Skip Menu |

This queue is for tickets about the namespace-clean CPAN distribution.

Report information
The Basics
Id: 107343
Status: resolved
Priority: 0/
Queue: namespace-clean

People
Owner: Nobody in particular
Requestors: ether [...] cpan.org
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: 0.27



Subject: tests fail, darwin 5.8.1
A cpantesters report has been submitted... However, there was also this on stdout: ... Building and testing B-Hooks-EndOfScope-0.15 ... OK Successfully installed B-Hooks-EndOfScope-0.15 Building and testing namespace-clean-0.25 ... A scope-end callback raised an exception, which can not be propagated when B::Hooks::EndOfScope operates in pure-perl mode. Your program will CONTINUE EXECUTION AS IF NOTHING HAPPENED AFTER THIS WARNING. Below is the complete exception text, followed by a stack-trace of the callback execution: namespace::clean::deleted::Foo is not a module name at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 227. at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 227. namespace::clean::__ANON__[/Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm:252]("Foo", HASH(0x7fec8882ac10), "baz", "foo") called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 315 namespace::clean::__ANON__[/Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm:316]() called at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP.pm line 33 eval {...} called at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP.pm line 33 B::Hooks::EndOfScope::PP::__invoke_callback(CODE(0x7fec8885d400)) called at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP/HintHash.pm line 39 B::Hooks::EndOfScope::PP::_SG_STACK::DESTROY(B::Hooks::EndOfScope::PP::_SG_STACK=ARRAY(0x7fec8885d4e0)) called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 25 eval {...} called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 25 at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP.pm line 36. B::Hooks::EndOfScope::PP::__invoke_callback(CODE(0x7fec8885d400)) called at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP/HintHash.pm line 39 B::Hooks::EndOfScope::PP::_SG_STACK::DESTROY(B::Hooks::EndOfScope::PP::_SG_STACK=ARRAY(0x7fec8885d4e0)) called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 25 eval {...} called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 25 at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP.pm line 36. B::Hooks::EndOfScope::PP::__invoke_callback(CODE(0x7fec8885d400)) called at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/PP/HintHash.pm line 39 B::Hooks::EndOfScope::PP::_SG_STACK::DESTROY(B::Hooks::EndOfScope::PP::_SG_STACK=ARRAY(0x7fec8885d4e0)) called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 25 eval {...} called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 25 FAIL ! Installing namespace::clean failed. See /Users/ether/.cpanm/work/1443220556.69252/build.log for details. Retry with --force to force install it. ... ----------- Building and testing namespace-clean-0.25 cp lib/namespace/clean.pm blib/lib/namespace/clean.pm Manifying 1 pod document PERL_DL_NONLAZY=1 "/Users/ether/perl5/perlbrew/perls/8.1/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/00-basic.t ............. ok t/01-function-wipeout.t .. ok t/02-inheritance.t ....... ok t/03-unimport.t .......... ok t/04-except.t ............ ok t/05-syntax-error.t ...... ok t/08-const-sub.t ......... ok t/06-other-types.t ....... ok namespace::clean::deleted::Foo is not a module name at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 227. at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 227. namespace::clean::__ANON__[/Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm:252]("Foo", HASH(0x7feac103d2f0), "foo", "baz") called at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 315 namespace::clean::__ANON__[/Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm:316]() called at /Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/XS.pm line 17 B::Hooks::EndOfScope::XS::__ANON__[/Users/ether/.perlbrew/libs/8.1@std/lib/perl5/B/Hooks/EndOfScope/XS.pm:17](HASH(0x7feac1818790), ARRAY(0x7feac1061780)) called at t/07-debugger.t line 25 eval {...} called at t/07-debugger.t line 25 Execution of t/07-debugger.t aborted due to compilation errors. at t/07-debugger.t line 32. t/07-debugger.t .......... Dubious, test returned 255 (wstat 65280, 0xff00) No subtests run t/09-fiddle-hinthash.t ... ok t/05-explicit-cleanee.t .. ok # Failed test 'foo cleaned up' # at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 27. # Failed test 'baz cleaned up' # at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t line 28. # Looks like you failed 2 tests of 4. # Failed test 'Exit 512 from: /Users/ether/perl5/perlbrew/perls/8.1/bin/perl /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/t/07-debugger.t' # at t/10-pure-perl.t line 55. # Looks like you failed 1 test of 12. t/10-pure-perl.t ......... Dubious, test returned 1 (wstat 256, 0x100) Failed 1/12 subtests Test Summary Report ------------------- t/07-debugger.t (Wstat: 65280 Tests: 0 Failed: 0) Non-zero exit status: 255 Parse errors: No plan found in TAP output t/10-pure-perl.t (Wstat: 256 Tests: 12 Failed: 1) Failed test: 10 Non-zero exit status: 1 Files=12, Tests=2090, 1 wallclock secs ( 0.39 usr 0.04 sys + 1.23 cusr 0.24 csys = 1.90 CPU) Result: FAIL Failed 2/12 test programs. 1/2090 subtests failed.
