This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perl5.git
12 years agopatch submisson(symbian/symbian_utils.cpp)
Adam Russell [Wed, 15 Jul 2009 05:27:26 +0000 (01:27 -0400)]
patch submisson(symbian/symbian_utils.cpp)

This patch adds some compatibility for older Symbian
SDKs.

>From 146c431f9030e275fcf9aca35d79a72eece128c2 Mon Sep 17 00:00:00 2001
From: Osvaldo Villalon <ovillalon@dextratech.com>
Date: Fri, 3 Jul 2009 01:26:33 -0400
Subject: [PATCH 1690/1692] Osvaldo Villalon's changes to symbian_utils.cpp.

12 years agoExtUtils::MakeMaker 6.55_01
David Mitchell [Thu, 16 Jul 2009 09:53:04 +0000 (10:53 +0100)]
ExtUtils::MakeMaker 6.55_01

12 years agoAn update to the pod is in order for the PERL_VMS_POSIX_EXIT logical
John E. Malmberg [Thu, 16 Jul 2009 02:45:57 +0000 (21:45 -0500)]
An update to the pod is in order for the PERL_VMS_POSIX_EXIT logical
name.  That should help you with what to put in the delta file.

I forgot to indicate in the file that the DECC$FILENAME_UNIX_REPORT
feature also activates the PERL_VMS_POSIX_EXIT mode.

12 years agoexpand Porting/release_managers_guide.pod
David Mitchell [Wed, 15 Jul 2009 22:59:21 +0000 (23:59 +0100)]
expand Porting/release_managers_guide.pod
and remove some similar or out-of-date stuff from Porting/pumpkin.pod

12 years agoDocument PL_opfreehook
Vincent Pit [Wed, 15 Jul 2009 20:32:36 +0000 (22:32 +0200)]
Document PL_opfreehook

12 years agoThe op_free() hook doesn't need to return an OP*, so introduce a new typedef for...
Vincent Pit [Wed, 15 Jul 2009 20:08:39 +0000 (22:08 +0200)]
The op_free() hook doesn't need to return an OP*, so introduce a new typedef for OP hooks

12 years agodocument given/when in perltrap
Moritz Lenz [Wed, 15 Jul 2009 14:53:37 +0000 (16:53 +0200)]
document given/when in perltrap

12 years agoadd Porting/bump-perl-version to MANIFEST
David Mitchell [Wed, 15 Jul 2009 10:51:23 +0000 (11:51 +0100)]
add Porting/bump-perl-version to MANIFEST

12 years agotest_harness, not just test, needs unpack_files on VMS.
Craig A. Berry [Tue, 14 Jul 2009 23:53:27 +0000 (18:53 -0500)]
test_harness, not just test, needs unpack_files on VMS.

12 years agoadd Porting/bump-perl-version
David Mitchell [Tue, 14 Jul 2009 22:22:37 +0000 (23:22 +0100)]
add Porting/bump-perl-version

utility for bumping the value of the perl version in lots of files

12 years agofix some perl version numbers
David Mitchell [Tue, 14 Jul 2009 18:20:06 +0000 (19:20 +0100)]
fix some perl version numbers

12 years agoUpdate my address.
Abhijit Menon-Sen [Tue, 14 Jul 2009 14:44:43 +0000 (20:14 +0530)]
Update my address.

12 years agoExplain camel and dromedary repository servers
David Golden [Tue, 14 Jul 2009 03:16:35 +0000 (23:16 -0400)]
Explain camel and dromedary repository servers

Patch provided by Dennis Kaarsemaker.

12 years agoDon't preload Carp::Heavy in warnings.pm
Rafael Garcia-Suarez [Tue, 14 Jul 2009 07:41:42 +0000 (09:41 +0200)]
Don't preload Carp::Heavy in warnings.pm

Use only Carp, now that all the machinery is in it.

12 years agoMerge Carp::Heavy into Carp
Rafael Garcia-Suarez [Tue, 14 Jul 2009 07:16:17 +0000 (09:16 +0200)]
Merge Carp::Heavy into Carp

12 years agoupate pumpking list
David Golden [Mon, 13 Jul 2009 13:03:47 +0000 (14:03 +0100)]
upate pumpking list

12 years agoCosmetic change
Stepan Kasal [Mon, 13 Jul 2009 08:26:27 +0000 (10:26 +0200)]
Cosmetic change

12 years agoUpgrade to Encode 2.35
Rafael Garcia-Suarez [Mon, 13 Jul 2009 07:40:04 +0000 (09:40 +0200)]
Upgrade to Encode 2.35

12 years agoClean up t/op/sort.t to constrain tests which use a particular package.
Michael G. Schwern [Sat, 11 Jul 2009 07:34:31 +0000 (00:34 -0700)]
Clean up t/op/sort.t to constrain tests which use a particular package.

Also make use of is() instead of a home rolled conflicting Test.pm style ok().

