> wrote:
> The go spec, at https://golang.org/ref/spec#ExprSwitchStmt, says:
> `In a case or default clause, the last non-empty statement may be a
> (possibly labeled) "fallthrough" statement to indicate that control should
> flow from the end of this clause to the first statement of the next clause`.
>

> [...] 

What correction do you propose?

Add an example.

 

To clarify, a labelled fallthrough statement is something like

    label:
        fallthrough

That is, a fallthrough statement may be the target of a goto
statement.

I really failed to figure this, thanks!

Of course now with an actual example, the text "labeled fallthrough" make sense.

Thanks  Manlio 

[Manlio Perillo]

Il giorno martedì 24 novembre 2015 16:20:00 UTC+1, Manlio Perillo ha scritto:

Il giorno martedì 24 novembre 2015 16:10:53 UTC+1, Ian Lance Taylor ha scritto:

> [...] 

clarify, a labelled fallthrough statement is something like

    label:
        fallthrough

That is, a fallthrough statement may be the target of a goto
statement.

I really failed to figure this, thanks!

Of course now with an actual example, the text "labeled fallthrough" make sense.

To clarify, I misinterpreted the "(optionally labeled) fallthrough" text with "fallthrough with a label" (the same as with goto).

The reason is that the text "(possibly labeled)" is superfluous (any statement can be labeled), so the fact that a language specification has such an additional text make me assume that there is a special case for the fallthrough statement.

Thanks  Manlio

[Matt Harden]

I agree; the best course may be to just remove the "(possibly labeled)" text.

--
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/d/optout.

[Konstantin Khomoutov]

On Tue, 24 Nov 2015 07:33:01 -0800 (PST)
Manlio Perillo <manlio....@gmail.com> wrote:

> > clarify, a labelled fallthrough statement is something like
> >>
> >> label:
> >> fallthrough
> >>
> >> That is, a fallthrough statement may be the target of a goto
> >> statement.
> >>
> > I really failed to figure this, thanks!
> > Of course now with an actual example, the text "labeled
> > fallthrough" make sense.
> >
>
> To clarify, I misinterpreted the "(optionally labeled) fallthrough"
> text with "fallthrough with a label" (the same as with goto).
> The reason is that the text "(possibly labeled

> <http://localhost:8000/ref/spec#Labeled_statements>)" is superfluous

> (any statement can be labeled), so the fact that a language
> specification has such an additional text make me assume that there
> is a special case for the fallthrough statement.

(Disclaimer: I'm not a native English-speaker either.)

In English, a phrase "(possibly labeled) something" stands for
"something, which might have a label attached to it (or may be not)".

That "-ed" ending of the verb "to label" flexes it to signalize the
possessive case; so a "labeled something" means that that
"something" has been labeled -- by attaching a label to it somehow,
and not the other way round ;-)

Hence I'm afraid your issue with the spec might be a purely linguistic
one.