rvosa / bio-phylo

Bio::Phylo - Phyloinformatic analysis using Perl
http://search.cpan.org/dist/Bio-Phylo
GNU General Public License v3.0
16 stars 6 forks source link

Issues with IO::Handle and perl 5.10 #19

Closed cjfields closed 8 years ago

cjfields commented 9 years ago

In BioPerl we're seeing some issues with Bio::Phylo using perl v 5.10:

https://travis-ci.org/bioperl/bioperl-live/jobs/47662896

Same in CPAN Testers for Bio::Phylo:

http://www.cpantesters.org/cpan/report/87d57d52-6bf9-1014-99a2-46f257c061e0

Seems very specific to the IO::Handle version, so maybe the fix is to require at least IO::Handle 1.28 (the one released w/ perl 5.12, which passes. May test this in BioPerl via Travis-CI

cjfields commented 9 years ago

May be more complicated than that; seems to pop up only under certain circumstances. Only one BioPerl test appears to be failing at the moment, so I can focus on that.

rvosa commented 9 years ago

Thanks! If you figure out what is happening let me know and I'll push out a fix.

On 19:03, Tue, Jan 20, 2015 Chris Fields notifications@github.com wrote:

May be more complicated than that; seems to pop up only under certain circumstances. Only one BioPerl test appears to be failing at the moment, so I can focus on that.

— Reply to this email directly or view it on GitHub https://github.com/rvosa/bio-phylo/issues/19#issuecomment-70701185.

rvosa commented 8 years ago

Hi @cjfields - did you ever manage to figure this out?

rvosa commented 8 years ago

I looked at the failing logs and for some reason it is unhappy about me doing $handle->eof() because I am getting Can't find object method... errors. However, I should equally be able to use the built in function, i.e. change the idiom to eof $handle. This appears to pass the tests on my end on travis (including 5.10) so I wonder if this is any help on your end?

cjfields commented 8 years ago

May need to check but I would assume this should fix it. Let me add a pointer to this in the bioperl issues queue.

cjfields commented 8 years ago

@rvosa I think this is a non-issue, apparently the travis test was rerun and doesn't show the same problem, I would ignore it. Closing this ticket.