Bug #100790 for RDF-Trine-Node-Literal-XML: Use of reference as array index

Mon Dec 08 23:20:37 2014 ANDK [...] cpan.org - Ticket created

Subject:

Use of reference as array index

Bleadperl has a new warning that looks interesting. Bleadperl commit: v5.21.6-336-gfedf30e Sample fail report: http://www.cpantesters.org/cpan/report/c7ded7f8-7ed8-11e4-aaeb-9f7752b25338 HTH&&Thanks, (Disclaimer: this was found while running a cpan cmoker, I'm not talking as a user)

Mon Dec 08 23:22:24 2014 ANDK [...] cpan.org - Correspondence added

Maybe also of interest a ticket about other affected modules: https://rt.perl.org/rt3/Ticket/Display.html?id=123390

Tue Dec 09 06:13:18 2014 kjetilk [...] cpan.org - Correspondence added

Thanks a lot! Looking through the test report, I get the impression that the actual problem is in RDF::Trine, it says e.g. # Use of reference "HASH(0x2a40c58)" as array index at /home/sand/src/perl/repoperls/installed-perls/perl/v5.21.6-336-gfedf30e/9980/lib/site_perl/5.21.7/RDF/Trine/Store/Hexastore.pm line 308. # at /home/sand/src/perl/repoperls/installed-perls/perl/v5.21.6-336-gfedf30e/9980/lib/site_perl/5.21.7/RDF/Trine/Store/Hexastore.pm line 308. The only line it points out in this module is line 14, which is use base qw(RDF::Trine::Node::Literal); RDF::Trine::Node::Literal is also in the RDF-Trine distribution, and RDF::Trine loads a lot of its packages right there, so it is quite likely it loads ::Hexastore too. Could you please have a look and see if you agree. I can forward the bug to RDF::Trine if you haven't already.

Tue Dec 09 06:13:18 2014 The RT System itself - Status changed from 'new' to 'open'

Tue Dec 09 06:13:19 2014 kjetilk [...] cpan.org - Taken

Tue Dec 09 16:06:04 2014 ANDK [...] cpan.org - Correspondence added

I must admit, I had not looked at line 308 of Hexastore.pm when I sent you this report. Now I saw it and I would say, the warning from bleadperl is wrong. I'll report back to perl bugtracker and I'd suggest we close this issue here unless you come to a different conclusion. Look: https://metacpan.org/source/GWILLIAMS/RDF-Trine-1.011/lib/RDF/Trine/Store/Hexastore.pm#L308 https://metacpan.org/source/GWILLIAMS/RDF-Trine-1.011/lib/RDF/Trine/Store/Hexastore.pm#L38 The two relevant lines are 308 and 38 and the term NODEMAP->{ $_ } is always a scalar in the range 0..3, never a reference. So the warning looks wrong. I'm sorry for the false alarm noise and thank you for the quick reaction! I leave the ticket open for now, please close it if you agree. Thanks!

Tue Dec 09 16:48:18 2014 kjetilk [...] cpan.org - Correspondence added

ty. 09. des. 2014 16.06.04 skreiv ANDK: Show quoted text

> I must admit, I had not looked at line 308 of Hexastore.pm when I sent > you this report. Now I saw it and I would say, the warning from > bleadperl is wrong. I'll report back to perl bugtracker and I'd > suggest we close this issue here unless you come to a different > conclusion.

Yeah, I couldn't see anything wrong there, so my guess would be that it is not our bug. :-) But then, I don't follow bleadperl, so I can't be sure, but if it is your conclusion, I think it is most likely not here. Show quoted text

> I'm sorry for the false alarm noise and thank you for the quick reaction!

No problem at all! I find service you provide to the community so valuable, that an occasional false alarm is a small price to pay! :-) Cheers, Kjetil

Tue Dec 09 16:48:20 2014 kjetilk [...] cpan.org - Status changed from 'open' to 'rejected'