This is a live mirror of the Perl 5 development currently hosted at
11 years agoRemove all space optimisations from SV body structures.
Nicholas Clark [Fri, 22 Jan 2010 15:31:49 +0000 (15:31 +0000)]
Remove all space optimisations from SV body structures.

11 years agoThe size of a character in C is per definition 1
H.Merijn Brand [Fri, 21 May 2010 06:01:54 +0000 (08:01 +0200)]
The size of a character in C is per definition 1

11 years agoFix CCINCDIR and CCLIBDIR for mingw64 cross compiler
Sisyphus [Fri, 21 May 2010 00:56:07 +0000 (17:56 -0700)]
Fix CCINCDIR and CCLIBDIR for mingw64 cross compiler

When building perl with the mingw64 x64 cross-compiler 'incpath', 'libpth',
'ldflags', 'lddlflags' and 'ldflags_nolargefiles' values in and are not being set correctly because, with that compiler, the
include and lib directories are not immediately below $(CCHOME).

11 years agofully test package-version-block syntax
Zefram [Thu, 20 May 2010 20:48:03 +0000 (21:48 +0100)]
fully test package-version-block syntax

Extend the exhaustive package-version tests in t/op/packagev.t
to test each case using package-block syntax in addition to the
package-declaration syntax.

11 years agoadditional tests for package block syntax
Zefram [Sat, 1 May 2010 16:51:21 +0000 (17:51 +0100)]
additional tests for package block syntax

Test that __PACKAGE__ propagates into string eval correctly.  Test that
__LINE__ is correct.  Test that goto into and out of package blocks
works correctly.

11 years agofix SEGV with eval("package Foo {")
Zefram [Sat, 1 May 2010 16:30:26 +0000 (17:30 +0100)]
fix SEGV with eval("package Foo {")

OPs relating to the package name and version were subject to double
freeing during error recovery from an incomplete package block.  Fixed by
using the op_latefree mechanism to shift the op free time.

11 years agosupport "package Foo { ... }"
Zefram [Mon, 19 Apr 2010 19:37:25 +0000 (20:37 +0100)]
support "package Foo { ... }"

Package block syntax limits the scope of the package declaration to the
attached block.  It's cleaner than requiring the declaration to come
inside the block.

11 years agoPL_endav can be NULL, so in S_ithread_create() no need to set it to newAV().
Nicholas Clark [Tue, 23 Feb 2010 20:35:29 +0000 (20:35 +0000)]
PL_endav can be NULL, so in S_ithread_create() no need to set it to newAV().

11 years agoRemove redundant hv_exists() calls from ithread_create()'s spec parser.
Nicholas Clark [Sat, 13 Feb 2010 09:05:18 +0000 (09:05 +0000)]
Remove redundant hv_exists() calls from ithread_create()'s spec parser.

hv_fetch(..., 0) won't create the element if it doesn't exist, returning a NULL
pointer, so hv_exists() and hv_fetch() is doing two hash lookups where one would

On this machine, reduces the object code by 3K, about 7%. Everyone's a winner.

11 years agoChange S_ithread_create() params from a single AV* to a pair of SV** pointers.
Nicholas Clark [Sat, 13 Feb 2010 08:31:55 +0000 (08:31 +0000)]
Change S_ithread_create() params from a single AV* to a pair of SV** pointers.

This saves creating, duplicating and freeing and AV, which is only ever used for
an internal calling convention.

11 years agoIn threads.xs, convert thread->params from RV to AV.
Nicholas Clark [Sat, 13 Feb 2010 08:01:45 +0000 (08:01 +0000)]
In threads.xs, convert thread->params from RV to AV.

Pass around and store the array directly, rather than creating, holding and
dereferencing a reference to it.

11 years agoSvIVX() isn't valid on SVt_REGEXP
Nicholas Clark [Tue, 26 Jan 2010 12:00:56 +0000 (12:00 +0000)]
SvIVX() isn't valid on SVt_REGEXP

11 years agoIn sv.c, _all_ {new,del}_X* macros can be *_body_allocated.
Nicholas Clark [Mon, 25 Jan 2010 10:53:33 +0000 (10:53 +0000)]
In sv.c, _all_ {new,del}_X* macros can be *_body_allocated.

