http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58001
--- Comment #6 from Steve Kargl <sgk at troutmask dot apl.washington.edu> ---
On Tue, Jul 30, 2013 at 12:59:00PM +0000, dominiq at lps dot ens.fr wrote:
>
> If compiled with -Wno-tabs, it gives
>
> pr58001.f90:2.1:
>
> print 1894
> 1
> Warning: Nonconforming tab character at (1)
> pr58001.f90:3.14:
>
> 1894 format( '123')
> 1
> Warning: Nonconforming tab character at (1)
> pr58001.f90:3.14:
>
> 1894 format( '123')
> 1
> Warning: Extension: Tab character in format at (1)
>
> Is this the expected behavior?
With -Wno-tabs, gfortran should report a warning for the
occurence of every tab used in the context of a character
from the Fortran character set. Taking your code, which I've
deleted here, and replace all whitespace by tabs, I get the
expected number of warnings. Historically, I wanted to use
-Wtabs as you expected, but there was too much wailing on
the gfortran lists, so it was changed to -Wno-tabs with the
meaning that "no tabs are allowed" in the context of the
Fortran character set. You'll notice that tabs within
a literal character string are not flagged. The history
of -Wtabs verse -Wno-tabs can be found in the mailinglist
archive.
Now, for the problem at hand, it seems that there is a
bug in io.c. Here the code in question:
/* Eat up the spaces and return a character. */
static char
next_char_not_space (bool *error)
{
char c;
do
{
error_element = c = next_char (NONSTRING);
if (c == '\t')
{
if (gfc_option.allow_std & GFC_STD_GNU)
gfc_warning ("Extension: Tab character in format at %C");
else
{
gfc_error ("Extension: Tab character in format at %C");
*error = true;
return c;
}
}
}
while (gfc_is_whitespace (c));
return c;
}
Notice that there is no check for gfc_option.warn_tabs.
