Bug #85001 for Time-Format: Test failures due to Date::Manip alteration

Thu May 02 11:01:18 2013 sendu [...] sendu.me.uk - Ticket created

Subject:	Test failures due to Date::Manip alteration
Date:	Thu, 2 May 2013 16:01:02 +0100
To:	bug-Time-Format [...] rt.cpan.org
From:	Sendu Bala <sendu [...] sendu.me.uk>

Running perl 5.12.2 on an Ubuntu precise system, Date::Manip 6.39 installed, trying to install Time::Format 1.12 I get the failures you see at the end of this email. If I alter the 3 failing tests (t/doc.t, t/funcs.t and t/manip.t) to remove this line: %Date::Manip:: = (); Then all tests pass fine. I can't work out why that line is there in the test scripts. It isn't used in the Time::Format module itself, so presumably the test failures don't indicate a real problem? Perhaps you could remove those lines so people can install Time::Format easily in the future. $ perl Build test t/0-signature.t .. ok t/1-load.t ....... ok t/DateManip.t .... ok t/DateTime.t ..... ok t/die.t .......... ok t/doc.t .......... 1/26 Can't locate package Date::Manip::Obj for @Date::Manip::TZ::ISA at /software/vertres/lib/perl/lib/perl5/Date/Manip/TZ.pm line 1367, <DATA> line 1. Can't locate package Date::Manip::TZ_Base for @Date::Manip::TZ::ISA at /software/vertres/lib/perl/lib/perl5/Date/Manip/TZ.pm line 1367, <DATA> line 1. Can't locate object method "_now" via package "Date::Manip::TZ" at /software/vertres/lib/perl/lib/perl5/Date/Manip/TZ.pm line 1367, <DATA> line 1. # Looks like you planned 26 tests but ran 23. # Looks like your test exited with 255 just after 23. t/doc.t .......... Dubious, test returned 255 (wstat 65280, 0xff00) Failed 3/26 subtests t/epoch.t ........ ok t/export1.t ...... ok t/export2.t ...... ok t/export3.t ...... ok t/funcs.t ........ 1/18 Can't locate package Date::Manip::Obj for @Date::Manip::TZ::ISA at /software/vertres/lib/perl/lib/perl5/Date/Manip/Date.pm line 1861, <DATA> line 1. Can't locate package Date::Manip::TZ_Base for @Date::Manip::TZ::ISA at /software/vertres/lib/perl/lib/perl5/Date/Manip/Date.pm line 1861, <DATA> line 1. Can't locate object method "_fix_year" via package "Date::Manip::TZ" at /software/vertres/lib/perl/lib/perl5/Date/Manip/Date.pm line 1861, <DATA> line 1. # Looks like you planned 18 tests but ran 12. # Looks like your test exited with 255 just after 12. t/funcs.t ........ Dubious, test returned 255 (wstat 65280, 0xff00) Failed 6/18 subtests t/locale.t ....... ok t/manip.t ........ 1/6 Can't locate package Date::Manip::Obj for @Date::Manip::TZ::ISA at /software/vertres/lib/perl/lib/perl5/Date/Manip/Date.pm line 1861. Can't locate package Date::Manip::TZ_Base for @Date::Manip::TZ::ISA at /software/vertres/lib/perl/lib/perl5/Date/Manip/Date.pm line 1861. Can't locate object method "_fix_year" via package "Date::Manip::TZ" at /software/vertres/lib/perl/lib/perl5/Date/Manip/Date.pm line 1861. # Looks like you planned 6 tests but ran 1. # Looks like your test exited with 255 just after 1. t/manip.t ........ Dubious, test returned 255 (wstat 65280, 0xff00) Failed 5/6 subtests t/msec.t ......... ok t/past.t ......... ok t/quot.t ......... ok t/strftime.t ..... ok t/string.t ....... ok t/time.t ......... ok t/xs_DateTime.t .. ok t/xs_doc.t ....... ok t/xs_funcs.t ..... ok t/xs_locale.t .... ok t/xs_quot.t ...... ok t/xs_time.t ...... ok Test Summary Report ------------------- t/doc.t (Wstat: 65280 Tests: 23 Failed: 0) Non-zero exit status: 255 Parse errors: Bad plan. You planned 26 tests but ran 23. t/funcs.t (Wstat: 65280 Tests: 12 Failed: 0) Non-zero exit status: 255 Parse errors: Bad plan. You planned 18 tests but ran 12. t/manip.t (Wstat: 65280 Tests: 1 Failed: 0) Non-zero exit status: 255 Parse errors: Bad plan. You planned 6 tests but ran 1. Files=25, Tests=459, 6 wallclock secs ( 0.14 usr 0.04 sys + 1.07 cusr 0.30 csys = 1.55 CPU) Result: FAIL Failed 3/25 test programs. 0/459 subtests failed.