Previously those where bodies_by_type[sv_type].offset was zero were using
{new,del}_body_typed. However, the optimiser can spot this, and generates the
same object code. This allows simplification of the C code, and more
flexibility to rearrange the structures without generating bugs.

11 years agoupdate Ricardo Signes's email
Ricardo Signes [Thu, 20 May 2010 14:14:39 +0000 (10:14 -0400)]
update Ricardo Signes's email

11 years agoDo not try to load a feature bundle when doing "no VERSION"
Rafael Garcia-Suarez [Wed, 19 May 2010 18:36:12 +0000 (20:36 +0200)]
Do not try to load a feature bundle when doing "no VERSION"

11 years agodo not check for deprecation of removed lib v5.13.1
Ricardo Signes [Thu, 20 May 2010 12:29:40 +0000 (08:29 -0400)]
do not check for deprecation of removed lib

11 years agoadd the v5.13.1 epigraph
Ricardo Signes [Thu, 20 May 2010 11:49:35 +0000 (07:49 -0400)]
add the v5.13.1 epigraph

11 years agoremove =over/=back from epigraphs.pod, add empty sections
Ricardo Signes [Thu, 20 May 2010 11:47:09 +0000 (07:47 -0400)]
remove =over/=back from epigraphs.pod, add empty sections

...even though they are *totally legal* and the suggested way to make a
blockquote-like paragraph.  Pod::Checker doesn't like them.

11 years agoadd new release to perlhist
Ricardo Signes [Thu, 20 May 2010 11:25:47 +0000 (07:25 -0400)]
add new release to perlhist

11 years agoupdate MANIFEST, remove delta for uninstalled file
Ricardo Signes [Thu, 20 May 2010 11:24:16 +0000 (07:24 -0400)]
update MANIFEST, remove delta for uninstalled file

11 years agofurther updates to CoreList
Ricardo Signes [Thu, 20 May 2010 11:18:17 +0000 (07:18 -0400)]
further updates to CoreList

Who knew there was a helper to do this, now!

11 years agoUpdate META.yml
Ricardo Signes [Thu, 20 May 2010 11:06:16 +0000 (07:06 -0400)]
Update META.yml

Notice the glorious deletion of Switch from the manifest!

11 years agobump CoreList version
Ricardo Signes [Thu, 20 May 2010 11:01:43 +0000 (07:01 -0400)]
bump CoreList version

11 years agoadd vertical space to
Ricardo Signes [Thu, 20 May 2010 11:00:42 +0000 (07:00 -0400)]
add vertical space to

This makes it easy to move through the file with { and } in vi.
Yes, this really helps.

11 years agoremove empty perldelta section
Ricardo Signes [Thu, 20 May 2010 04:50:55 +0000 (00:50 -0400)]
remove empty perldelta section

11 years agoupdate change command
Ricardo Signes [Thu, 20 May 2010 04:50:44 +0000 (00:50 -0400)]
update change command

11 years agoBump the perl version in various places for 5.13.1
Ricardo Signes [Thu, 20 May 2010 04:45:27 +0000 (00:45 -0400)]
Bump the perl version in various places for 5.13.1

11 years agoreplace 2nd 5.013000 block with 5.013001
Ricardo Signes [Thu, 20 May 2010 04:31:11 +0000 (00:31 -0400)]
replace 2nd 5.013000 block with 5.013001

11 years agoupdate CoreList for 5.13.1
Ricardo Signes [Thu, 20 May 2010 04:29:02 +0000 (00:29 -0400)]
update CoreList for 5.13.1

11 years agobump version on modules changed since 5.13.0
Ricardo Signes [Thu, 20 May 2010 04:28:42 +0000 (00:28 -0400)]
bump version on modules changed since 5.13.0

11 years agopass through perl5131delta in prep of release
Ricardo Signes [Thu, 20 May 2010 04:12:47 +0000 (00:12 -0400)]
pass through perl5131delta in prep of release

11 years agoAdd a note in perl5131delta about given return values
Vincent Pit [Thu, 20 May 2010 00:44:22 +0000 (02:44 +0200)]
Add a note in perl5131delta about given return values

And tweak its documentation.

