https://bz.apache.org/SpamAssassin/show_bug.cgi?id=7992
Giovanni Bechis <giova...@paclan.it> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |giova...@paclan.it
--- Comment #23 from Giovanni Bechis <giova...@paclan.it> ---
(In reply to Henrik Krohns from comment #22)
> ... and reopening.
>
> This is why pre-releases are handy, because it makes people run tests on
> more varied systems.
>
> t/regexp_named_capture does fail atleast on Perl 5.30. But not on my main
> 5.22 dev box.
>
this has changed in Perl 5.26
> invalid regexp for TEST_CAPTURE_6 'm,www\.%{TESTCAP1}\.,i': Unescaped left
> brace in regex is passed through in regex
>
> invalid regexp for TEST_CAPTURE_7 'm{www\.%{HEADER(From:addr:domain)}/}':
> Unescaped left brace in regex is passed through in regex
>
> Classic error. This is why I mentioned my disdain for using any regex meta
> characters. I had false sense of security for it working fine on my single
> Perl install, duh.
>
every regex meta character may need to be escaped in future Perl releases and
this may break rules with future Perl versions
> Now we need to decide if the format needs changing again. Or if it's safe to
> ignore the error. Or whether we should manually add escape the found "%{"
> string into "%\{" and then look for that in the regex template replace code.
--
You are receiving this mail because:
You are the assignee for the bug.