12 years agoBetter flock detection for autodie tests
Paul Fenwick [Mon, 13 Jul 2009 01:30:28 +0000 (11:30 +1000)]
Better flock detection for autodie tests

G'day p5p,

Craig A. Berry discovered that autodie's tests which attempt to detect
flock() support can cause bogus test failures on older VMS systems:

https://rt.cpan.org/Ticket/Display.html?id=47812

Attached is a trivial patch from Craig that fixes this issue.  It's already
been applied to my upstream repository.

Dave - I have no idea if this would be an issue for people building Perl
5.10.1 on old VMS systems.  You may wish to consider it for inclusion into
maint-5.10 just in case.

All the very best,

Paul

--
Paul Fenwick <pjf@perltraining.com.au> | http://perltraining.com.au/
Director of Training                   | Ph:  +61 3 9354 6001
Perl Training Australia                | Fax: +61 3 9354 2681

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoh2ph: handle "#if defined SYMBOL" better
Stepan Kasal [Thu, 9 Jul 2009 13:49:57 +0000 (15:49 +0200)]
h2ph: handle "#if defined SYMBOL" better

expr() contains a code to handle "defined(SYM)" in #if directives.
Unfortunately, this code is not executed for "defined SYM", without
parentheses.  This patch fixes it.

12 years agoSome missing realclean items for VMS.
Craig A. Berry [Sun, 12 Jul 2009 14:27:48 +0000 (09:27 -0500)]
Some missing realclean items for VMS.

12 years agoWrong skip count in t/op/stash.t
Michael G Schwern [Sun, 12 Jul 2009 09:38:42 +0000 (02:38 -0700)]
Wrong skip count in t/op/stash.t

stash.t has the wrong number of tests to skip when dynamic loading is not
available breaking minitest.

12 years ago[perl #61520] Segfault in debugger with tr// and UTF8
David Mitchell [Sat, 11 Jul 2009 16:56:06 +0000 (17:56 +0100)]
[perl #61520] Segfault in debugger with tr// and UTF8

commit 043e41b8 (29765), which made tr// threadsafe by moving the
swash into the pad, didn't mark the pad SV as read-only, so it was getting
removed from anon sub prototypes

12 years agoSubject: [PATCH] Have add-package.pl include the Changes file
H.Merijn Brand [Sat, 11 Jul 2009 15:54:40 +0000 (17:54 +0200)]
Subject: [PATCH] Have add-package.pl include the Changes file
From: David Golden <dagolden@cpan.org>
Date: Fri, 10 Jul 2009 08:20:06 -0400
Message-Id: <1247228406-29001-1-git-send-email-dagolden@cpan.org>

Also adds a "-c FILE" option for an alternate change log name

I've extended it With additional support for ChangeLog and open for more

12 years agoMaintainers.pl: bump Term::ANSIColor version
David Mitchell [Fri, 10 Jul 2009 15:58:58 +0000 (16:58 +0100)]
Maintainers.pl: bump Term::ANSIColor version

12 years agoExtUtils::MakeMaker 6.54
David Mitchell [Fri, 10 Jul 2009 14:16:14 +0000 (15:16 +0100)]
ExtUtils::MakeMaker 6.54

12 years agoRemove %patchlevel hash
Rafael Garcia-Suarez [Fri, 10 Jul 2009 14:00:37 +0000 (16:00 +0200)]
Remove %patchlevel hash

(as the perforce repository has been decomissionned, this information
is no longer useful and can be retrieved by other means. Every release
listed here has a tag in git)

12 years agoEncode 2.34
David Mitchell [Fri, 10 Jul 2009 13:33:57 +0000 (14:33 +0100)]
Encode 2.34

12 years agoadd lib/ExtUtils/ParseXS/Changes to MANIFEST
H.Merijn Brand [Fri, 10 Jul 2009 10:49:24 +0000 (12:49 +0200)]
add lib/ExtUtils/ParseXS/Changes to MANIFEST

12 years agoFix compiler warning in doio.c
Jerry D. Hedden [Thu, 9 Jul 2009 20:29:23 +0000 (16:29 -0400)]
Fix compiler warning in doio.c

12 years agoAvoid test failure on Linux when STDIN is /dev/null
Gisle Aas [Thu, 9 Jul 2009 17:27:12 +0000 (19:27 +0200)]
Avoid test failure on Linux when STDIN is /dev/null

sysseek($fh, -1, 0) succeed on Linux when $fh is a handle to /dev/null.
In our build environment the perl test suite runs with STDIN redirected
from /dev/null and without a /dev/tty.  The hope is that the same seek
on a regular file will always fail.

12 years agoUpgrade to Term::ANSIColor 2.01
Rafael Garcia-Suarez [Thu, 9 Jul 2009 13:41:34 +0000 (15:41 +0200)]
Upgrade to Term::ANSIColor 2.01

12 years agoMake kill() fatal for non-numeric pids
David Golden [Wed, 8 Jul 2009 17:28:54 +0000 (13:28 -0400)]
Make kill() fatal for non-numeric pids

As the debate over the best way to deal with floating point
pids stalled, this is just for non-numeric, which at least
squashes the bug even if it's not the Platonic ideal for
everyone.