11 years agoAdding script used for recent perlhist changes.
Curtis Jewell [Mon, 17 May 2010 07:55:28 +0000 (01:55 -0600)]
Adding script used for recent perlhist changes.

11 years agoUpdate perlhist.pod for 5.10.1-5.12.1
Curtis Jewell [Mon, 17 May 2010 07:50:19 +0000 (01:50 -0600)]
Update perlhist.pod for 5.10.1-5.12.1

11 years agobumped File::Find version to 1.16
David Golden [Wed, 19 May 2010 21:04:35 +0000 (17:04 -0400)]
bumped File::Find version to 1.16

11 years agoFixed typo: "effecting data" -> "affecting data".
Dan Dascalescu [Tue, 18 May 2010 22:12:19 +0000 (15:12 -0700)]
Fixed typo: "effecting data" -> "affecting data".

11 years agoMerge branch 'vincent/rvalue_stmt_given' into blead
Vincent Pit [Wed, 19 May 2010 20:59:58 +0000 (22:59 +0200)]
Merge branch 'vincent/rvalue_stmt_given' into blead

11 years agoperldelta entries for my changes in 5.13.1.
Nicholas Clark [Wed, 19 May 2010 19:02:59 +0000 (20:02 +0100)]
perldelta entries for my changes in 5.13.1.

11 years agoadd reminder about updating epigraphs.pod to RMG
David Golden [Wed, 19 May 2010 16:30:31 +0000 (12:30 -0400)]
add reminder about updating epigraphs.pod to RMG

11 years agothe proper term is epigraph, not epigram
David Golden [Wed, 19 May 2010 16:27:52 +0000 (12:27 -0400)]
the proper term is epigraph, not epigram

11 years agodocument historical release epigrams
David Golden [Wed, 19 May 2010 16:20:40 +0000 (12:20 -0400)]
document historical release epigrams

11 years agoActually note that is deprecated for 5.13 and 5.14, so we can
Jesse Vincent [Tue, 18 May 2010 16:17:24 +0000 (12:17 -0400)]
Actually note that is deprecated for 5.13 and 5.14, so we can
remove it in 5.15 in the Spring of 2011.

11 years was missing its deprecation warning in 5.12. So it can't be
Jesse Vincent [Tue, 18 May 2010 16:15:41 +0000 (12:15 -0400)] was missing its deprecation warning in 5.12. So it can't be
removed in 5.14.

Revert "Remove Shell from the core distribution. Get it from CPAN now."

This reverts commit 28d302d426b73ed76fdcc816dd51bb1a8f93332b.

11 years agoMerge branch 'blead' of ssh:// into blead
Jesse Vincent [Tue, 18 May 2010 15:55:15 +0000 (11:55 -0400)]
Merge branch 'blead' of ssh:// into blead

* 'blead' of ssh://
  Fix groups.t test on AIX
  further fix for RT #23810
  use the correct format codes in warnings from gmtime/localtime
  Better perldiag heading for "\cX" deprecation warning
  PATCH: Clean up EBCDIC handling of \cX
  [perl #41530] s/non-utf8/is-utf8/ fails.
  Fix for README.aix and XL C/C++ V11

11 years agoAdding Chris Williams as our November 2010 sucker
Jesse Vincent [Tue, 18 May 2010 15:54:01 +0000 (11:54 -0400)]
Adding Chris Williams as our November 2010 sucker

11 years agoFix groups.t test on AIX
Rainer Tammer [Sun, 16 May 2010 09:37:57 +0000 (11:37 +0200)]
Fix groups.t test on AIX

'id -a' doesn't work, so try it without the -a if that fails.

11 years agofurther fix for RT #23810
David Mitchell [Mon, 17 May 2010 23:54:50 +0000 (00:54 +0100)]
further fix for RT #23810

The fix for #23810, 27e904532594b7fb, introduced a regression, spotted by
Nicholas as RT #75146.

