Bug #107566 for GraphViz2: GraphViz2::Parse::Regexp default graph parameter

Tue Oct 06 04:01:46 2015 user42_kevin [...] yahoo.com.au - Ticket created

Subject:	GraphViz2::Parse::Regexp default graph parameter
Date:	Tue, 06 Oct 2015 19:01:15 +1100
To:	bug-GraphViz2 [...] rt.cpan.org
From:	Kevin Ryde <user42_kevin [...] yahoo.com.au>

The docs of GraphViz2::Parse::Regexp on the "graph" parameter say "This key is optional", but in recent debian i386 perl 5.20.2 attempting to do that with use strict; use GraphViz2::Parse::Regexp; my $parse = GraphViz2::Parse::Regexp->new; $parse->create(regexp => '(([abcd0-9])|(foo))'); gets an error Can't call method "add_node" on unblessed reference at /usr/share/perl5/GraphViz2/Parse/Regexp.pm line 93. Nosing around the code I wonder if the BUILD bit would be the "default" subr of the graph parameter ... but I don't know if that makes some subtle difference.

Tue Oct 06 05:52:57 2015 RSAVAGE [...] cpan.org - Correspondence added 15 min

Thanx for the report. V 2.36 is now on CPAN. V 2.35 was never released, so the new version should have been 2.35.

Tue Oct 06 05:52:57 2015 The RT System itself - Status changed from 'new' to 'open'

Tue Oct 06 05:52:59 2015 RSAVAGE [...] cpan.org - Status changed from 'open' to 'resolved'

Wed Oct 07 08:13:47 2015 user42_kevin [...] yahoo.com.au - Correspondence added

Subject:	Re: [rt.cpan.org #107566] GraphViz2::Parse::Regexp default graph parameter
Date:	Wed, 07 Oct 2015 23:12:11 +1100
To:	"RSAVAGE via RT" <bug-GraphViz2 [...] rt.cpan.org>
From:	Kevin Ryde <user42_kevin [...] yahoo.com.au>

"RSAVAGE via RT" <bug-GraphViz2@rt.cpan.org> writes: Show quoted text

> > Thanx for the report. V 2.36 is now on CPAN.

Oh, I should have mentioned I think the other Parse modules could be similar. Eg. use GraphViz2::Parse::ISA; my $parse = GraphViz2::Parse::ISA->new; $parse->add(class => 'GraphViz2::Parse::ISA', ignore => []); $parse->generate_graph; => Can't call method "dependency" on unblessed reference at /usr/share/perl5/GraphViz2/Parse/ISA.pm line 110. On the bright side, if it works then it might be a way to compactify the "synopsis" section of the pod :) ... let the parser make the graph. But maybe that depends what is more typical ... Incidentally, I notice in GraphViz2::Parse::ISA the "isa" property accessor shadows the UNIVERSAL.pm ->isa() method. If that property hasn't yet been a documented feature it might be worth renaming to avoid clashing. -- Biology jargon elucidated for the layman: "Parthenogenesis" -- cellular division resulting in a large Greek building.

Wed Oct 07 18:28:10 2015 ron [...] savage.net.au - Correspondence added

Subject:	Re: [rt.cpan.org #107566] GraphViz2::Parse::Regexp default graph parameter
Date:	Thu, 08 Oct 2015 09:35:33 +1100
To:	bug-GraphViz2 [...] rt.cpan.org
From:	Ron Savage <ron [...] savage.net.au>

Hi Kevin On 07/10/15 23:13, Kevin Ryde via RT wrote: Show quoted text

> Queue: GraphViz2 > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=107566 > > > "RSAVAGE via RT" <bug-GraphViz2@rt.cpan.org> writes:

>> >> Thanx for the report. V 2.36 is now on CPAN.

> > Oh, I should have mentioned I think the other Parse modules could be > similar. Eg.

Yes. That eventually occurred to me :-(. Sigh. I'll fix them shortly. Show quoted text

> use GraphViz2::Parse::ISA; > my $parse = GraphViz2::Parse::ISA->new; > $parse->add(class => 'GraphViz2::Parse::ISA', ignore => []); > $parse->generate_graph; > > => > Can't call method "dependency" on unblessed reference at /usr/share/perl5/GraphViz2/Parse/ISA.pm line 110. > > On the bright side, if it works then it might be a way to compactify the > "synopsis" section of the pod :) ... let the parser make the graph. But > maybe that depends what is more typical ...

I have no way of knowing exactly how people use the module, and any reports I get simply may not be representative. So I'll just make the same patch to all of them. Show quoted text

> Incidentally, I notice in GraphViz2::Parse::ISA the "isa" property > accessor shadows the UNIVERSAL.pm ->isa() method. If that property > hasn't yet been a documented feature it might be worth renaming to avoid > clashing.

Ahhgg. I hadn't thought of that. isa() is not documented, so I will change it. Thanx for noticing! -- Ron Savage - savage.net.au