Bug#825475: perl: regular expression wide character and problem with catch group $1

[Niko Tyni]

On Fri, May 27, 2016 at 08:16:02AM +0200, Leszek Dubiel wrote:
> Package: perl
> Version: 5.20.2-3+deb8u4
> Severity: normal
> 
> Dear Maintainer,

> 	$x =~ s/([[:alpha:]])([[:alpha:]])/lc($1) . lc($2); /ge; 
> 	$y =~ s/([[:alpha:]])([[:alpha:]])/my $z = $1; lc($z) . lc($2); /ge; 
> 
> 	print "$x\n$y\n"; 
> 
> prints: 
> 
> 	1ąąą
> 	ąąąą

Thanks for the report. This was fixed upstream in 5.21.11 by
 http://perl5.git.perl.org/perl.git/commit/3553f4fa11fd9e8bb0797ace43605cc33ebf32fa
and in 5.20.3 by
 http://perl5.git.perl.org/perl.git/commit/17d9707d444517764c7bcb479c236a8c58a1d605

The fix has been backported for the Debian jessie packages,
and it is included in 5.20.2-3+deb8u5, which will be pushed
out on the next jessie point release. The upcoming package
is already available in jessie-proposed-updates, see
https://www.debian.org/releases/proposed-updates

If you want to help, testing of the updated packages would be appreciated.
-- 
Niko Tyni   ntyni at debian.org