Basically, in S_doeval() if the yyparse() fails due to dying (rather than
just bailing out with with a syntax error, say), then the topmost EVAL
context will have been popped. My improved error handling code mostly
understood the difference, but forgot that in the die case, PL_eval_root
will have been restored to its previous value by the CX pop, and thus
its value shouldn't be messed with.

11 years agouse the correct format codes in warnings from gmtime/localtime
Tony Cook [Tue, 23 Feb 2010 13:35:35 +0000 (00:35 +1100)]
use the correct format codes in warnings from gmtime/localtime

11 years agoBetter perldiag heading for "\cX" deprecation warning
Rafael Garcia-Suarez [Mon, 17 May 2010 09:47:48 +0000 (11:47 +0200)]
Better perldiag heading for "\cX" deprecation warning

11 years agoPATCH: Clean up EBCDIC handling of \cX
Karl Williamson [Wed, 12 May 2010 17:50:19 +0000 (11:50 -0600)]
PATCH: Clean up EBCDIC handling of \cX

The function perl_ebcdic_control() is unnecessary, as the toCTRL macro
that calls it can be changed to just map EBCDIC to ASCII first, and then
doing the normal procedure.

This means that EBCDIC and ASCII will no longer diverge.  Currently,
EBCIDIC gives a syntax error for inputs outside its domain, whereas the
ASCII version accepts some of them.

11 years ago[perl #41530] s/non-utf8/is-utf8/ fails.
Karl Williamson [Tue, 11 May 2010 16:57:41 +0000 (10:57 -0600)]
[perl #41530] s/non-utf8/is-utf8/ fails.

When the replacement is in utf8, there was failure to upgrade the result
when the source and the pattern weren't in utf8.  This simply checks
that when there is a match that will lead to the replacement being done.
It then does the upgrade.  If this led to changes in the source, we redo
the match because pointers to saved buffers could have changed.  There
may be other cases where we don't need to redo the match, but I don't
know the code well-enough to easily figure it out.

11 years agoFix for README.aix and XL C/C++ V11
Rainer Tammer [Fri, 14 May 2010 07:53:51 +0000 (09:53 +0200)]
Fix for README.aix and XL C/C++ V11

I have attached a small doc fix for README.aix. It contains information
regarding the new XL C/C++ V11 compiler. This patch is only for bleed.
I have already posted the correct fix for maint-5.12.



From f66807e32b4c01bb58b61cb3e48d89da9c2c4bd7 Mon Sep 17 00:00:00 2001
From: Rainer Tammer <>
Date: Fri, 14 May 2010 09:48:25 +0200
Subject: [PATCH] FIX to README.aix for V11 compiler

Signed-off-by: H.Merijn Brand <>
11 years agoPerl 5.12.1 now includes 2.27. Module::CoreList should say so
Jesse Vincent [Thu, 13 May 2010 20:00:26 +0000 (16:00 -0400)]
Perl 5.12.1 now includes 2.27. Module::CoreList should say so

11 years agoNew release date for 5.12.1 in light of the new RC
Jesse Vincent [Thu, 13 May 2010 19:58:51 +0000 (15:58 -0400)]
New release date for 5.12.1 in light of the new RC

11 years agoMerge branch 'blead' of ssh:// into blead
Jesse Vincent [Thu, 13 May 2010 17:04:37 +0000 (13:04 -0400)]
Merge branch 'blead' of ssh:// into blead

* 'blead' of ssh://
  Return 0 (with a warning) for sprintf("%.0g") and sprintf("%.0f")
  Updating list of DOSish perls in perlport.pod.
  Add Porting/ to MANIFEST
  separate the 0755 from the . $TAINT0. this silences the warning "Dot after octal literal is concatenation".
  Clarify that mail to perlbug also goes to p5p

11 years agoCorrecting the release date of 5.12.1
Jesse Vincent [Thu, 13 May 2010 17:04:10 +0000 (13:04 -0400)]
Correcting the release date of 5.12.1

11 years agoReturn 0 (with a warning) for sprintf("%.0g") and sprintf("%.0f")
Nicholas Clark [Thu, 13 May 2010 11:04:55 +0000 (12:04 +0100)]
Return 0 (with a warning) for sprintf("%.0g") and sprintf("%.0f")

There is special case code in the sprintf implementation, for simple %f and %g
formats, conditionally compiled in only when NVs are doubles. Under long
doubles, these are handled by the general purpose code, which always returns
0 if the argument is missing. Note that sprintf(" %.0g"), ie a leading space,
sufficient to bypass the special case code, would return the string " 0".