It also doesn't address overloaded objects that might not have
IV, NV or PV appropriately set, but the approach mirrors what is
done elsewhere in doio.c so I recommend applying this patch now and
fixing the problem of overloaded objects at some other time when
it can be done more globally, either through an improvement or
replacement of looks_like_number

Also updated POD for kill when process is 0 or negative and
fixed Test-Harness tests that used kill with a string pid.
(Test-Harness test fix also submitted upstream)

12 years agoMake INSTALL -Duserelocatableinc aware
nperez [Wed, 24 Jun 2009 01:32:31 +0000 (20:32 -0500)]
Make INSTALL -Duserelocatableinc aware

12 years agoMerge commit '1640e9f0f39bf8108370c2f5d99bb100fe6c9418' into blead
Chip Salzenberg [Thu, 9 Jul 2009 07:44:18 +0000 (00:44 -0700)]
Merge commit '1640e9f0f39bf8108370c2f5d99bb100fe6c9418' into blead

12 years agofix segfault in -Dx without threads
Chip Salzenberg [Thu, 9 Jul 2009 07:39:41 +0000 (00:39 -0700)]
fix segfault in -Dx without threads

12 years agoSkip File::Copy return tests on Windows and VMS
Paul Fenwick [Thu, 9 Jul 2009 04:04:35 +0000 (14:04 +1000)]
Skip File::Copy return tests on Windows and VMS

Many thanks to Craig Berry for tracking this down.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoModify the return value of Gv_AMupdate to indicate a compilation error
Rafael Garcia-Suarez [Wed, 8 Jul 2009 22:03:47 +0000 (00:03 +0200)]
Modify the return value of Gv_AMupdate to indicate a compilation error

This way we'll restore most of the performance on object desctruction
lost by the previous commit

12 years agoAdd a parameter "destructing" to Gv_AMupdate()
Rafael Garcia-Suarez [Wed, 8 Jul 2009 14:56:50 +0000 (16:56 +0200)]
Add a parameter "destructing" to Gv_AMupdate()

This boolean parameter indicates if the function has been called
to update the overload magic table while looking up the DESTROY
method. In this case, it's probably best to avoid croaking if
those tables could not be updated (for example due to a method
that could not be loaded.)

12 years agoUpdate ExtUtils::ParseXS to 2.20_01
David Golden [Wed, 8 Jul 2009 17:13:56 +0000 (13:13 -0400)]
Update ExtUtils::ParseXS to 2.20_01

2.20_01 - Wed Jul  8 12:12:47 EDT 2009

 - Fix XSUsage prototypes for testing [Jan Dubois]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoAdd a pluggable hook in op_free()
Vincent Pit [Wed, 8 Jul 2009 14:49:36 +0000 (16:49 +0200)]
Add a pluggable hook in op_free()

12 years agoMention in perldelta that "use 5.011" enables strictures
Rafael Garcia-Suarez [Wed, 8 Jul 2009 09:58:41 +0000 (11:58 +0200)]
Mention in perldelta that "use 5.011" enables strictures

12 years agoIgnore libperl.so.
Florian Ragwitz [Wed, 8 Jul 2009 09:23:26 +0000 (11:23 +0200)]
Ignore libperl.so.

12 years agoFix compiler warning when cuserid is absent and sizeof(int) != sizeof(pointer)
Tim Jenness [Tue, 7 Jul 2009 21:49:11 +0000 (11:49 -1000)]
Fix compiler warning when cuserid is absent and sizeof(int) != sizeof(pointer)

The attached patch fixes a compilation warning from POSIX.xs when cuserid
is missing and ints and pointers have different size. eg on Mac OSX when
using 64-bit mode.

gcc-4.2 -c   -fno-common -DPERL_DARWIN -no-cpp-precomp -arch x86_64 -DDEBUGGING -fno-strict-aliasing -pipe -fstack-protector
-I/usr/local/include -I/opt/local/include -O3 -g   -DVERSION=\"1.17\"
-DXS_VERSION=\"1.17\"  "-I../.."   POSIX.c
POSIX.c: In function 'XS_POSIX_cuserid':
POSIX.c:4096: warning: cast to pointer from integer of different size
Running Mkbootstrap for POSIX ()
chmod 644 POSIX.bs
rm -f ../../lib/auto/POSIX/POSIX.bundle

Also relevant for 5.10.1.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoUpdating Module::Build to 0.34
David Golden [Wed, 8 Jul 2009 02:15:59 +0000 (22:15 -0400)]
Updating Module::Build to 0.34

0.34 - Tue Jul  7 16:56:47 EDT 2009

 No changes from 0.33_06

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoMake File::Copy::rmscopy return the documented values.
Craig A. Berry [Wed, 8 Jul 2009 03:29:10 +0000 (22:29 -0500)]
Make File::Copy::rmscopy return the documented values.

These were always supposed to be 0 for failure and 1 for success,
not booleans.  On VMS, rmscopy implements syscopy.

