New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Bleadperl v5.19.2-257-gc30fc27 breaks FANGLY/Getopt-Euclid-0.4.3.tar.gz #13145
Comments
From @andkgit bisect commit c30fc27 Handle /[#]/ and /[(?#]/ with code blocks diagnostics http://www.cpantesters.org/cpan/report/ee513274-fa8f-11e2-af9a-37acf1ff63fb perl -V Summary of my perl5 (revision 5 version 19 subversion 3) configuration: Characteristics of this binary (from libperl): -- |
From @rjbs* "Andreas J. Koenig via RT" <perlbug-followup@perl.org> [2013-08-02T04:04:53]
So, a patch meant to 1/3 fix Damian's regexp libraries 100% breaks his getopt -- |
The RT System itself - Status changed from 'new' to 'open' |
From @iabynOn Fri, Aug 02, 2013 at 08:39:43AM -0400, Ricardo Signes wrote:
It appears to be a bug in Getopt::Euclid, which was masked by a bug in Consider the following code: my $x = 'abc'; print "hash in char class is not a comment\n" if "#" =~ /[#]/x; which prior to the 'Handle /[#]/ and /[(?#]/ with code blocks' fix (?^:[abc]-[#abc]) (?^x:[abc]-[#$x]) and after, outputs: (?^:[abc]-[#abc]) (?^x:[abc]-[#abc]) Note that in both cases, perl agrees that a '#' within a charclass ([]), Further, note that in the non-extended case both before and after, '$x' However in the extended case, previously a '#' before a variable name In the case of Getopt::Euclid, it has a regex with the char class The fix for Getopt::Euclid is simple; the following diff makes all tests Inline Patch--- lib/Getopt/Euclid.pm- 2013-08-02 23:17:31.624828488 +0100
+++ lib/Getopt/Euclid.pm 2013-08-02 23:17:52.508088261 +0100
@@ -1055,7 +1055,7 @@
sub _escape_specials {
# Escape quotemeta special characters
my $arg = shift;
- $arg =~ s{([@#$^*()+{}?])}{\\$1}gxms; #?
+ $arg =~ s{([@#\$^*()+{}?])}{\\$1}gxms; #?
return $arg;
}
-- |
From @cpansproutOn Fri Aug 02 15:42:07 2013, davem wrote:
If we remove your fix from the maint branch, we retain a serious If we leave it in, we introduce a tiny ‘regression’ that doesn’t count Other maintenance releases have had bug fixes (surprise!) and every bug So I say leave it in. But what does my opinion count? -- Father Chrysostomos |
From @rjbs* Dave Mitchell <davem@iabyn.com> [2013-08-02T18:41:27]
This is my feeling, too, but I'll read this again in the morning when my brain Thanks. -- |
From @andk(Andreas J. Koenig) (via RT) <perlbug-followup@perl.org> writes:
Another affected module: ANDREWF/LaTeX-Encode-0.08.tar.gz -- |
From @iabynOn Sat, Aug 03, 2013 at 08:15:33AM +0200, Andreas Koenig wrote:
I haven't tried building and testing it, but a quick grep of the src Encode/EncodingTable.pm: which looks like another "bare $var inadvertently protected by a I wonder how many more of these there are. I tried doing a search on -- |
From @iabynOn Fri, Aug 02, 2013 at 06:09:41PM -0700, Father Chrysostomos via RT wrote:
The third option is that I could (probably) modify the fix in maint The main thing I don't like about the fix for for the long-term bug is Perhaps in blead we should have deprecation cycle, where initially -- |
From @maukeOn 03.08.2013 11:13, Dave Mitchell wrote:
I found lots of false positives but also: HTML::StripScripts - DBD::PO - Mail::SpamAssassin::Plugin::FreeMail - Net::validMX - Tripletail::DateTime - On the other hand, SGML::Parser - -- |
From @rjbs* Dave Mitchell <davem@iabyn.com> [2013-08-03T05:22:39]
If you can do that, that would be nice.