Tue May 24 10:17:09 2016 EDAVIS [...] cpan.org - Correspondence added

I can confirm this with perl 5.22 and Date::Manip 6.53. Here's the suggested change as a patch: diff -ru Time-Format-1.12-orig/t/doc.t Time-Format-1.12/t/doc.t --- Time-Format-1.12-orig/t/doc.t 2012-09-27 17:32:39.000000000 +0100 +++ Time-Format-1.12/t/doc.t 2016-05-24 15:12:47.598894130 +0100 @@ -20,7 +20,6 @@ $dm_notz = eval('Date::Manip::Date_TimeZone (); 1')? 0 : 1; } delete $INC{'Date/Manip.pm'}; - %Date::Manip:: = (); } # Were all variables imported? (3) diff -ru Time-Format-1.12-orig/t/funcs.t Time-Format-1.12/t/funcs.t --- Time-Format-1.12-orig/t/funcs.t 2012-09-27 17:32:39.000000000 +0100 +++ Time-Format-1.12/t/funcs.t 2016-05-24 15:12:47.598894130 +0100 @@ -23,7 +23,6 @@ $manip_notz = eval ('Date::Manip::Date_TimeZone (); 1')? 0 : 1; } delete $INC{'Date/Manip.pm'}; - %Date::Manip:: = (); } # Get day/month names in current locale diff -ru Time-Format-1.12-orig/t/manip.t Time-Format-1.12/t/manip.t --- Time-Format-1.12-orig/t/manip.t 2012-09-27 17:32:39.000000000 +0100 +++ Time-Format-1.12/t/manip.t 2016-05-24 15:12:47.598894130 +0100 @@ -19,7 +19,6 @@ unless eval 'Date::Manip::Date_TimeZone(); 1'; } delete $INC{'Date/Manip.pm'}; - %Date::Manip:: = (); } my $t = 'first thursday in june 2003';

Tue May 24 10:17:09 2016 The RT System itself - Status changed from 'new' to 'open'

Tue May 24 14:55:48 2016 eric.roode.cpan [...] gmail.com - Correspondence added

On Thu May 02 11:01:18 2013, sendu@sendu.me.uk wrote: [...] Show quoted text

> > If I alter the 3 failing tests (t/doc.t, t/funcs.t and t/manip.t) to > remove this line: > %Date::Manip:: = (); > Then all tests pass fine. > > I can't work out why that line is there in the test scripts. It isn't > used in the Time::Format module itself, so presumably the test > failures don't indicate a real problem? > > Perhaps you could remove those lines so people can install > Time::Format easily in the future.

You are correct that removing those lines appears to resolve the problem, but they are not actually the problem. Those lines are a sort of hacky way I unloaded the Date::Manip module. Part of the test is to ensure that Time::Format correctly loads Date::Manip if it needs; that is, if the user references %manip. I do not know what changed in Date::Manip that resulted in my crude module-unloading failing. I am trying to come up with a workaround.

Wed Sep 21 12:48:31 2016 PLICEASE [...] cpan.org - Cc PLICEASE added

Mon Feb 26 20:57:48 2018 jkeenan [...] cpan.org - Correspondence added

On Tue May 24 14:55:48 2016, ROODE wrote: Show quoted text

> On Thu May 02 11:01:18 2013, sendu@sendu.me.uk wrote: > [...]