The special case code used to return an empty string, meaning that the
behaviour of sprintf("%.0g") and sprintf("%.0f") was inconsistent between a
perl built with doubles, and a perl with long doubles, and the behaviour of
sprintf("%.0g") and sprintf(" %.0g") was inconsistent.

5b98cd54dff3b163 fixed #62874 - the special case code did not warn, but
changed behaviour to return 0. d347ad18ecf3da70 undid the behaviour change,
viewing it as a regression. However, the tests added in 5b98cd54dff3b163
expose the inconsistency in behaviour between doubles and long doubles.

There should be no inconsistency, hence the only logically consistent
conclusion is that the special case implementation was wrong - it cannot
give results inconsistent with the general code. Hence this commit changes
it to return 0 (with a warning). This is achieved by simply skipping the
special case code, if there are insufficient arguments.

11 years agoUpdating list of DOSish perls in perlport.pod.
Curtis Jewell [Wed, 12 May 2010 12:55:56 +0000 (06:55 -0600)]
Updating list of DOSish perls in perlport.pod.

11 years agoAdd Porting/ to MANIFEST
H.Merijn Brand [Tue, 11 May 2010 05:54:51 +0000 (07:54 +0200)]
Add Porting/ to MANIFEST

11 years agoseparate the 0755 from the . $TAINT0. this silences the warning "Dot after octal...
Richard Soderberg [Wed, 28 Apr 2010 21:44:47 +0000 (14:44 -0700)]
separate the 0755 from the . $TAINT0. this silences the warning "Dot after octal literal is concatenation".

11 years agoClarify that mail to perlbug also goes to p5p
Ævar Arnfjörð Bjarmason [Sun, 9 May 2010 12:38:29 +0000 (12:38 +0000)]
Clarify that mail to perlbug also goes to p5p

The wording added in 64a8e22bd3 didn't clarify that mail to perlbug is
also forwarded to perl5-porters. Make this section clearer so that it
now explains that perlbug is just perl5-porters + RT tracking.

See the discussion after Craig A. Berry's comment in RT #73914 for

11 years agoPATCH [perl #74978] dot after } breaks \N{}
Karl Williamson [Sat, 8 May 2010 20:06:10 +0000 (14:06 -0600)]
PATCH [perl #74978] dot after } breaks \N{}

The problem is that a dot can come between the braces in \N{},
but when searching for it, I didn't stop looking at the right brace, so
it generated an error inappropriately.

This is essentially a minimum patch; efficiency could be improved
slightly with a little more work.

11 years agoNote in comment that many \N{...} tests won't work here
Karl Williamson [Sat, 8 May 2010 19:38:27 +0000 (13:38 -0600)]
Note in comment that many \N{...} tests won't work here

11 years agoComment where to find file's format
Karl Williamson [Sat, 8 May 2010 19:12:53 +0000 (13:12 -0600)]
Comment where to find file's format

11 years agoperldelta item on PL_restartjmpenv
Zefram [Wed, 5 May 2010 19:41:28 +0000 (20:41 +0100)]
perldelta item on PL_restartjmpenv

11 years agoperldelta item on $[ usage
Zefram [Wed, 5 May 2010 19:35:43 +0000 (20:35 +0100)]
perldelta item on $[ usage

11 years agoperldelta item on B::Deparse package+label fix
Zefram [Wed, 5 May 2010 19:32:33 +0000 (20:32 +0100)]
perldelta item on B::Deparse package+label fix

11 years agoperldelta item on reliable exception handling
Zefram [Wed, 5 May 2010 19:26:01 +0000 (20:26 +0100)]
perldelta item on reliable exception handling

