Hi,
thanks for the fast follow up.
when using Clone 0.43, I now all of a sudden when running tests I see
"Unsupported magic_ptr clone at lib/Geo/Address/Formatter.pm line 721."
this is the line in question:
https://github.com/OpenCageData/perl-Geo-Address-Formatter/blob/master/lib/Geo/Address/Formatter.pm#L721
my $context = clone($components);
In this specific test $components is a reference to hash that looks like
this:
{
'city' => 'Palo Alto',
'country' => 'USA',
'country_code' => 'US',
'county' => 'Santa Clara County',
'house_number' => '301',
'neighbourhood' => 'Crescent Park',
'postcode' => '94303',
'road' => 'Hamilton Ave',
'state' => 'California',
'state_code' => 'CA';
};
just strings as keys and values, nothing remarkable in my opinion.
Thanks for having a look
Ed
Am Mo., 29. Juli 2019 um 21:02 Uhr schrieb Breno G. de Oliveira via RT <
bug-Geo-Address-Formatter@rt.cpan.org>:
Show quoted text> Queue: Geo-Address-Formatter
> Ticket <URL:
https://rt.cpan.org/Ticket/Display.html?id=130195 >
>
> Hey, Slaven! Hi Ed!
>
> Thank you for letting me know. Clone 0.42 changes the way magic is cloned
> to fix a segmentation fault. It was not supposed to break anything and
> we're looking into it. I'll let you know when we release 0.43 so all you
> need to do (then) is set the minimal version to 0.43.
>
> Thanks again!
>
> garu
>
> On Mon Jul 29 12:27:44 2019, SREZIC wrote:
> > Adding GARU to the discussion.
> >
> > On 2019-07-29 06:49:43, edf@sloan.mit.edu wrote:
> > > yes, definitely possible, but would prefer to understand why Clone.pm
> > > all
> > > of a sudden thinks my code is in error.
> > > I did not touch that code, and as far as I can tell I am following
> > > their
> > > publicly stated usage.
> > >
> > > I am on it, though not with massive urgency as I am on holiday
> > >
> > > thanks,
> > > Ed
> > >
> > > Am Sa., 27. Juli 2019 um 15:45 Uhr schrieb Slaven_Rezic via RT <
> > > bug-Geo-Address-Formatter@rt.cpan.org>:
> > >
> > > > <URL:
https://rt.cpan.org/Ticket/Display.html?id=130195 >
> > > >
> > > > Another idea: maybe in your case Storable::dclone would also work as
> > > > a
> > > > cloning function. As a plus, Storable.pm is perl core module, so you
> > > > have a
> > > > dependency less.
> > > >
> > > > On 2019-07-27 09:37:40, SREZIC wrote:
> > > > > But in the CPAN world it is not possible to depend on an older
> > > > > version
> > > > > than the latest stable. I see that your Makefile.PL and META.*
> > > > > files
> > > > > now specify Clone => 0.41, but this means only *minimum* version.
> > > > > So
> > > > > if 0.42 is already installed, things will still fail.
> > > > >
> > > > > Do you know if the problem is a known issue in Clone.pm? Maybe it
> > > > > should be reported in their issue tracker?
> > > > >
> > > > > On 2019-07-27 08:21:47, edf@sloan.mit.edu wrote:
> > > > > > thanks for making me aware. Have released new version now which
> > > > > > requires
> > > > > > older version of Clone.pm
> > > > > >
> > > > > > Am Fr., 26. Juli 2019 um 19:29 Uhr schrieb Slaven_Rezic via RT <
> > > > > > bug-Geo-Address-Formatter@rt.cpan.org>:
> > > > > >
> > > > > > > Fri Jul 26 13:29:10 2019: Request 130195 was acted upon.
> > > > > > > Transaction: Ticket created by SREZIC
> > > > > > > Queue: Geo-Address-Formatter
> > > > > > > Subject: Unsupported magic_ptr clone (with Clone 0.42?)
> > > > > > > Broken in: 1.77
> > > > > > > Severity: (no value)
> > > > > > > Owner: Nobody
> > > > > > > Requestors: SREZIC@cpan.org
> > > > > > > Status: new
> > > > > > > Ticket <URL:
>
https://rt.cpan.org/Ticket/Display.html?id=130195
> > > > > > > >
> > > > > > >
> > > > > > >
> > > > > > > On some of my smokers the test suite fails:
> > > > > > >
> > > > > > > ...
> > > > > > > Unsupported magic_ptr clone at lib/Geo/Address/Formatter.pm
> > > > > > > line
> > > > > > > 643.
> > > > > > > # Tests were run but no plan was declared and done_testing()
> > > > > > > was
> > > > > > > not seen.
> > > > > > > # Looks like your test exited with 255 just after 25.
> > > > > > > t/blackbox/countries.t ..
> > > > > > > Dubious, test returned 255 (wstat 65280, 0xff00)
> > > > > > > All 25 subtests passed
> > > > > > > Unsupported magic_ptr clone at lib/Geo/Address/Formatter.pm
> > > > > > > line
> > > > > > > 643.
> > > > > > > # Tests were run but no plan was declared and done_testing()
> > > > > > > was
> > > > > > > not seen.
> > > > > > > # Looks like your test exited with 255 just after 3.
> > > > > > > t/unit/abbreviate.t .....
> > > > > > > Dubious, test returned 255 (wstat 65280, 0xff00)
> > > > > > > All 3 subtests passed
> > > > > > > ...
> > > > > > >
> > > > > > > This seems to happen with the most recent Clone.pm version.
> > > > > > > Statistical
> > > > > > > analysis (theta=-1 means "bad"):
> > > > > > >
> > > > > > >
> ****************************************************************
> > > > > > > Regression 'mod:Clone'
> > > > > > >
> ****************************************************************
> > > > > > > Name Theta StdErr T-stat
> > > > > > > [0='const'] 1.0000 0.0000
> > > > > > > 11034699526378720.00
> > > > > > > [1='eq_0.39'] -0.0000 0.0000 -1.06
> > > > > > > [2='eq_0.41'] -0.0000 0.0000 -3.36
> > > > > > > [3='eq_0.42'] -1.0000 0.0000
> > > > > > > -10564920452501012.00
> > > > > > >
> > > > > > > R^2= 1.000, N= 80, K= 4
> > > > > > >
> ****************************************************************
> > > > > > >
> > > > > > >
> > > >
> > > >
> > > >
> > > >
> >
> >
>
>
>
>