> > > > If I alter the 3 failing tests (t/doc.t, t/funcs.t and t/manip.t) to > > remove this line: > > %Date::Manip:: = (); > > Then all tests pass fine. > > > > I can't work out why that line is there in the test scripts. It isn't > > used in the Time::Format module itself, so presumably the test > > failures don't indicate a real problem? > > > > Perhaps you could remove those lines so people can install > > Time::Format easily in the future.

> > You are correct that removing those lines appears to resolve the > problem, but they are not actually the problem. > > Those lines are a sort of hacky way I unloaded the Date::Manip module. > Part of the test is to ensure that Time::Format correctly loads > Date::Manip if it needs; that is, if the user references %manip. > > I do not know what changed in Date::Manip that resulted in my crude > module-unloading failing. I am trying to come up with a workaround.

Has there been any progress on this since your last post? I ask because Time-Format is fairly "high up on the CPAN river", so other CPAN distributions which depend upon it no longer can be tested. Today I tested Time-Format against perl-5.27.9 and got the same failures as reported two years ago. Thank you very much. Jim Keenan

Fri Mar 16 04:04:09 2018 KENTNL [...] cpan.org - Correspondence added

Looking at what you're trying to achieve, it might be better to execute this block in a seperate perl process somehow, and pass the result back:

---
BEGIN
{
    $dm_notok = eval('use Date::Manip (); 1')? 0 : 1;
    unless ($dm_notok)
    {
        # If Date::Manip can't determine the time zone, it'll bomb out of the tests.
        $dm_notz = eval('Date::Manip::Date_TimeZone (); 1')? 0 : 1;
    }
}

---

That'd be more future-proof against both perl design changes and Date::Manip internal detail changes, I'd imagine.

--
- CPAN kentnl@cpan.org
- Gentoo Perl Maintainer kentnl@gentoo.org ( perl@gentoo.org )

Thu Jul 11 19:50:32 2019 jkeenan [...] cpan.org - Correspondence added

On Mon Feb 26 20:57:48 2018, JKEENAN wrote: Show quoted text

> On Tue May 24 14:55:48 2016, ROODE wrote:

> > On Thu May 02 11:01:18 2013, sendu@sendu.me.uk wrote: > > [...]

> > > > > > If I alter the 3 failing tests (t/doc.t, t/funcs.t and t/manip.t) > > > to > > > remove this line: > > > %Date::Manip:: = (); > > > Then all tests pass fine. > > > > > > I can't work out why that line is there in the test scripts. It > > > isn't > > > used in the Time::Format module itself, so presumably the test > > > failures don't indicate a real problem? > > > > > > Perhaps you could remove those lines so people can install > > > Time::Format easily in the future.

> > > > You are correct that removing those lines appears to resolve the > > problem, but they are not actually the problem. > > > > Those lines are a sort of hacky way I unloaded the Date::Manip > > module. > > Part of the test is to ensure that Time::Format correctly loads > > Date::Manip if it needs; that is, if the user references %manip. > > > > I do not know what changed in Date::Manip that resulted in my crude > > module-unloading failing. I am trying to come up with a workaround.

> > Has there been any progress on this since your last post? > > I ask because Time-Format is fairly "high up on the CPAN river", so > other CPAN distributions which depend upon it no longer can be tested. > Today I tested Time-Format against perl-5.27.9 and got the same > failures as reported two years ago. > > Thank you very much. > Jim Keenan

Still a problem when tested against perl-5.31.0 on FreeBSD-12 today. Can you investigate? Thank you very much. Jim Keenan

Fri Jul 12 09:06:10 2019 roode [...] cpan.org - Correspondence added

Subject:	Re: [rt.cpan.org #85001] Test failures due to Date::Manip alteration
Date:	Fri, 12 Jul 2019 09:06:13 -0400
To:	bug-Time-Format [...] rt.cpan.org
From:	Eric Roode <eric.roode.cpan [...] gmail.com>

Hello James, Yes, I will see if I can fix this. -- Eric On Thu, Jul 11, 2019 at 7:50 PM James E Keenan via RT < bug-Time-Format@rt.cpan.org> wrote: Show quoted text

> Queue: Time-Format > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=85001 > > > On Mon Feb 26 20:57:48 2018, JKEENAN wrote:

> > On Tue May 24 14:55:48 2016, ROODE wrote:

> > > On Thu May 02 11:01:18 2013, sendu@sendu.me.uk wrote: > > > [...]