I think we're better off with just listening to the BBC for breakage, so far. -- |
From @iabynOn Sat, Aug 03, 2013 at 08:30:37AM -0400, Ricardo Signes wrote:
Now done and pushed as davem/maint-5.18-119125 -- |
From @nwc10On Fri, Aug 02, 2013 at 11:41:27PM +0100, Dave Mitchell wrote:
A bit older than that. But it turns out that it is actually a regression: bisect.pl --start=perl-5.000 --end=perl-5.002 --target=miniperl -e 'unless ( and the culprit is: commit 748a930 Perl 5.001 I suspect that it's this one: +NETaa13369: # is now a comment character, and \# should be left for regcomp. but we don't have any granularity on the changes. And we don't have access
I like the approach that I think you've taken subsequently to this message - I think that fixing this in a maintenance release is going to cause a few too Nicholas Clark |
From @cpansproutThis is resolved in 5.18.1 by commit 0268238. -- Father Chrysostomos |
From [Unknown Contact. See original ticket]This is resolved in 5.18.1 by commit 0268238. -- Father Chrysostomos |
@cpansprout - Status changed from 'open' to 'resolved' |
From @cpansproutOn Mon Aug 12 20:13:20 2013, andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:
Yes, it does. You’re right. The bug here is that the modules are buggy and need patches. -- Father Chrysostomos |
@cpansprout - Status changed from 'resolved' to 'open' |
From @cpansproutOn Thu Aug 15 06:07:27 2013, andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:
https://metacpan.org/source/EHOOD/perlSGML.1997Sep18/lib/SGML/Parser.pm#L216
Thank you for taking care of that. I have added most of these to the -- Father Chrysostomos |
@cpansprout - Status changed from 'open' to 'resolved' |
From @iabynOn Sat, Aug 03, 2013 at 04:08:22PM +0100, Dave Mitchell wrote:
Just reviving an old ticket. Ricardo privately pointed out that even with I've just pushed out the branch commit 40e7e8d RT #119125: fix two issues with/[#]/x -- |
From @jkeenanIf we're "reviving an old ticket," then we have to change its status in RT back to "open". |
@jkeenan - Status changed from 'resolved' to 'open' |
From @rjbs* Dave Mitchell <davem@iabyn.com> [2014-02-18T14:02:01]
Thanks, with this, Damian's test program passes. I will apply this to maint-5.18 before 5.18.3, and probably quite soon. -- |
From @tonycozOn Fri Feb 21 19:32:52 2014, perl.p5p@rjbs.manxome.org wrote:
Apparently you applied it as 1f621a8. Tony |
From @jkeenanOn Fri Feb 21 19:32:52 2014, perl.p5p@rjbs.manxome.org wrote:
I'm trying to determine whether this older ticket is closable. I went to matrix.cpantesters.org and tried to locate reports for the most recent officially released versions of CPAN distributions mentioned over the course of this RT's existence. Here are my impressions: * Getopt-Euclid 0.4.5 (latest distribution): both blead (5.21.5) and 5.18.3 doing well So, with one exception, these distributions are either (a) well maintained and consistently PASS a very high majority of their smoke test runs; or (b) they're unmaintained junk that P5P shouldn't worry about. The one exception: This distribution, though not widely smoked, had consistent PASSes from 5.17.10 to 5.21.1. It has consistently FAILed in recent months. (See: http://matrix.cpantesters.org/?dist=HTML%3A%3AStripScripts) So this is the only one that I think P5P might be on the hook for. Thank you very much. -- |
From @jkeenanOn Fri Sep 26 19:11:06 2014, jkeenan wrote:
And on closer inspection, I don't think we're on the hook for HTML::StripScripts, either. ##### ... and the smoke tests are die-ing on: ##### So if the author were to correct the situation giving the warning, this distribution would probably PASS once again (though the fatalization of warnings is sub-optimal, IMO). Thank you very much. -- |
From @rjbsI think this ticket has reached the end of its usefulness. Thanks very much for your previous investigations, Jim. -- |
@rjbs - Status changed from 'open' to 'resolved' |
Migrated from rt.perl.org#119125 (status was 'resolved')
Searchable as RT119125$
The text was updated successfully, but these errors were encountered: