Bug #128427 for Data-Sah-Coerce: Data::Sah::Coerce fails on t/js-date.t

Tue Feb 05 18:38:21 2019 james2vegas [...] aim.com - Ticket created

Subject:

Data::Sah::Coerce fails on t/js-date.t

$ make test
PERL_DL_NONLAZY=1 "/usr/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/00-compile.t ........... ok
t/01-basic.t ............. ok
t/author-critic.t ........ skipped: these tests are for testing by the author ]
t/author-pod-coverage.t .. skipped: these tests are for testing by the author
t/author-pod-syntax.t .... skipped: these tests are for testing by the author
t/js-bool.t .............. ok
t/js-date.t ..............
        #   Failed test at t/js-date.t line 26.
        #          got: '2016-01-01T17:34:56.000Z'
        #     expected: '2016-01-01T12:34:56.000Z'
        # Looks like you failed 1 test of 2.

    #   Failed test 'from string'
    #   at t/js-date.t line 27.
    # Looks like you failed 1 test of 3.
t/js-date.t .............. 1/?
Show quoted text

# Failed test 'coerce_to=Date'
# at t/js-date.t line 28.
# Looks like you failed 1 test of 1.

t/js-date.t .............. Dubious, test returned 1 (wstat 256, 0x100)

Mon May 13 15:44:00 2019 SREZIC [...] cpan.org - Correspondence added

On 2019-02-05 18:38:21, JWRIGHT wrote: Show quoted text

> $ make test > PERL_DL_NONLAZY=1 "/usr/bin/perl" "-MExtUtils::Command::MM" "- > MTest::Harness" > "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', > 'blib/arch')" > t/*.t > t/00-compile.t ........... ok > t/01-basic.t ............. ok > t/author-critic.t ........ skipped: these tests are for testing by the > author ] > t/author-pod-coverage.t .. skipped: these tests are for testing by the > author > t/author-pod-syntax.t .... skipped: these tests are for testing by the > author > t/js-bool.t .............. ok > t/js-date.t .............. > # Failed test at t/js-date.t line 26. > # got: '2016-01-01T17:34:56.000Z' > # expected: '2016-01-01T12:34:56.000Z' > # Looks like you failed 1 test of 2. > > # Failed test 'from string' > # at t/js-date.t line 27. > # Looks like you failed 1 test of 3. > t/js-date.t .............. 1/? > # Failed test 'coerce_to=Date' > # at t/js-date.t line 28. > # Looks like you failed 1 test of 1. > > > t/js-date.t .............. Dubious, test returned 1 (wstat 256, 0x100)

Looks like a timezone issue. There's no problem if TZ=UTC is used.

Mon May 13 15:44:00 2019 The RT System itself - Status changed from 'new' to 'open'

Mon May 13 16:07:47 2019 SREZIC [...] cpan.org - Correspondence added

On 2019-05-13 15:44:00, SREZIC wrote: Show quoted text

> On 2019-02-05 18:38:21, JWRIGHT wrote:

> > $ make test > > PERL_DL_NONLAZY=1 "/usr/bin/perl" "-MExtUtils::Command::MM" "- > > MTest::Harness" > > "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', > > 'blib/arch')" > > t/*.t > > t/00-compile.t ........... ok > > t/01-basic.t ............. ok > > t/author-critic.t ........ skipped: these tests are for testing by the > > author ] > > t/author-pod-coverage.t .. skipped: these tests are for testing by the > > author > > t/author-pod-syntax.t .... skipped: these tests are for testing by the > > author > > t/js-bool.t .............. ok > > t/js-date.t .............. > > # Failed test at t/js-date.t line 26. > > # got: '2016-01-01T17:34:56.000Z' > > # expected: '2016-01-01T12:34:56.000Z' > > # Looks like you failed 1 test of 2. > > > > # Failed test 'from string' > > # at t/js-date.t line 27. > > # Looks like you failed 1 test of 3. > > t/js-date.t .............. 1/? > > # Failed test 'coerce_to=Date' > > # at t/js-date.t line 28. > > # Looks like you failed 1 test of 1. > > > > > > t/js-date.t .............. Dubious, test returned 1 (wstat 256, 0x100)

> > Looks like a timezone issue. There's no problem if TZ=UTC is used.

I can reproduce it on my ubuntu 18.04 smoker if TZ=America/Anchorage is set. However, I cannot reproduce it on other linux systems (debian/jessie, .../stretch, centos6) using the same TZ value.

Mon Dec 30 05:12:26 2019 SREZIC [...] cpan.org - Correspondence added

On 2019-05-13 16:07:47, SREZIC wrote: Show quoted text

> On 2019-05-13 15:44:00, SREZIC wrote:

> > On 2019-02-05 18:38:21, JWRIGHT wrote:

> > > $ make test > > > PERL_DL_NONLAZY=1 "/usr/bin/perl" "-MExtUtils::Command::MM" "- > > > MTest::Harness" > > > "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', > > > 'blib/arch')" > > > t/*.t > > > t/00-compile.t ........... ok > > > t/01-basic.t ............. ok > > > t/author-critic.t ........ skipped: these tests are for testing by > > > the > > > author ] > > > t/author-pod-coverage.t .. skipped: these tests are for testing by > > > the > > > author > > > t/author-pod-syntax.t .... skipped: these tests are for testing by > > > the > > > author > > > t/js-bool.t .............. ok > > > t/js-date.t .............. > > > # Failed test at t/js-date.t line 26. > > > # got: '2016-01-01T17:34:56.000Z' > > > # expected: '2016-01-01T12:34:56.000Z' > > > # Looks like you failed 1 test of 2. > > > > > > # Failed test 'from string' > > > # at t/js-date.t line 27. > > > # Looks like you failed 1 test of 3. > > > t/js-date.t .............. 1/? > > > # Failed test 'coerce_to=Date' > > > # at t/js-date.t line 28. > > > # Looks like you failed 1 test of 1. > > > > > > > > > t/js-date.t .............. Dubious, test returned 1 (wstat 256, > > > 0x100)

> > > > Looks like a timezone issue. There's no problem if TZ=UTC is used.

> > I can reproduce it on my ubuntu 18.04 smoker if TZ=America/Anchorage > is set. > However, I cannot reproduce it on other linux systems (debian/jessie, > .../stretch, centos6) using the same TZ value.

I don't see this kind of failure anymore with 0.040. The previously failing test file js_date.t does not exist anymore --- maybe it's now split up into js-To_date.t and perl-To_date.t, but both don't fail anymore in this way. So maybe this issue may be set to resolved?

Mon Dec 30 10:37:13 2019 PERLANCAR [...] cpan.org - Correspondence added

On Mon, 30 Dec 2019 10:12:26 GMT, SREZIC wrote: Show quoted text

> I don't see this kind of failure anymore with 0.040. The previously > failing test file js_date.t does not exist anymore --- maybe it's now > split up into js-To_date.t and perl-To_date.t, but both don't fail > anymore in this way. > > So maybe this issue may be set to resolved?

Closing as resolved for now. Thanks.

Mon Dec 30 10:37:14 2019 PERLANCAR [...] cpan.org - Status changed from 'open' to 'resolved'