On Wed, May 18, 2005 at 12:13:23AM +0200, Gerrit P. Haase wrote:
> Hi,
>
>
> Yitzchak Scott-Thoennes wrote:
>
> >On Fri, May 13, 2005 at 02:21:28PM +0100, Orton, Yves wrote:
> >
> >>>>--- perl/cygwin/perlld.in 2003-11-24 14:44:26.000000000 -0800
> >>>>+++ perlpatch/cygwin/perlld.in 2005-05-11
> >>>
> >>>23:28:20.203129600 -0700
> >>>
> >>>Thanks, applied as #24461
> >>>
> >>>
> >>>>@@ -79,7 +79,7 @@
> >>>>
> >>>
> >>>#-------------------------------------------------------------
> >>>--------------
> >>>
> >>>>sub shellexec{
> >>>> my $command =shift;
> >>>>- print $command;
> >>>>+ print STDERR $command;
> >>>> print DEBUGFILE $command if $DEBUG;
> >>>> system($command) == 0
> >>>> or die "perlld: *** system() failed to execute\n$command\n";
> >>>>
> >>>
> >>Shouldnt there be a "# " on the front there?
> >
> >
> >No, STDERR is ignored. And I'm not 100% certain that $command is always
> >one line, anyway. If it is, printing "# $command\n" to STDOUT would be
> >ok. Gerrit?
>
> I don't really understand the problem. Is perl reading STDOUT and you
> want to comment it so it is ignored like STDERR?
Yes. t/TEST only is bothered by non-#'d extra output to STDOUT.
> Anyway, I'm not sure
> if it is useful at all that perlld prints to STDOUT, the compiler output
> arrives at STDOUT anyway and if there is s.th. wrong with $command we
> have a die at the end. So it seems to be ok to write to STDERR or even
> comment the whole line (seems to be better).
Thread Previous
|
Thread Next