On Thu, 23 May 2019 17:00:02 GMT, nigel_horne@hotmail.com wrote:
Show quoted text
The bin/cstocs.PL turns into blib/script/cstocs and line 152 of blib/script/cstocs is actually line 167 in bin/cstocs.PL: while (<>) { ...
In that line 152 of bin/cstocs.PL we reexec the same perl interpreter with -i option to run the in-place replacement. And then in line 167 of bin/cstocs.PL the <> mechanism tries to read from @ARGV, and the test gives it t/out.tst to run.
If you could turn that
system "$^X @libs blib/script/cstocs -i.bak utf8 ascii $TSTFILE";
line in t/cstocs.t to say
system "ls -l $TSTFILE";
or
system "$^X -ne 'print' $TSTFILE";
and rerun make test -- that would allow us to determine where things call apart.
Thanks,
Jan
Show quoted text>
> -Nigel
>
> On 5/22/19, 3:36 PM, "Slaven_Rezic via RT" <bug-Cstools@rt.cpan.org>
> wrote:
>
> <URL:
https://rt.cpan.org/Ticket/Display.html?id=117730 >
>
> On 2019-05-22 05:15:11, JANPAZ wrote:
> > On Tue, 9 Apr 2019 20:21:23 GMT, SREZIC wrote:
> > >
> > > It seems that the optional module MIME::Words has to be installed
> > > to
> > > reproduce the failures.
> >
> > I've fixed this in Cstools 3.44.
> >
> > The other big group of issues (on Windows, on FreeBSD, even on Linux)
> > is the failure
> >
> > Can't open t/out.tst: No such file or directory at blib/script/cstocs
> > line 152.
> >
> > The code before that test 27 is
> >
> > open OUT, "> $TSTFILE";
> > print OUT $il2_to_utf8->conv('malièký je¾eèek'), "\n";
> > close OUT;
> >
> > print "Running: $^X @libs blib/script/cstocs -i.bak utf8 ascii
> > $TSTFILE\n";
> > system "$^X @libs blib/script/cstocs -i.bak utf8 ascii $TSTFILE";
> >
> > so it's not clear to me why that $TSTFILE cannot be found when we
> > just
> > finished writing to it and closed it the line before.
>
>
http://matrix.cpantesters.org/?dist=Cstools%203.44;reports=1 -> all
> failures on linux & freebsd happen on Nigel Horne's systems. Put him
> on Cc, maybe he has an idea...
>
>
>
RT for rt.cpan.org