Bug #96321 for CPANPLUS-Dist-Arch: Uninitialised value in pattern match in

Sat Jun 07 18:57:09 2014 cpandistarch [...] volkerschatz.com - Ticket created

Subject:	Uninitialised value in pattern match in _scanvspec()
Date:	Sun, 8 Jun 2014 00:56:50 +0200 (CEST)
To:	bug-CPANPLUS-Dist-Arch [...] rt.cpan.org
From:	Volker Schatz <cpandistarch [...] volkerschatz.com>

Hi Justin, I am having trouble with CPANPLUS::Dist::Arch when trying to install Text::LevenshteinXS. It aborts with the error message: [ERROR] Use of uninitialized value $vspec in pattern match (m//) at /usr/share/perl5/vendor_perl/CPANPLUS/Dist/Arch.pm line 1065. This is inside the _scanvspec() function. A possibly related earlier message is: Unparsable version '' for prerequisite Test at /home/vs/.cpanplus/5.20.0/build/Text-LevenshteinXS-0.03/Makefile.PL line 4. It looks to me as though Text::LevenshteinXS has a non-versioned dependency and CPANPLUS::Dist::Arch cannot cope with that. Best regards, Volker

Sun Jun 08 17:37:24 2014 jrcd83 [...] gmail.com - Correspondence added

Subject:	Re: [rt.cpan.org #96321] Uninitialised value in pattern match in _scanvspec()
Date:	Sun, 8 Jun 2014 17:37:16 -0400
To:	bug-CPANPLUS-Dist-Arch [...] rt.cpan.org
From:	Justin Davis <jrcd83 [...] gmail.com>

Volker, Thank you very much for the bug report! I am investigating this bug right now. I will try to fix this as quickly as I can. Justin

Sun Jun 08 17:37:25 2014 The RT System itself - Status changed from 'new' to 'open'

Sun Jun 08 18:02:59 2014 jrcd83 [...] gmail.com - Correspondence added

Subject:	Re: [rt.cpan.org #96321] Uninitialised value in pattern match in _scanvspec()
Date:	Sun, 8 Jun 2014 18:02:50 -0400
To:	bug-CPANPLUS-Dist-Arch [...] rt.cpan.org
From:	Justin Davis <jrcd83 [...] gmail.com>

I think I found the cause of the problem and you were dead on. I should have anticipated an undefined version value. After I test my solution I will submit a new release on CPAN. Might I also suggest avoiding Text::LevenshteinXS and maybe trying Text::Levenshtein::Damerau which also has an XS version? The former is a tiny wrapper over a tiny C function, lacks tests, and apparently breaks on UTF8 strings, while the latter has several tests, a nice interface, and works on UTF8. On Sun, Jun 8, 2014 at 5:37 PM, Justin Davis via RT <bug-CPANPLUS-Dist-Arch@rt.cpan.org> wrote: Show quoted text

> Queue: CPANPLUS-Dist-Arch > Ticket <URL: https://rt.cpan.org/Ticket/Display.html?id=96321 > > > Volker, > > Thank you very much for the bug report! I am investigating this bug > right now. I will try to fix this as quickly as I can. > > Justin >

-- -Justin

Sun Jun 08 18:26:39 2014 cpandistarch [...] volkerschatz.com - Correspondence added

Subject:	Re: [rt.cpan.org #96321] Uninitialised value in pattern match in _scanvspec()
Date:	Mon, 9 Jun 2014 00:26:17 +0200 (CEST)
To:	Justin Davis via RT <bug-CPANPLUS-Dist-Arch [...] rt.cpan.org>
From:	Volker Schatz <cpandistarch [...] volkerschatz.com>

Thank you for the quick reaction! Also thanks for pointing me to Text::Levenshtein::Damerau, though this is a slightly different algorithm. I have yet to test Text::LevenshteinXS with unicode chars; one example in the manual suggests it is supposed to work with non-ASCII characters, for what that is worth.

Wed Aug 12 10:41:44 2015 JUSTER [...] cpan.org - Taken

Wed Aug 12 10:43:11 2015 JUSTER [...] cpan.org - Correspondence added

I forgot to close this! Sorry.

Wed Aug 12 10:43:14 2015 JUSTER [...] cpan.org - Status changed from 'open' to 'resolved'

Bug #96321 for CPANPLUS-Dist-Arch: Uninitialised value in pattern match in _scanvspec()