Re: libstdc++/9439: filebuf::sputbackc ignores beginning-of-file

[Paolo Carlini <pcarlini@unitus.it>]

The following reply was made to PR libstdc++/9439; it has been noted by GNATS.

From: Paolo Carlini <pcarlini@unitus.it>
To: =?ISO-8859-1?Q?P=E9tur_Run=F3lfsson?= <peturr02@ru.is>
Cc: paolo@gcc.gnu.org,  bkoz@redhat.com,  gcc-bugs@gcc.gnu.org, 
 nobody@gcc.gnu.org,  gcc-gnats@gcc.gnu.org
Subject: Re: libstdc++/9439: filebuf::sputbackc ignores beginning-of-file
Date: Mon, 03 Feb 2003 15:21:09 +0100

 P=E9tur Run=F3lfsson wrote:
 
 >Hi,
 > =20
 >
 >>    Hi. I don't think there is a bug here: I cannot find in the
 >>    standard a specific prescription for the behaviour you
 >>    expect (in particular in 27.5.2.4.4) Can you?
 >>   =20
 >>
 >
 >Yes. 27.8.1.4 [lib.filebuf.virtuals] p5 describes how pbackfail may
 >put back the character c. The only cases that can apply here are the
 >ones starting with
 >
 [snip]
 
 Ok. I agree with your reading.
 
 Therefore it seems that seekoff should be changed to return error in=20
 this case. Then pbackfail would pay attention to the negative returned=20
 value. The latter changes will also fix libstdc++/9425.
 
 >>    Indeed, sputbackc calls, as expected, pbackfail, which in=20
 >>turns calls seekoff (fstream.tcc, line 218) (pay attention
 >>    to the preceding comment which means that this specific=20
 >>situation was considered and _not_ supposed to lead to an
 >>    obvious failure).
 >>   =20
 >>
 >I read "at the beginning of the buffer" as simply meaning that
 >gptr() =3D=3D eback(), not as "at the beginning of the file".
 >
 Yes, you are right.
 
 Paolo.
 
 http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=3Dview%20audit-trail&database=3D=
 gcc&pr=3D9439