Re: sizeof(type)

[<tlaronde%kergis.com@localhost>]

On Fri, Nov 29, 2024 at 09:59:19AM +0000, David Holland wrote:
> On Mon, Nov 25, 2024 at 07:53:05PM -0500, Mouse wrote:
>  > There's also
>  > 
>  > 	sizeof (int) - 1
>  > 
>  > which is ambiguous even _with_ the parens; it means either
>  > 	(sizeof(int))-1
>  > or
>  > 	sizeof((int)-1)
>  > 
>  > In this case, I think the ambiguity is resolved by operator precedence
> 
> C99 6.5.4 says
>    cast-expression ::=
>         unary-expression
>       | ( type-name ) cast-expression
> 
> and 6.5.3 says
>    unary-expression ::=
>         postfix-expression
>       | ++ unary-expression
>       | -- unary-expression
>       | unary-operator cast-expression
>       | sizeof unary-expression
>       | sizeof ( type-name )
> 
> The reason these are stratified is so that sizeof ( type-name ) is
> unambiguous.
> 
> As written it also rules out ++(int)x, but that's a defect -- it is
> not semantically valid but there's no reason to gratuitously make it
> a syntax error. Especially since ++ +(int)x is accepted and that's
> semantically equally invalid.
> 

It is also specified in K&R (I didn't look at the standard), p. 200,
for A7. "Expressions":

"The precedence of expressions operators is the same as the order of
the major subsections of this [A7] section, highest precedence first.
Thus, for example, the expressions referred to as the operands of +
(A7.7) are those expressions defined in A7.1--A7.6. Within each
subsection, the operators have the same precedence. [...]"

So, when considering precedence, this too has to be taken into
account.

FWIW,
-- 
        Thierry Laronde <tlaronde +AT+ kergis +dot+ com>
                     http://www.kergis.com/
                    http://kertex.kergis.com/
Key fingerprint = 0FF7 E906 FBAF FE95 FD89  250D 52B1 AE95 6006 F40C