12 years agoMerge branch 'strict-by-default' into blead
Rafael Garcia-Suarez [Tue, 7 Jul 2009 20:32:39 +0000 (22:32 +0200)]
Merge branch 'strict-by-default' into blead

12 years agosome fixes to makerel:
David Mitchell [Tue, 7 Jul 2009 17:48:36 +0000 (18:48 +0100)]
some fixes to makerel:
 * add cmd-line switches
 * make ExtUtils::Manifest silent
 * remove some autogenerated files from the fixup list

12 years agoupdate to CPANPLUS 0.88
David Mitchell [Tue, 7 Jul 2009 16:06:43 +0000 (17:06 +0100)]
update to CPANPLUS 0.88

12 years agorebump Hash::Util::FieldHash from 1.03_01 to 1.04
David Mitchell [Tue, 7 Jul 2009 14:40:58 +0000 (15:40 +0100)]
rebump Hash::Util::FieldHash from 1.03_01 to 1.04
(it's not dual-lived so there's no need for a devel version number)

12 years agoFast enabling of strictures when version 5.11.0 is required
Rafael Garcia-Suarez [Tue, 7 Jul 2009 09:04:13 +0000 (11:04 +0200)]
Fast enabling of strictures when version 5.11.0 is required

We don't load strict.pm, we just manipulate the hint bits.
Plus more tests.

12 years agouse strict by default if "use 5.011" is in effect!
Steffen Mueller [Tue, 7 Jul 2009 08:27:50 +0000 (10:27 +0200)]
use strict by default if "use 5.011" is in effect!

12 years agolimit what add-package.pl might try to delete
David Golden [Tue, 7 Jul 2009 08:10:33 +0000 (04:10 -0400)]
limit what add-package.pl might try to delete

Old way looked in perl source for files matching exclude
list to delete.  New way looks only in source distribution.

This means that files already in perl source will need to be
manually removed, but avoids the risk that a bad exclusion
regex blows away unrelated things in the source tree.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoupgrade to ExtUtils::MakeMaker 6.53_03
David Mitchell [Mon, 6 Jul 2009 21:16:57 +0000 (22:16 +0100)]
upgrade to ExtUtils::MakeMaker 6.53_03

12 years agoRegex fails when string is too long
hv@crypt.org [Mon, 6 Jul 2009 14:45:12 +0000 (15:45 +0100)]
Regex fails when string is too long

This looks to be a simple oversight. All tests pass here.

Hugo

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoDon't dump autodie from core (was Re: Coring Variable::Magic / autodie fights with...
Paul Fenwick [Mon, 6 Jul 2009 14:17:12 +0000 (00:17 +1000)]
Don't dump autodie from core (was Re: Coring Variable::Magic / autodie fights with string eval in Perl 5.10.x)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

G'day Dave,

Dave Mitchell wrote:

> Okay I think you've convinced me. So basically you'll release a 2.06 which
> is 2.05 + bug documentation; 2.06 goes in 5.10.1; then the bug fix goes in
> 2.07?

Patch attached, including documentation fixes, and TODO tests.

I did slip a small code change in, which fixes a regression that was
introduced (but not spotted) around autodie 1.997ish.  Many thanks to rafl
and particularly to Vincent for their help in debugging and documenting.

All the best,

Paul

- --
Paul Fenwick <pjf@perltraining.com.au> | http://perltraining.com.au/
Director of Training                   | Ph:  +61 3 9354 6001
Perl Training Australia                | Fax: +61 3 9354 2681
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (MingW32)

iD8DBQFKUgdnx5N6j7FHnlURAhcbAJ4zQ5WujSXK5uNpEfdOp0RMi8FxVQCghU02
Z3ZBYlXddRMOmm9SVC7UPTE=
=qf29
-----END PGP SIGNATURE-----

>From 689bc09292ba5173542d208dca85b6f514c91aba Mon Sep 17 00:00:00 2001
From: Paul Fenwick <pjf@perltraining.com.au>
Date: Tue, 7 Jul 2009 00:11:42 +1000
Subject: [PATCH] Merge autodie 2.06 with core.

NOTE: This commit is different from that on the CPAN by three lines
of POD.  A documented bug was fixed, but documentation about this bug
was accidentally not removed from the CPAN release.

Significant changes in this release are:

        * BUG: Explicitly documented that autodie does NOT play
          nicely with string evals, especially under Perl 5.10.x.
          Please avoid using string evals while autodie is in scope.

        * TEST: Check for autodie leaking out of scope in the
          presence of string evals.  (string-eval-leak.t)
          Thanks to Florian Ragwitz and Vincent Pit for identifying
          this.

        * BUGFIX: autodie once again correctly works when used
          inside a string eval.  (This was accidently broken
          somewhere around 1.997-1.998).

