Hello.

I found 2 bugs with nested comments. Not sure about second but first indeed.

So the first one:

> 
> module core;
> 
> /+
> char[] string = "/+";
> +/
> 
> int main(char[][] args)
> {
> 	return 0;
> }

D:\proj\dtest>dmd core.d
core.d(12): unterminated /+ +/ comment

And second is:

> 
> module core;
> 
> /+
> int[] x = 5; // comment with /+ in it
> +/
> 
> int main(char[][] args)
> {
> 	return 0;
> }

Compiloation gives exactly same result.

-- 
Victor (aka nail) Nakoryakov
nail-mail<at>mail<dot>ru

Krasnoznamensk, Moscow, Russia

Victor Nakoryakov <nail-mail@mail.ru> wrote:

> I found 2 bugs with nested comments.

No bugs at all:

| This means that embedded strings and comments do not prevent | recognition of comment openings and closings

The first comment token appearing in normal source code determines the type of the comment and that of the the following source code that belongs to the type determined as above is considered to control the state of the comment.

-manfred

Manfred Nowak wrote:
> Victor Nakoryakov <nail-mail@mail.ru> wrote:
> 
> 
>>I found 2 bugs with nested comments.
> 
> 
> No bugs at all:
> 
> | This means that embedded strings and comments do not prevent
> | recognition of comment openings and closings
> 
> The first comment token appearing in normal source code determines the type of the comment and that of the the following source code that belongs to the type determined as above is considered to control the state of the comment.
> 
> -manfred


Aaargh! I'm stupid, stupid, stupid... to many beer yesterday :)
Of course there is no bug there, sorry.

-- 
Victor (aka nail) Nakoryakov
nail-mail<at>mail<dot>ru

Krasnoznamensk, Moscow, Russia

Victor Nakoryakov <nail-mail@mail.ru> wrote:

> Aaargh! I'm stupid, stupid, stupid...

No, you are not. I think the knowledge of the many failures software professionals make, do also make them for errors the most tolerant professional group.

-manfred