Bug #120912 for Regexp-Debugger: Regexp::Debugger bug

Mon Apr 03 16:31:15 2017 dpettit [...] bluehost.com - Ticket created

Subject:	Regexp::Debugger bug
Date:	Mon, 3 Apr 2017 14:30:59 -0600
To:	bug-regexp-debugger [...] rt.cpan.org
From:	Daina Pettit <dpettit [...] bluehost.com>

I tried to submit a bug report on rt.cpan.org but it failed to record the bug. I've had trouble in the past year or so to get Regexp::Debugger to work. Even with simple stuff it blows up. I get a core dump with the following test case for rxrx: *'fdas dfsa dfsa adfs sdfa fdsa asdf sdfa dfas sdfa fsda dfsa dfsa dfas fdsa dsfa fdas dafs dfas'** **/a/* Like this: substr outside of string at /usr/local/share/perl5/Regexp/Debugger.pm line 2183, <> line 6. Segmentation fault (core dumped) For the much shorter match string *'a'* but the same RE I get Use of uninitialized value $str_pos in subtraction (-) at /usr/local/share/perl5/Regexp/Debugger.pm line 1775, <> line 4. This is running Regexp::Debugger 0.001022 Perl 5.10.1 CentOS 6.7 Linux -- <http://bluehost.com/> Daina Pettit | Senior Developer | 801.765.9400 x5704

Mon Apr 03 17:57:52 2017 damian [...] conway.org - Correspondence added

Subject:	Re: [rt.cpan.org #120912] Regexp::Debugger bug
Date:	Tue, 4 Apr 2017 07:56:48 +1000
To:	bug-Regexp-Debugger [...] rt.cpan.org
From:	Damian Conway <damian [...] conway.org>

Hi Daina, I'm so sorry you've had difficulties with Regexp::Debugger. I've been working on some significant improvements and enhancements to the module, and your message prompted me to make a new CPAN release of the work-in-progress. Please give version 0.002000 a try (once it propagates) and see if it works any better for you. All the best, Damian

Mon Apr 03 17:57:53 2017 The RT System itself - Status changed from 'new' to 'open'

Mon Apr 03 19:21:51 2017 DAINA [...] cpan.org - Correspondence added

RT-Send-CC:

damian [...] conway.org

Updated to 0.002000 and tried the two test cases and it still fails both in the same way. Checked dependencies again and updated/installed just to be sure. It worked! However, captures and other information (like number of steps) don't remain on the screen at the end. I thought they used to stay on the screen.

Tue Apr 04 00:24:18 2017 DCONWAY [...] cpan.org - Status changed from 'open' to 'resolved'

Tue Apr 04 00:24:31 2017 damian [...] conway.org - Correspondence added

Subject:	Re: [rt.cpan.org #120912] Regexp::Debugger bug
Date:	Tue, 4 Apr 2017 14:23:36 +1000
To:	bug-Regexp-Debugger [...] rt.cpan.org
From:	Damian Conway <damian [...] conway.org>

Show quoted text

> It worked! However, captures and other information (like number of > steps) don't remain on the screen at the end. I thought they used to > stay on the screen.

They stay on the screen at the final step, but if you quit matching (either explicitly with 'q' or implicitly by stepping past the end), then the captures and other info disappear. This is unchanged from earlier versions. Damian

Tue Apr 04 12:37:48 2017 DAINA [...] cpan.org - Correspondence added

RT-Send-CC:

damian [...] conway.org

On Tue Apr 04 00:24:31 2017, damian@conway.org wrote: Show quoted text

> > It worked! However, captures and other information (like number of > > steps) don't remain on the screen at the end. I thought they used to > > stay on the screen.

> > They stay on the screen at the final step, but if you quit matching > (either explicitly with 'q' or implicitly by stepping past the end), then > the captures and other info disappear. This is unchanged from earlier > versions.

'c' - continue will run through the steps all the way to the end and *then* clear away the results. That's normal?

Tue Apr 04 12:40:39 2017 DAINA [...] cpan.org - Correspondence added

On Tue Apr 04 12:37:48 2017, DAINA wrote: Show quoted text

> On Tue Apr 04 00:24:31 2017, damian@conway.org wrote:

> > > It worked! However, captures and other information (like number of > > > steps) don't remain on the screen at the end. I thought they used > > > to > > > stay on the screen.

> > > > They stay on the screen at the final step, but if you quit matching > > (either explicitly with 'q' or implicitly by stepping past the end), > > then > > the captures and other info disappear. This is unchanged from earlier > > versions.

> > 'c' - continue will run through the steps all the way to the end and > *then* clear away the results. That's normal?

Oh, I see what happens. I hit 'return' *after* the 'c' and that 'return' is queued up and *that's* what clears the screen once the continue completes. Sorry for the excitement. Thanks for your replies.