12 years agore-add Test::Simple exclude accidentally removed during 0.92 upgrade
David Mitchell [Mon, 6 Jul 2009 14:47:43 +0000 (15:47 +0100)]
re-add Test::Simple exclude accidentally removed during 0.92 upgrade

12 years agoclarify usage of Porting/Maintainers
David Mitchell [Mon, 6 Jul 2009 14:43:17 +0000 (15:43 +0100)]
clarify usage of Porting/Maintainers

12 years agoadd core boilerplate to three new Test::Simple tests
David Mitchell [Mon, 6 Jul 2009 14:39:34 +0000 (15:39 +0100)]
add core boilerplate to three new Test::Simple tests

12 years agoupdate to Test::Simple 0.92
David Mitchell [Mon, 6 Jul 2009 13:59:06 +0000 (14:59 +0100)]
update to Test::Simple 0.92

12 years agoMaintainers.pl: Explanation as to why autodie tests are excluded.
Paul Fenwick [Mon, 6 Jul 2009 10:27:49 +0000 (20:27 +1000)]
Maintainers.pl: Explanation as to why autodie tests are excluded.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agobump autodie veresion in Maintainers.pl
David Mitchell [Mon, 6 Jul 2009 10:33:33 +0000 (11:33 +0100)]
bump autodie veresion in Maintainers.pl

12 years agosync Module::Build with CPAN
David Mitchell [Mon, 6 Jul 2009 09:43:53 +0000 (10:43 +0100)]
sync Module::Build with CPAN

12 years agosync ExtUtils::CBuilder with CPAM
David Mitchell [Mon, 6 Jul 2009 09:41:10 +0000 (10:41 +0100)]
sync ExtUtils::CBuilder with CPAM

12 years agoAdded missing 2009 to perl.h copyright header
Offer Kaye [Mon, 6 Jul 2009 09:02:52 +0000 (12:02 +0300)]
Added missing 2009 to perl.h copyright header

12 years agoUpdate Module::Build to 0.33_06
David Golden [Mon, 6 Jul 2009 03:37:32 +0000 (23:37 -0400)]
Update Module::Build to 0.33_06

0.33_06 - Sun Jul  5 10:11:40 EDT 2009

 Bug-fixes:
 - Bundled version code will use pure Perl on 5.10.0 to work around
   a corner case involving eval and locale [John Peacock]
 - Reversed VMS patch from 0.33_03 [Craig Berry]
 - PL_files in Build.PL that are in the bin/scripts directory should not be
   installed as if they are scripts (fixed for case-tolerant systems).
   [David Golden, reported by Craig Berry]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoUpdated ExtUtils::CBuilder to 0.2602
David Golden [Sat, 4 Jul 2009 15:23:22 +0000 (11:23 -0400)]
Updated ExtUtils::CBuilder to 0.2602

0.2602 - Sat Jul  4 10:57:12 EDT 2009

 Bugs fixed:
 - Fixed 00-have-compiler.t if $^X is a relative path [David Wheeler]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoUpdate CPANPLUS::Dist::Build to CPAN version 0.36
Chris 'BinGOs' Williams [Sat, 4 Jul 2009 14:06:19 +0000 (15:06 +0100)]
Update CPANPLUS::Dist::Build to CPAN version 0.36

Hi,

Attached is a patch to update CPANPLUS::Dist::Build in core to
CPAN version 0.36.

Highlights from the Change Log:

==============================================================
0.36 Sat Jul  4 14:51:42 BST 2009
  - Make install() honour the buildflags directive,
    absence spotted by Matt Trout
==============================================================

Many thanks,

--
Chris Williams
aka BinGOs
PGP ID 0x4658671F
http://www.gumbynet.org.uk
==========================

From 43b0b0b2f93858517fa5d21566bcd807d4bf434a Mon Sep 17 00:00:00 2001
From: Chris Williams <chris@bingosnet.co.uk>
Date: Sat, 4 Jul 2009 15:02:09 +0100
Subject: [PATCH] Update CPANPLUS::Dist::Build to CPAN version 0.36

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoMerge autodie 2.05 into core.
Paul Fenwick [Sat, 4 Jul 2009 06:41:39 +0000 (16:41 +1000)]
Merge autodie 2.05 into core.

2.05  Sat Jul  4 16:33:01 AUSEST 2009

        * BUGFIX: format_default() in autodie::exception no longer
          returns a string with file and line attached.  This would
          cause the file and line information to appear twice when
          format handlers would choose to fall back to the defaults.
          The file and line information is now always added by
          stringify().  (RT #47520, thanks to Michael Schwern)

        * BUGFIX: Exceptions thrown by 2-argument open() are more likely
          to specify the mode as 'for reading' when no explicit
          mode was given.  (RT #47520, thanks to Michael Schwern)

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoThe 2nd arg to save_alloc() must be cast to I32 to avoid pointer truncation warnings...
Jan Dubois [Sat, 4 Jul 2009 00:56:11 +0000 (17:56 -0700)]
The 2nd arg to save_alloc() must be cast to I32 to avoid pointer truncation warnings on 64-bit platforms.