On 2015-09-25 15:41:42, ETHER wrote: Show quoted text
> A cpantesters report has been submitted... > > However, there was also this on stdout: > > ... > Building and testing B-Hooks-EndOfScope-0.15 ... OK > Successfully installed B-Hooks-EndOfScope-0.15 > Building and testing namespace-clean-0.25 ... A scope-end callback > raised an exception, which can not be propagated when > B::Hooks::EndOfScope operates in pure-perl mode. Your program will > CONTINUE EXECUTION AS IF NOTHING HAPPENED AFTER THIS WARNING. Below is > the complete exception text, followed by a stack-trace of the callback > execution:
I'm not sure why the PP version was being run here -- Variable::Magic 0.58 was successfully installed just prior to BHEOS. And also from the build log, I see that the XS version of BHEOS was being tested: Building and testing B-Hooks-EndOfScope-0.15 cp lib/B/Hooks/EndOfScope.pm blib/lib/B/Hooks/EndOfScope.pm cp lib/B/Hooks/EndOfScope/XS.pm blib/lib/B/Hooks/EndOfScope/XS.pm cp lib/B/Hooks/EndOfScope/PP.pm blib/lib/B/Hooks/EndOfScope/PP.pm cp lib/B/Hooks/EndOfScope/PP/FieldHash.pm blib/lib/B/Hooks/EndOfScope/PP/FieldHash.pm cp lib/B/Hooks/EndOfScope/PP/HintHash.pm blib/lib/B/Hooks/EndOfScope/PP/HintHash.pm Manifying 3 pod documents PERL_DL_NONLAZY=1 "/Users/ether/perl5/perlbrew/perls/8.1/bin/perl" "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/00-basic.t .................... ok t/01-eval.t ..................... ok t/05-exception_xs.t ............. ok t/02-localise.t ................. ok t/06-exception_pp.t ............. skipped: Skipping PP fallback test in XS mode t/11-direct_xs.t ................ ok t/12-direct_pp.t ................ ok t/10-test_without_vm_pure_pp.t .. ok Let me know how I can help.
I've attached the entire build log as it may contain more things of use.
Subject: build.log
Download build.log
application/octet-stream 124.5k

Message body not shown because it is not plain text.

Subject: Re: [rt.cpan.org #107343] tests fail, darwin 5.8.1
Date: Sun, 27 Sep 2015 19:39:32 +0200
To: bug-namespace-clean [...] rt.cpan.org
From: Peter Rabbitson <ribasushi [...] cpan.org>
On 09/26/2015 12:41 AM, Karen Etheridge via RT wrote: Show quoted text
> namespace::clean::deleted::Foo is not a module name at /Users/ether/.cpanm/work/1443220556.69252/namespace-clean-0.25/blib/lib/namespace/clean.pm line 227. >
This is new breakage introduced by Package::Stash::XS 0.36+. It has not been detected until now, as the tests in question do not run unless Sub::Identify is available. Both of these will be rectified in commits coming up shortly.
On Sat Sep 26 00:46:34 2015, ETHER wrote: Show quoted text
> > I'm not sure why the PP version was being run here -- Variable::Magic > 0.58 was > successfully installed just prior to BHEOS. >
Any distribution with a dual implementation generally must run both sets of tests (normal and PP). So this part is not only expected, but not having it would constitute a bug. You raising this issue prompted me to double check what I run, and realized that Package::Stash::PP is not being smoked - this has now been rectified as well: https://metacpan.org/source/RIBASUSHI/namespace-clean-0.25_01/t/10-pure-perl.t#L5-27
On Sun Sep 27 19:39:47 2015, RIBASUSHI wrote: Show quoted text
> This is new breakage introduced by Package::Stash::XS 0.36+. > It has not been detected until now, as the tests in question do not > run > unless Sub::Identify is available. > > Both of these will be rectified in commits coming up shortly.
The main issue has been fixed by hard-coding Package::Stash::PP to always run on perls < 5.8.7: https://metacpan.org/source/RIBASUSHI/namespace-clean-0.25_01/lib/namespace/clean/_Util.pm#L17-40 This will stave off future problems until RT#74151 is resolved Additionally t/07-debugger.t is now skipped only when there is no way to make it run: https://metacpan.org/source/RIBASUSHI/namespace-clean-0.25_01/t/07-debugger.t#L3-12 Please test https://metacpan.org/release/RIBASUSHI/namespace-clean-0.25_01 on a fresh perl when time permits, should be all covered. Cheers!
