This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [patch, libgfortran] Fix PR24459 Namelist problem
- From: Jerry DeLisle <jvdelisle at verizon dot net>
- To: Paul Thomas <paulthomas2 at wanadoo dot fr>, Fortran List <fortran at gcc dot gnu dot org>, gcc-patches <gcc-patches at gcc dot gnu dot org>
- Date: Mon, 08 May 2006 19:46:38 -0700
- Subject: Re: [patch, libgfortran] Fix PR24459 Namelist problem
- References: <A7E4CB800B73A2409E113012513E03940235E3@EX1V.rl.gov> <445991DF.4090303@verizon.net> <445AD025.3060406@verizon.net> <445EFBB3.9050208@wanadoo.fr>
Paul Thomas wrote:
Jerry,
:REVIEWMAIL:
I am sorry to take so long to get back to you on this.
The new versions of the patch and the testcase are OK, except that I
think something should go to stderr on detection of the extension! It
might be a tad on the mysterious side, otherwise.
Paul
I need some clarification here.
The extension only occurs (gets used) if there is data available. So to issue a
warning, we would have to set a flag in nml_parse_qualifier and then if extra
data is found by nml_read_obj, issue a warning. We would then clear the flag
after the warning so that we only issue it once.
Take the case of:
names(1,1) = 'onedata'
names2 = ...
In this case, even though the extension is enabled, it does not actually occur
because there is no extra data. In this case we do not want to issue a warning.
However, in this case:
names(1,1) = 'onedata' 'twodata'
names2 = ...
The only way we know that we have done something non-standard is to detect
actually reading in the second data item successfully.
I believe I can do this, but want to make sure this is the intent of your comment.
Regards,
Jerry