12 years agobump versions of non-dual-life modules that
David Mitchell [Fri, 3 Jul 2009 21:53:42 +0000 (22:53 +0100)]
bump versions of non-dual-life modules that
* differ between 5.10.0 and maint-5.10, or
* differ between 5.8.9 and maint-5.10

12 years agomake cmpVERSION.pl ignore Config.pm
David Mitchell [Fri, 3 Jul 2009 21:09:59 +0000 (22:09 +0100)]
make cmpVERSION.pl ignore Config.pm

12 years agoadd -x option to Porting/cmpVERSION.pl
David Mitchell [Fri, 3 Jul 2009 20:31:15 +0000 (21:31 +0100)]
add -x option to Porting/cmpVERSION.pl
which ignores dual-life modules

12 years agomark I18N::LangTags as no longer dual-life
David Mitchell [Fri, 3 Jul 2009 16:15:48 +0000 (17:15 +0100)]
mark I18N::LangTags as no longer dual-life

12 years agosync Module::Build with CPAN
David Mitchell [Fri, 3 Jul 2009 16:14:22 +0000 (17:14 +0100)]
sync Module::Build with CPAN

12 years agosync ExtUtils::CBuilder with CPAN
David Mitchell [Fri, 3 Jul 2009 16:13:39 +0000 (17:13 +0100)]
sync ExtUtils::CBuilder with CPAN

12 years agosync CPAN with, er..., CPAN
David Mitchell [Fri, 3 Jul 2009 16:13:14 +0000 (17:13 +0100)]
sync CPAN with, er..., CPAN

12 years agoupate Maintainers.pl versions
David Mitchell [Fri, 3 Jul 2009 16:12:40 +0000 (17:12 +0100)]
upate Maintainers.pl versions

12 years agoRevert "Bump Module-Pluggable VERSION following
David Mitchell [Fri, 3 Jul 2009 15:03:45 +0000 (16:03 +0100)]
Revert "Bump Module-Pluggable VERSION following
4a28828fc8f160c18323be1125f0f8473bcd000f"

(There is no need to bump the version number for just a core-specific
Makefile.PL hack)

This reverts commit 53102b2bf304e114abfc50b0b9efcd8bbcab5c41.

12 years agobump memoize version to distinguish it from 5.8.9
David Mitchell [Fri, 3 Jul 2009 14:58:05 +0000 (15:58 +0100)]
bump memoize version to distinguish it from 5.8.9

12 years agoOsvaldo Villalon's changes to sdk.pl.
Adam Russell [Fri, 3 Jul 2009 05:33:30 +0000 (01:33 -0400)]
Osvaldo Villalon's changes to sdk.pl.

This patch cleans up symbian/sdk.pl.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoOsvaldo Villalon's changes to symbian_dll.cpp.=0A=
Adam Russell [Fri, 3 Jul 2009 05:35:46 +0000 (01:35 -0400)]
Osvaldo Villalon's changes to symbian_dll.cpp.=0A=

This patch adds some backwards compatibility for older
SDKs.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoChanges made by Osvaldo Villalon.
Adam Russell [Fri, 3 Jul 2009 05:32:00 +0000 (01:32 -0400)]
Changes made by Osvaldo Villalon.

This patch cleans up config.pl

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoSome bugs in Perl regexp (core Perl issues)
hv@crypt.org [Thu, 2 Jul 2009 10:36:08 +0000 (11:36 +0100)]
Some bugs in Perl regexp (core Perl issues)

"Hugo van der Sanden via RT" <perlbug-followup@perl.org> wrote:
:This is caused by a failure of the start_class optimization in the case
:of lookahead, as per the attached comment.
:
:In more detail: at the point study_chunk() attempts to deal with the
:start_class discovered for the lookahead chunk, we have
:SCF_DO_STCLASS_OR set, and_withp has the starting value of ANYOF_EOS |
:ANYOF_UNICODE_ALL, and data->start_class has [a] | ANYOF_EOS.
[...]
:In other words, we need to stack an alternation of ANDs and ORs to cope
:with this situation, and we don't have a mechanism to do that except to
:recurse into study_chunk() some more.
:
:A simpler short-term fix is instead to throw up our hands in this
:situation, and just nullify start_class. I'm not sure exactly how to do
:that, but it seems the more likely to be achievable for 5.10.1.

This patch implements the simple fix, and passes all tests including
Abigail's test cases for the bug.

Yves: note that I've preserved the 'was' code in this chunk, introduced
by you in the patch [1], discussed in the thread [2]. As far as I can
see the 3 lines propagating ANYOF_EOS via 'was' (and the copy of those
3 lines a little later) are simply doing the wrong thing - they seem
to be saying "when we combine two start classes using SCF_DO_STCLASS_AND,
claim that end-of-string is valid if the first class says it would be
even though the second says it wouldn't be". Removing those lines doesn't
cause any test failures - can you remember why you introduced those lines,
and maybe add a test case that fails without them?

Hugo

[1] http://perl5.git.perl.org/perl.git/commit/b515a41db88584b4fd1c30cf890c92d3f9697760
[2] http://groups.google.co.uk/group/perl.perl5.porters/browse_thread/thread/436187077ef96918/f11c3268394abf89

Message-Id: <200907021036.n62Aa8rv029500@zen.crypt.org>
rt.perl.org #56690

12 years agoRemove fake facility strings from VMS start-up warnings.
Craig A. Berry [Thu, 2 Jul 2009 17:32:18 +0000 (12:32 -0500)]
Remove fake facility strings from VMS start-up warnings.

