Hey everyone,

Following is the p5p (Perl 5 Porters) mailing list summary for the past week. Enjoy!

February 2nd-7th

Edit: The last paragraph in the report mentioned a wrong ticket number, which led to a 404. This was spotted by Tony Cook, Ben Bullock, and Choroba. Thank you all for noticing and for the correction!

News and updates

Dave Mitchell has merged his work on the context stack. You can read more here. He also resolved two outstanding tickets as part of the work:

• Perl #124248
• Perl #124156

Thanks to Craig A. Berry, podlators 4.06 is now officially in blead.

Issues

New issues

• Perl #127455: Regex qr/(?[ [B] | ! ( [^B] ) ])/ fails to compile on x64 debugging optimized builds.
• Perl #127462: Bogus file descriptor warning in fdopen.
• Perl #127463: Third part of C-style for has messed up scoping.
• Perl #127481: Tracking reminder for do EXPR documentation issue.

Robert Drake alerted the list that perl 5.23.2+ and net-snmp do not play well together.

Resolved issues

• Perl #119667: Smartmatch example in perlop is broken.
• Perl #123994: Null pointer dereferencing.
• Perl #124156: Death during unwinding causes crash.
• Perl #124248: Temporary objects created in returning statements are DESTROYed too late.
• Perl #124387: AUTOLOAD not called when DESTROY is missing in 5.20.1.
• Perl #125471: Document behavior of sprintf with spaces and plus signs.
• Perl #125540: Null pointer dereferencing.
• Perl #126410: Magic for overload broken 5.18..blead.
• Perl #126544: Document variables used in the synopsis of the fcntl function.
• Perl #126621: Bleadperl breaks Tie::DBIx::Class.
• Perl #126815: Segfault in S_scan_heredoc (duplicate).
• Perl #126953: Add renaming capabilities to Exporter. (In favor of Perl #127384.)
• Perl #127474: Perl_cast_uv broken with CASTFLAGS=7, no quad math, uvsize=4.
• Perl #127475: Bleadperl breaks Algorithm::Permute.

Proposed patches

Tony Cook provides a patch for Perl #125833 to disallow leading colons on require.

Tony also provided a patch for Perl #127386 to clarify exit codes in perl a bit more.

Bulk88 provides a patch for Perl #126414 with a conservative approach.

James E. Keenan provided a patch to revise an outdated portion of perlmodlib to resolve Perl #127435.

Discussion

On the matter of a script beginning with a colon, Abigail has found the behavior is indeed document, albeit in a possibly obscure place.

Karl Williamson explains that the problem in Perl #127392 (more memory consumption in constant) is not specific to constant. As he expands it is a Unicode-related loading that was changed from run-time to compile-time.

Karl hopes to tackle this in 5.24.

Tony Cook weighs in on Perl #125569 saying it might not be worth the effort.