11 years agoperlunifaq, uniintro: fix for 80 col display
Karl Williamson [Wed, 5 May 2010 18:16:48 +0000 (12:16 -0600)]
perlunifaq, uniintro: fix for 80 col display

11 years agoperlunicode: fix for 80 col display
Karl Williamson [Wed, 5 May 2010 18:15:14 +0000 (12:15 -0600)]
perlunicode: fix for 80 col display

11 years agoperlreref: missing info, 80 col display
Karl Williamson [Wed, 5 May 2010 18:13:02 +0000 (12:13 -0600)]
perlreref: missing info, 80 col display

The \p{Posix...} classes had not gotten added yet to the ref pod; there
were some reformattings to make things display properly in an 80 column

11 years agoperlrecharclass: Clarify \p{Punct}, fix for 80 col
Karl Williamson [Wed, 5 May 2010 18:11:02 +0000 (12:11 -0600)]
perlrecharclass: Clarify \p{Punct}, fix for 80 col

While not strictly wrong, the hre was missing info for what \p{Punct}

11 years agoperlrebackslash: fix for 80 col display
Karl Williamson [Wed, 5 May 2010 18:09:47 +0000 (12:09 -0600)]
perlrebackslash: fix for 80 col display

11 years agoperlre: fix for 80 col display
Karl Williamson [Wed, 5 May 2010 18:08:53 +0000 (12:08 -0600)]
perlre: fix for 80 col display

11 years agoperlport: fix 80 col display; broken link
Karl Williamson [Wed, 5 May 2010 18:07:42 +0000 (12:07 -0600)]
perlport: fix 80 col display; broken link

11 years agoFix to display ok in 80 columns
Karl Williamson [Wed, 5 May 2010 18:06:18 +0000 (12:06 -0600)]
Fix to display ok in 80 columns

11 years agoFix perlebcdic for for 80 column tty; fix C<>s
Karl Williamson [Wed, 5 May 2010 18:03:16 +0000 (12:03 -0600)]
Fix perlebcdic for for 80 column tty; fix C<>s

Two C<>'s were unclosed

11 years agoAdd =encoding utf8 to all core non-ASCII POD files + test script
Ævar Arnfjörð Bjarmason [Tue, 30 Mar 2010 14:09:24 +0000 (14:09 +0000)]
Add =encoding utf8 to all core non-ASCII POD files + test script

It's now possible to run:

    perl Porting/

To check if the core contains any naughty POD that uses non-ASCII
without declaring an encoding.

With this patch all the POD in core (except POD tests we're ignoring)
has a correct =encoding directive. The script also flags problems in
lib/* and cpan/* which aren't being fixed as part of this commit.

11 years agoRT #34604 didn't honour tied overloaded values
David Mitchell [Sat, 8 May 2010 20:25:47 +0000 (21:25 +0100)]
RT #34604 didn't honour tied overloaded values

A tied hash lookup could return an overloaded object but sort wouldn't
notice that it was overloaded because it checked for overload before doing

11 years agoadd SV_SKIP_OVERLOAD flag to sv_2*v_flags fns
David Mitchell [Sat, 8 May 2010 16:23:56 +0000 (17:23 +0100)]
add SV_SKIP_OVERLOAD flag to sv_2*v_flags fns

While trying to coerce an SV into a string or whatever, stop if you
suddenly discover it's overloaded (this may not happen until after you've
called it's get magic)

11 years agoadd flags arg to sv_2nv (as sv_2nv_flags)
David Mitchell [Sat, 8 May 2010 16:06:45 +0000 (17:06 +0100)]
add flags arg to sv_2nv (as sv_2nv_flags)

11 years agoModify Module::CoreList docs to state the availability of the removed_from() functions.
Chris 'BinGOs' Williams [Sat, 8 May 2010 13:40:24 +0000 (14:40 +0100)]
Modify Module::CoreList docs to state the availability of the removed_from() functions.