They weren't consistent and the second one caused a compiler
warning.  Plus there's no particular reason to pretend we're
using the native message API when we're not.

12 years agoUnused 'cv'
Jerry D. Hedden [Mon, 29 Jun 2009 19:13:18 +0000 (15:13 -0400)]
Unused 'cv'

12 years agoUpgrade autodie to 2.04
Paul Fenwick [Thu, 2 Jul 2009 13:55:08 +0000 (23:55 +1000)]
Upgrade autodie to 2.04

G'day p5p,

Sorry about the slew of autodie patches.  This is hopefully the last one
you'll see for a while, and merely fixes some nits in the test suite.
Generated against the latest blead, but there's a chance some of these
changes may already be applied.  Apologies for any inconvenience caused.

All the best,

Paul

--
Paul Fenwick <pjf@perltraining.com.au> | http://perltraining.com.au/
Director of Training                   | Ph:  +61 3 9354 6001
Perl Training Australia                | Fax: +61 3 9354 2681

From 2ac05f822da3f96c6453dd32570bc76e4a8e1489 Mon Sep 17 00:00:00 2001
From: Paul Fenwick <pjf@perltraining.com.au>
Date: Thu, 2 Jul 2009 23:46:53 +1000
Subject: [PATCH] Upgrade autodie to 2.04

* TEST: Removed spurious warning about insufficient credit.

* TEST: hints.t produces less debugging output when testing
  the Perl core.  (Thanks to Jerry D. Hedden)

* TEST: hints.t no longer spuriously fails when checking
  the return values from File::Copy under Windows before
  Perl 5.10.1.  (Thanks to Curtis Jewell)

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoAdded docs from Wolfgang Laun to perlpacktut about Intel HEX
H.Merijn Brand [Thu, 2 Jul 2009 10:27:54 +0000 (12:27 +0200)]
Added docs from Wolfgang Laun to perlpacktut about Intel HEX

12 years agomake manisort.
Nicholas Clark [Wed, 1 Jul 2009 21:40:36 +0000 (22:40 +0100)]
make manisort.

12 years agoUpdating ExtUtils-ParseXS to 2.20
David Golden [Wed, 1 Jul 2009 17:54:43 +0000 (13:54 -0400)]
Updating ExtUtils-ParseXS to 2.20

2.20 - Wed Jul  1 13:42:11 EDT 2009

 - No changes from 2.19_04

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoTabs, not spaces, in MANIFEST.
Nicholas Clark [Wed, 1 Jul 2009 15:33:34 +0000 (16:33 +0100)]
Tabs, not spaces, in MANIFEST.

12 years ago[Configure]: avoid leaking Win32CORE.c into output of extension probing.
Andy Dougherty [Wed, 1 Jul 2009 14:28:47 +0000 (10:28 -0400)]
[Configure]: avoid leaking Win32CORE.c into output of extension probing.

12 years agoautodie 2.03
Paul Fenwick [Wed, 1 Jul 2009 13:34:21 +0000 (23:34 +1000)]
autodie 2.03

G'day Everyone,

Attached is a patch that brings blead up to autodie 2.03.  The failing tests
are fixed, and a number of nits have been picked.  Full changes below, as
well as in the commit message.

Cheerio,

Paul

Revision history for autodie

        * TEST: Removed spurious warning about insufficient credit.

2.03  Wed Jul  1 15:39:16 AUSEST 2009

        * BUGFIX: Stopped blog_hints.t from booching under Perl
          5.8.x. because parent.pm is not installed.

2.02  Wed Jul  1 15:06:21 AUSEST 2009

        * FEATURE: autodie::exception now supports ->context() to
          discover the context of the failing subroutine, and
          ->return() to get a list of what it returned.

        * BUGFIX: ->function from autodie::exception now returns
          the original name of the dying sub, rather than its imported
          name.  For example, 'File::Copy::copy' rather than 'main::copy'.
          Core functions continue to always return 'CORE::whatever'.

        * TEST: blog_hints.t tests new hinting features against
          examples in my blog at http://pjf.id.au/blog/

2.01  Wed Jul  1 01:31:24 AUSEST 2009

        * DOCUMENTATION: General copyediting and tidy-up
          (Thanks to Toby Corkindale)

        * BUGFIX: Warnings are no longer emitted when undefined values
          are compared by hinting routines.

        * BUGFIX: Hints for File::Copy now operate correctly under
          Perl 5.10.1.

        * BUGFIX: Inheritance is now considered sufficient to declare
          allegiance to the hints provider role under Perl 5.8.x.
          (Thanks to Glenn Fowler)

        * TEST: hints.t no longer throws failures under Perl 5.10.1.

