On Thu, Sep 09, 2004 at 12:28:27PM -0400, Jeff Urlwin wrote:
> Doh.  Yes.  Didn't think about File::Spec and devnull()...I learn something
> new every day. -- yet it seems that it's still not fast enough ;)

The patch would have been perfect[1] if it hadn't been wrapped by your mailer :)

Thanks.

Tim.

[1] "perfect" means I can apply it from my mail reader (mutt) by typing
        | cd ~/dbi/trunk && patch -p0

> Jeff
> 
> > 
> > Could you rework that using File::Spec?
> > 
> 
> Index: t/01basics.t
> ===================================================================
> --- t/01basics.t      (revision 452)
> +++ t/01basics.t      (working copy)
> @@ -3,6 +3,7 @@
>  use strict;
>  
>  use Test::More tests => 110;
> +use File::Spec;
>  
>  ##
> ----------------------------------------------------------------------------
>  ## 01basic.t - test of some basic DBI functions 
> @@ -147,10 +148,11 @@
>  ## testing dbi_debug
>  
>  cmp_ok($DBI::dbi_debug, '==',  0, "... DBI::dbi_debug's initial state is
> 0");
> +my $null = File::Spec::->devnull();
>  
>  SKIP: {
> -    skip "cannot find : /dev/null", 2 unless (-e "/dev/null");       
> -    DBI->trace(42,"/dev/null");
> +    skip "cannot find : $null", 2 unless ($^O eq "MSWin32" || -e $null);
> 
> +    DBI->trace(42,$null);
>      cmp_ok($DBI::dbi_debug, '==', 42, "... DBI::dbi_debug is 42");
>       DBI->trace(0, undef);
>      cmp_ok($DBI::dbi_debug, '==',  0, "... DBI::dbi_debug is 0");
> 
> 

Reply via email to