> > > > > > > > If I alter the 3 failing tests (t/doc.t, t/funcs.t and t/manip.t) > > > > to > > > > remove this line: > > > > %Date::Manip:: = (); > > > > Then all tests pass fine. > > > > > > > > I can't work out why that line is there in the test scripts. It > > > > isn't > > > > used in the Time::Format module itself, so presumably the test > > > > failures don't indicate a real problem? > > > > > > > > Perhaps you could remove those lines so people can install > > > > Time::Format easily in the future.

> > > > > > You are correct that removing those lines appears to resolve the > > > problem, but they are not actually the problem. > > > > > > Those lines are a sort of hacky way I unloaded the Date::Manip > > > module. > > > Part of the test is to ensure that Time::Format correctly loads > > > Date::Manip if it needs; that is, if the user references %manip. > > > > > > I do not know what changed in Date::Manip that resulted in my crude > > > module-unloading failing. I am trying to come up with a workaround.

> > > > Has there been any progress on this since your last post? > > > > I ask because Time-Format is fairly "high up on the CPAN river", so > > other CPAN distributions which depend upon it no longer can be tested. > > Today I tested Time-Format against perl-5.27.9 and got the same > > failures as reported two years ago. > > > > Thank you very much. > > Jim Keenan

> > Still a problem when tested against perl-5.31.0 on FreeBSD-12 today. > > Can you investigate? > > Thank you very much. > Jim Keenan > >

Fri Jul 19 08:41:35 2019 eric.roode.cpan [...] gmail.com - Correspondence added

RT-Send-CC:

jkeenan [...] cpan.org

I reworked the test suite to test for optional modules differently. No code changes were needed in the library itself.

Fri Jul 19 08:41:36 2019 eric.roode.cpan [...] gmail.com - Status changed from 'open' to 'resolved'

Fri Jul 19 08:41:37 2019 eric.roode.cpan [...] gmail.com - Fixed in 1.13 added

Fri Jul 19 08:52:40 2019 jkeenan [...] cpan.org - Correspondence added

On Fri Jul 19 08:41:35 2019, ROODE wrote: Show quoted text

> I reworked the test suite to test for optional modules differently. > No code changes were needed in the library itself.

LGTM. When Date::Manip is already installed, Time::Format passes its tests and installs. ##### [blead] $ ./bin/cpanm Date::Manip --> Working on Date::Manip Fetching http://www.cpan.org/authors/id/S/SB/SBECK/Date-Manip-6.77.tar.gz ... OK Configuring Date-Manip-6.77 ... OK ==> Found dependencies: Test::Inter --> Working on Test::Inter Fetching http://www.cpan.org/authors/id/S/SB/SBECK/Test-Inter-1.09.tar.gz ... OK Configuring Test-Inter-1.09 ... OK Building and testing Test-Inter-1.09 ... OK Successfully installed Test-Inter-1.09 Building and testing Date-Manip-6.77 ... OK Successfully installed Date-Manip-6.77 2 distributions installed [blead] $ ./bin/perl -MTime::Format -E 'say q|hello world|' Can't locate Time/Format.pm in @INC (you may need to install the Time::Format module) (@INC contains: /home/jkeenan/testing/blead/lib/perl5/site_perl/5.31.2/amd64-freebsd-thread-multi /home/jkeenan/testing/blead/lib/perl5/site_perl/5.31.2 /home/jkeenan/testing/blead/lib/perl5/5.31.2/amd64-freebsd-thread-multi /home/jkeenan/testing/blead/lib/perl5/5.31.2). BEGIN failed--compilation aborted. [blead] $ ./bin/cpanm Time::Format --> Working on Time::Format Fetching http://www.cpan.org/authors/id/R/RO/ROODE/Time-Format-1.13.tar.gz ... OK Configuring Time-Format-1.13 ... OK Building and testing Time-Format-1.13 ... OK Successfully installed Time-Format-1.13 1 distribution installed ##### Thank you very much. Jim Keenan