I've tested namespace-clean-0.25_02 on darwin on: 5.8.1, 5.8.5, 5.8.6, 5.8.8, 5.8.9, 5.10.0, 5.10.1, 5.12.5, 5.14.4, 5.16.3, 5.18.4, 5.20.3, 5.22.0, 5.23.3 and tests pass. Unfortunately cpanm-reporter refuses to send a report for some unknown reason, however -- there is something in the build.log that it is not recognizing.
On 2015-09-28 00:26:46, RIBASUSHI wrote: Show quoted text
> Please test https://metacpan.org/release/RIBASUSHI/namespace-clean- > 0.25_01 on a fresh perl when time permits, should be all covered.
Please add a $VERSION = eval $VERSION; line to avoid warnings like this all over the place: Argument "0.25_02" isn't numeric in subroutine entry at /Users/ether/.perlbrew/libs/8.5@std/lib/perl5/namespace/autoclean.pm line 12.
On 2015-09-30 18:45:56, ETHER wrote: Show quoted text
> On 2015-09-28 00:26:46, RIBASUSHI wrote: >
> > Please test https://metacpan.org/release/RIBASUSHI/namespace-clean- > > 0.25_01 on a fresh perl when time permits, should be all covered.
> > Please add a $VERSION = eval $VERSION; line to avoid warnings like > this all over the place: > > Argument "0.25_02" isn't numeric in subroutine entry at > /Users/ether/.perlbrew/libs/8.5@std/lib/perl5/namespace/autoclean.pm > line 12.
Probably better to write it $VERSION =~ s/_//; The current approach would turn 0.26_10 into 0.261, losing the trailing zero.
Subject: Re: [rt.cpan.org #107343] tests fail, darwin 5.8.1
Date: Thu, 08 Oct 2015 09:35:27 +0200
To: bug-namespace-clean [...] rt.cpan.org
From: Peter Rabbitson <ribasushi [...] cpan.org>
On 10/08/2015 07:42 AM, Slaven_Rezic via RT wrote: Show quoted text
> Queue: namespace-clean > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=107343 > > > On 2015-09-30 18:45:56, ETHER wrote:
>> On 2015-09-28 00:26:46, RIBASUSHI wrote: >>
>>> Please test https://metacpan.org/release/RIBASUSHI/namespace-clean- >>> 0.25_01 on a fresh perl when time permits, should be all covered.
>> Please add a $VERSION = eval $VERSION; line to avoid warnings like >> this all over the place: >> >> Argument "0.25_02" isn't numeric in subroutine entry at >> /Users/ether/.perlbrew/libs/8.5@std/lib/perl5/namespace/autoclean.pm >> line 12.
> Probably better to write it > > $VERSION =~ s/_//; > > The current approach would turn 0.26_10 into 0.261, losing the trailing zero.
Which should be of no consequence - devrels do not observe the "same amount of digits" rule anyway, so the problem of "losing a trailing zero" is moot anyway. I prefer the eval as a future-proofing technique - god knows what will come out of the version.pm mess. As an additional data point I have, in the past shipped trailing 0's like this with no observed adverse effect: https://metacpan.org/release/RIBASUSHI/DBIx-Class-0.082899_10 Thus I am very reluctant to incorporate the change you are proposing, unless you can demonstrate an actual tangible problem with the current approach. Cheers!
Subject: Re: [rt.cpan.org #107343] tests fail, darwin 5.8.1
Date: Thu, 8 Oct 2015 09:43:55 -0700
To: bug-namespace-clean [...] rt.cpan.org
From: Karen Etheridge <ether [...] cpan.org>
I encountered an issue last year with the Test-Simple trials and losing a trailing zero, but I do not recall what the specific problem was. It was probably something being too finicky in my tests and trying to do an exact string comparison somewhere. On Thu, Oct 8, 2015 at 12:35 AM, Peter Rabbitson via RT < bug-namespace-clean@rt.cpan.org> wrote: Show quoted text
> <URL: https://rt.cpan.org/Ticket/Display.html?id=107343 > > > On 10/08/2015 07:42 AM, Slaven_Rezic via RT wrote:
> > Queue: namespace-clean > > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=107343 > > > > > On 2015-09-30 18:45:56, ETHER wrote:
> >> On 2015-09-28 00:26:46, RIBASUSHI wrote: > >>
> >>> Please test https://metacpan.org/release/RIBASUSHI/namespace-clean- > >>> 0.25_01 on a fresh perl when time permits, should be all covered.
> >> Please add a $VERSION = eval $VERSION; line to avoid warnings like > >> this all over the place: > >> > >> Argument "0.25_02" isn't numeric in subroutine entry at > >> /Users/ether/.perlbrew/libs/8.5@std/lib/perl5/namespace/autoclean.pm > >> line 12.
