Bug #16288 for Perl-Critic: Violation->new() in RequireTrailingCommas.pm

Mon Dec 05 04:00:00 2005 Guest - Ticket created

Subject:

Violation->new() in RequireTrailingCommas.pm

Hi, when running Test::Perl::Critic against my code, I get following error message. # Failed test 'blib/lib/Biomax/My/Module.pm' # in /BioInformatics/soft/perl/lib/site_perl/5.8.0/Test/Perl/Critic.pm at line 83. # # Perl::Critic had errors in 'blib/lib/Biomax/My/Module.pm': # 3rd arg to Violation->new() must be ARRAY ref at /BioInformatics/soft/perl/lib/site_perl/5.8.0/Perl/Critic/Policy/CodeLayout/RequireTrailingCommas.pm line 34 I suppose that the same incorrect call of Violation->new() is lurking in other rules as well.

Mon Dec 05 04:29:10 2005 Guest - Correspondence added

From:

thaljef [...] cpan.org

I haven't seen that before. I double checked the code and it looks reasonable. Also, none of my test cases seem to raise that error. Would you be willing to send me the code that you're running it on? Thanks. -Jeff

Mon Dec 05 04:37:14 2005 thaljef [...] cpan.org - Status changed from 'new' to 'open'

Mon Dec 05 04:37:15 2005 thaljef [...] cpan.org - Broken in 0.13 added

Mon Dec 05 04:37:19 2005 thaljef [...] cpan.org - Taken

Tue Dec 06 10:19:49 2005 Guest - Correspondence added

From:

Bernhard Schmalhofer

[guest - Mon Dec 5 04:29:10 2005]: Show quoted text

> I haven't seen that before. I double checked the code and it looks > reasonable. Also, none of my test cases seem to raise that error. > Would you be willing to send me the code that you're running it on?

Hi Jeff, I have boiled it down to a small test-case and attached the two files. Thinking about it, I find the method calls in hash construction is somewhat questionable by itself. Perhaps I should at least wrap that like my %cds_tag = ( history => join( "\n", @history ), note => scalar( $prot->desc() ), translation => scalar( $prot->seq() ) ); Thanks for that great Module. Right now I just use it for my own code. But I hope to have a handle now, to raise code quality of the whole company. I also have found a documentation glitch, and added a note about it on http://annocpan.org/~THALJEF/Perl-Critic-0.13/lib/Perl/Critic.pm. PS: Sorry that the reply is so late. Initially I replied to comment-Perl-Critic@rt.cpan.org, but it looks like the reply mail never made it back into RT.

use strict; use warnings; use 5.008; sub merge_prot_with_dna { my ( $prot ); my @history; my %cds_tag = ( history => join( "\n", @history ), note => $prot->desc(), translation => $prot->seq() ); return; }

Tue Dec 06 10:21:07 2005 Guest - Correspondence added

Subject:

driver script for Util.p,

And here is the scripts taht actually checks the policy.

# pragmata use strict; use 5.008; use Perl::Critic; # Create Critic and load Policy by hand my $critic = Perl::Critic->new(-profile => 'NONE'); $critic->add_policy( -policy => 'CodeLayout::RequireTrailingCommas' ); $critic->critique( 'Util.pm' );

Tue Dec 06 19:02:13 2005 thaljef [...] cpan.org - Correspondence added

Thanks for the code sample -- that helped a lot. I think I've traced this to a bug in PPI. For any PPI::Node, the location is given by the location of the first child, but the empty list "()" has no children, so the location is undefined. I'll bring this up with Adam Kennedy (who wrote PPI). In the mean time, I can probably find a workaround for the next release of Perl::Critic, which should be ready by the end of the month. If you want to follow the latest developments, we have a repository at http://perlcritic.tigris.org and a mailing list at dev@perlcritic.tigris.org. -Jeff

Thu Dec 29 13:52:21 2005 thaljef [...] cpan.org - Status changed from 'open' to 'resolved'

Thu Dec 29 13:52:22 2005 thaljef [...] cpan.org - Fixed in 0.13_01 added

Bug #16288 for Perl-Critic: Violation->new() in RequireTrailingCommas.pm

Preferred bug tracker