        * TEST: pod-coverage.t (author test) no longer fails if
          Sub::Identify is not installed.
          (Thanks to Jonathan Yu. RT #47437)

--
Paul Fenwick <pjf@perltraining.com.au> | http://perltraining.com.au/
Director of Training                   | Ph:  +61 3 9354 6001
Perl Training Australia                | Fax: +61 3 9354 2681

From 715681c17c0f3f9f7a09971007c4c4de0d122c60 Mon Sep 17 00:00:00 2001
From: Paul Fenwick <pjf@perltraining.com.au>
Date: Wed, 1 Jul 2009 23:30:58 +1000
Subject: [PATCH] Merge autodie 2.03 into core.

Revision history for autodie

        * TEST: Removed spurious warning about insufficient credit.

2.03  Wed Jul  1 15:39:16 AUSEST 2009

        * BUGFIX: Stopped blog_hints.t from booching under Perl
          5.8.x. because parent.pm is not installed.

2.02  Wed Jul  1 15:06:21 AUSEST 2009

        * FEATURE: autodie::exception now supports ->context() to
          discover the context of the failing subroutine, and
          ->return() to get a list of what it returned.

        * BUGFIX: ->function from autodie::exception now returns
          the original name of the dying sub, rather than its imported
          name.  For example, 'File::Copy::copy' rather than 'main::copy'.
          Core functions continue to always return 'CORE::whatever'.

        * TEST: blog_hints.t tests new hinting features against
          examples in my blog at http://pjf.id.au/blog/

2.01  Wed Jul  1 01:31:24 AUSEST 2009

        * DOCUMENTATION: General copyediting and tidy-up
          (Thanks to Toby Corkindale)

        * BUGFIX: Warnings are no longer emitted when undefined values
          are compared by hinting routines.

        * BUGFIX: Hints for File::Copy now operate correctly under
          Perl 5.10.1.

        * BUGFIX: Inheritance is now considered sufficient to declare
          allegiance to the hints provider role under Perl 5.8.x.
          (Thanks to Glenn Fowler)

        * TEST: hints.t no longer throws failures under Perl 5.10.1.

        * TEST: pod-coverage.t (author test) no longer fails if
          Sub::Identify is not installed.
          (Thanks to Jonathan Yu. RT #47437)

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoUpdate ExtUtils::CBuilder to 0.2601
David Golden [Wed, 1 Jul 2009 13:51:08 +0000 (09:51 -0400)]
Update ExtUtils::CBuilder to 0.2601

0.2601 - Wed Jul  1 09:37:39 EDT 2009

 Bugs fixed:
 - On VMS, cleans up extra files generated during testing
   [John E. Malmberg, Craig Berry]

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agopatch submission(symbian/PerlBase.h)
Adam Russell [Wed, 1 Jul 2009 04:00:46 +0000 (00:00 -0400)]
patch submission(symbian/PerlBase.h)

The change made to PerlBase.cpp in a previous patch, of course, requires
a change in PerlBase.h!
Forgot to send this in earlier...
Best Regards,
Adam

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoMake C<undef ~~ 0> and C<undef ~~ ""> not match (like in 5.10.0)
Rafael Garcia-Suarez [Wed, 1 Jul 2009 10:28:04 +0000 (12:28 +0200)]
Make C<undef ~~ 0> and C<undef ~~ ""> not match (like in 5.10.0)

This makes ~~ commutative with regard to undef, and fixes an
inconsistency, since C<undef ~~ [0]> was not matching, and ~~
should be distributive in this case.

12 years agopatch submission(symbian/config.pl)
Adam Russell [Tue, 30 Jun 2009 19:33:35 +0000 (15:33 -0400)]
patch submission(symbian/config.pl)

Another symbian patch. All the credit for this one
really should go to Osvaldo Villalon(ovillalon@dextratech.com).
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.6001.18063" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>Another symbian patch. All the credit for this one
</FONT></DIV>
<DIV><FONT face=Arial size=2>really should go to Osvaldo Villalon(<A
href="mailto:ovillalon@dextratech.com">ovillalon@dextratech.com</A>).
</FONT></DIV></BODY></HTML>

From 8bfdfd3cd4ef775ae7f30bcade9f0f138ce8d8ac Mon Sep 17 00:00:00 2001
From: Adam Russell <arussell@cs.uml.edu>
Date: Tue, 30 Jun 2009 15:24:39 -0400
Subject: [PATCH 1685/1685] Changes made by Osvaldo Villalon update for latest SDKs.

Signed-off-by: H.Merijn Brand <h.m.brand@xs4all.nl>
12 years agoUpdate CPANPLUS::Dist::Build to CPAN version 0.35_03
Chris Williams [Tue, 30 Jun 2009 22:15:18 +0000 (23:15 +0100)]
Update CPANPLUS::Dist::Build to CPAN version 0.35_03

12 years agoQuiet t/op/stash.t output
Jerry D. Hedden [Tue, 30 Jun 2009 19:16:33 +0000 (15:16 -0400)]
Quiet t/op/stash.t output

 Attached patch to t/op/stash.t applies Craig Berry's fix (commit
 fbad106ea7f26c14f996d4ff2eb920ccffc44821) only on VMS. This keeps the
 noise from the affected tests from polluting the 'make test' output.