This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [v3] libstdc++/6518


Andreas Schwab wrote:

>Paolo Carlini <pcarlini@unitus.it> writes:
>
>|> Andreas Schwab wrote:
>|> 
>|> >I'm talking about an string of length 0, not a null pointer.  I cannot
>|> >find anything in the standard that requires badbit being set when an empty
>|> >string is passed to operator<<.
>|> >
>|> IMHO, this is a very good point:
>
>Why is it not a good point?  This causes correct programs to fail.
>
Andreas, I said it is! I agree completely with you!

But now: what to do about that? I mean, please excuse my naivete: how to 
tell a null terminated zero length string from a null pointer?!? I'm not 
able to do that. Perhaps you are implying that, since we are not able to 
do that, we must give up trying to diagnose the situation and therefore 
simply never set the badbit?

Paolo.



Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]