> > Probably better to write it > > > > $VERSION =~ s/_//; > > > > The current approach would turn 0.26_10 into 0.261, losing the trailing
> zero. > > Which should be of no consequence - devrels do not observe the "same > amount of digits" rule anyway, so the problem of "losing a trailing > zero" is moot anyway. I prefer the eval as a future-proofing technique - > god knows what will come out of the version.pm mess. > > As an additional data point I have, in the past shipped trailing 0's > like this with no observed adverse effect: > https://metacpan.org/release/RIBASUSHI/DBIx-Class-0.082899_10 > > Thus I am very reluctant to incorporate the change you are proposing, > unless you can demonstrate an actual tangible problem with the current > approach. > > Cheers! > >
On 2015-10-08 03:35:48, RIBASUSHI wrote: Show quoted text
> On 10/08/2015 07:42 AM, Slaven_Rezic via RT wrote:
> > Queue: namespace-clean > > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=107343 > > > > > On 2015-09-30 18:45:56, ETHER wrote:
> >> On 2015-09-28 00:26:46, RIBASUSHI wrote: > >>
> >>> Please test https://metacpan.org/release/RIBASUSHI/namespace-clean- > >>> 0.25_01 on a fresh perl when time permits, should be all covered.
> >> Please add a $VERSION = eval $VERSION; line to avoid warnings like > >> this all over the place: > >> > >> Argument "0.25_02" isn't numeric in subroutine entry at > >> /Users/ether/.perlbrew/libs/8.5@std/lib/perl5/namespace/autoclean.pm > >> line 12.
> > Probably better to write it > > > > $VERSION =~ s/_//; > > > > The current approach would turn 0.26_10 into 0.261, losing the > > trailing zero.
> > Which should be of no consequence - devrels do not observe the "same > amount of digits" rule anyway, so the problem of "losing a trailing > zero" is moot anyway. I prefer the eval as a future-proofing technique > - > god knows what will come out of the version.pm mess. > > As an additional data point I have, in the past shipped trailing 0's > like this with no observed adverse effect: > https://metacpan.org/release/RIBASUSHI/DBIx-Class-0.082899_10 > > Thus I am very reluctant to incorporate the change you are proposing, > unless you can demonstrate an actual tangible problem with the current > approach.
I was bitten by the problem in the Tk-804.030 release. But on a closer look it seems that the problem was caused by non-standard parsing and handling of VERSION and DISTVERSION in Tk's Makefile.PL... after all, I couldn't reproduce the problem with a clean testcase just using VERSION_FROM.
Subject: tests fail, now on 5.8.8
This one isn't fixed yet, I am failing it somehow on one of my travis smokes. Investigation pending... will be fixed in 0.27 tomorrow-ish. https://travis-ci.org/ribasushi/dbix-class/jobs/84409094 https://travis-ci.org/dbsrgits/dbix-class/jobs/84419442
Any progress on resolving this?
On Tue Mar 01 17:52:10 2016, DAGOLDEN wrote: Show quoted text
> Any progress on resolving this?
Sorry I placed this on the backburner - 5.8.8 doesn't fail as such, only 5.8.8 with -Duselongdouble does. I will get a devrel later today. Thanks for the poke.
On Tue Mar 01 17:54:44 2016, RIBASUSHI wrote: Show quoted text
> > I will get a devrel later today.
"Later today" turned out to be 5 weeks later :( A devrel is out, smokes clean on 5.8.8. There are a couple other enhancements, so I want ot give it a couple days. Stable 0.27 Tue-ish. Sorry for the "forever"
On Sat Apr 09 23:43:33 2016, RIBASUSHI wrote: Show quoted text
> On Tue Mar 01 17:54:44 2016, RIBASUSHI wrote:
> > > > I will get a devrel later today.
> > "Later today" turned out to be 5 weeks later :( > A devrel is out, smokes clean on 5.8.8. There are a couple other > enhancements, so I want ot give it a couple days. Stable 0.27 Tue-ish. > > Sorry for the "forever"
It's a bit beyond Tuesday.
RT-Send-CC: dagolden [...] cpan.org
On Wed Apr 27 21:11:21 2016, DAGOLDEN wrote: Show quoted text
> On Sat Apr 09 23:43:33 2016, RIBASUSHI wrote:
> > On Tue Mar 01 17:54:44 2016, RIBASUSHI wrote:
> > > > > > I will get a devrel later today.
> > > > "Later today" turned out to be 5 weeks later :( > > A devrel is out, smokes clean on 5.8.8. There are a couple other > > enhancements, so I want ot give it a couple days. Stable 0.27 Tue-ish. > > > > Sorry for the "forever"
> > It's a bit beyond Tuesday. >
Terribly sorry it took forever. Now fixed for good (and verified on my anal smokers). P.S. @xdg that puppy is still waiting: https://github.com/dagolden/Capture-Tiny/pull/42