On Mon, 26 Apr 2004, Steve Hay wrote:
> Randy Kobes wrote:
>
> >On Sat, 24 Apr 2004, Stas Bekman wrote:
> >
> >>Randy Kobes wrote:
> >>
> >>>Hi,
> >>> One Win32, t/filter/out_str)subreq_default fails
> >>>on Win32 because of different line endings. This:
> >>>=====================================================
> >>>Index: out_str_subreq_default.t
> >>>===================================================================
> >>>RCS file: /home/cvs/modperl-2.0/t/filter/out_str_subreq_default.t,v
> >>>retrieving revision 1.1
> >>>diff -u -r1.1 out_str_subreq_default.t
> >>>--- out_str_subreq_default.t 4 Mar 2004 15:16:10 -0000 1.1
> >>>+++ out_str_subreq_default.t 24 Apr 2004 05:47:53 -0000
> >>>@@ -16,6 +16,7 @@
> >>>
> >>> my $expected = join '', $content1, $subrequest, $content2, $filter;
> >>> my $received = GET_BODY $location;
> >>>+$received =~ s{\r}{}g;
> >>>
> >>> ok t_cmp($expected, $received,
> >>> "testing filter-originated lookup_uri() call to core served URI");
> >>>=================================================================
> >>>fixes it, but perhaps there's a cleaner way?
> >>>
> >>>
> >>+1, but please add '... if WIN32;' so it's clear why we do that.
> >>
> >>What's strange is that this test exists for quite a
> >>while, and only now it stopped working? Or does it have
> >>to do with how the server side files were checked out?
> >>
> >>
> >
> >I'm not sure what's happened - you're right, these have been
> >there for a while (since March 4). For me, the most
> >significant change since then is to upgrade to
> >Apache/2.0.49. Steve, have you noticed a change in this
> >test, and if so, does the above diff make any difference?
> >
> I just checked out the latest CVS mp2 and all tests run successfully,
> even without your patch.
>
> This is using Apache 2.0.49 and Perl 5.8.3 on WinXP.
>
> - Steve
Sigh :) Geoff reminded us that he also found a failure
in this test due to line endings - I guess it wouldn't
hurt to apply the patch (specifying it's for Win32), and
come back later to why it seems necessary for some and
not necessary for others.
--
best regards,
randy
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
