| Subject: | Doesn't handle closing parenthesis in charracter class correctly |
Hi,
1 bugs is reported by a Debian user :
This code doesn't work properly, but should without problem
--->8---
use strict;
use Regexp::Assemble;
my $ra = Regexp::Assemble->new();
$ra->add('([^)])s-[01]');
print $ra->re;
---8<---
It looks like Regexp::Assemble thinks that ) in [^)] need to be escaped
(It doesn't - only ]$^-\ have special mining and may need escaping (but
sometimes doesn't need)) but places escape character in wrong position:
---8<---
Building: (?:([^)\])s-[01]|((x)y))
Invalid [] range "s-[" in regex; marked by <-- HERE in m/(?:([^)\])s-[
<-- HERE
01]|((x)y))/ at /usr/share/perl5/Regexp/Assemble.pm line 1004.
--->8---
After escaping closing ], character class is still going on, and s-[ is
interpreted
as character class range, but s-[ is invalid due to the fact that s is
AFTER [
in encoding.
(Building message is just print STDERR "Building: $str\n";
added to sub _build_re in Regexp/Assemble.pm line 1004., to make debugging
easier).
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=675984