Re: [PATCH, committed] PR c++/29089

[Roger Sayle <roger at eyesopen dot com>]

Hi Dirk,

On Sun, 29 Oct 2006, Dirk Mueller wrote:
> another simple one from the huge diagnostic buglist of the c++ frontend.
> compiled and manually tested, committed as obvious.

I hate to be the bearer of bad news, but neither this nor your previous
obvious patch for PR29033 adhere to required the GNU/GCC code coding style
guidelines.


> -	  pedwarn ("ISO C++ forbids %sing an enum",
> -		   (code == PREINCREMENT_EXPR || code == POSTINCREMENT_EXPR)
> -		   ? "increment" : "decrement");
> +	  pedwarn ((code == PREINCREMENT_EXPR || code == POSTINCREMENT_EXPR) ?
> +                   "ISO C++ forbids incrementing an enum" :
> +                   "ISO C++ forbids decrementing an enum");


The GNU coding rules require that when wrapping lines, operators appear
on the second line, indented/aligned under the start of the expression.

Hence,

	a
	? b
	: c

is acceptable, as is

	a
	? b : c

and at a stretch even

	a ? b
	  : c

and the (in my opinion slightly uglier)

	a ? b
	: c



But your usage of

	a ?
	b :
	c

and previously of

	a ? b :
        c

are incorrect.

And placing a "?" on a line by itself is "right out" :-).


Sorry for any inconvenience.  I hope this helps.


Roger
--