11 years agoBackport patch from CPANPLUS-0.9004 to cpanp-run-perl
Chris 'BinGOs' Williams [Sat, 8 May 2010 11:57:00 +0000 (12:57 +0100)]
Backport patch from CPANPLUS-0.9004 to cpanp-run-perl

  This resolves RT #55964 and RT #57106 relating to the failure to
  install distributions that use Module::Install::DSL

  The addition of a BEGIN {} block to cpanp-run-perl resolves the

11 years agoBump Module::Corelist for cherrypicking to maint-5.12
Jesse Vincent [Fri, 7 May 2010 17:59:36 +0000 (13:59 -0400)]
Bump Module::Corelist for cherrypicking to maint-5.12

11 years agoMake t/op/sprintf.t test multiple warnings
Vincent Pit [Thu, 6 May 2010 20:20:10 +0000 (22:20 +0200)]
Make t/op/sprintf.t test multiple warnings

11 years agoStop returning 0 for sprintf("%.0g")
Vincent Pit [Thu, 6 May 2010 20:14:31 +0000 (22:14 +0200)]
Stop returning 0 for sprintf("%.0g")

And also from throwing two "Missing argument" warnings for this construct.

This was a regression introduced by 5b98cd54dff3b16344eab33ce6b09fb6fb1b89c2.

11 years agoReally test special sprintf formats
Vincent Pit [Thu, 6 May 2010 20:35:14 +0000 (22:35 +0200)]
Really test special sprintf formats

11 years agoMake sv_vcatpvfn() complain when special formats "%s" and "%.0f" can't find their...
Vincent Pit [Thu, 6 May 2010 18:28:58 +0000 (20:28 +0200)]
Make sv_vcatpvfn() complain when special formats "%s" and "%.0f" can't find their argument

This fixes [RT #62874] : printf does not print a warning when a lone %s conversion is used

11 years agoChange perlgpl.pod to GPL 1 to match README
David Golden [Thu, 6 May 2010 13:36:59 +0000 (09:36 -0400)]
Change perlgpl.pod to GPL 1 to match README

The README and Copying files are specific that Perl is licensed
under Artistic or GPL 1 (or later).  For whatever reason, perlgpl.pod
was created with GPL 2.  This commit changes it back to GPL 1 for
consistency with the actual license.

It also standardizes the introduction to both perlgpl and perlartistic
for consistent language explaining the dual license.

11 years agoMove code to after the variable declarations.
Nicholas Clark [Thu, 6 May 2010 09:16:02 +0000 (10:16 +0100)]
Move code to after the variable declarations.

Currently we only compile the top level code with "extreme warnings", so code
in ext/ isn't subject to -Wdeclaration-after-statement, hence why this was
missed. Now that dual life modules have been evicted from ext/, we could
probably consider notching up the warnings for all code in ext/.

11 years agoMerge branch 'blead' of ssh:// into blead
Jesse Vincent [Wed, 5 May 2010 23:22:23 +0000 (19:22 -0400)]
Merge branch 'blead' of ssh:// into blead

* 'blead' of ssh://
  bump PerlIO::scalar version
  RT 43789: "in memory" files don't call STORE

11 years agobump PerlIO::scalar version
David Mitchell [Wed, 5 May 2010 21:58:50 +0000 (22:58 +0100)]
bump PerlIO::scalar version

11 years agoRT 43789: "in memory" files don't call STORE
David Mitchell [Wed, 5 May 2010 21:39:24 +0000 (22:39 +0100)]
RT 43789: "in memory" files don't call STORE

The code in PerlIO-scalar that implements the open $fh, '>' \$buffer
feature did not, apart from accidentally, support get/set magic and thus
tied buffers. This patch remedies that: mostly by just blindly sprinkling
SvGETMAGIC/SvSETMAGIC about, rather than doing any deep analysis and
understanding of the code. One main change I did was to add a
PerlIOScalar_read() function, rather than rely on the default behaviour
(which implements it in terms of PerlIOScalar_get_ptr() etc), since that
approach had a tendency to call FETCH multiple times

11 years agoRevert "New deprecation warning: Dot after %s literal is concatenation"
Jesse Vincent [Wed, 5 May 2010 18:40:59 +0000 (14:40 -0400)]
Revert "New deprecation warning: Dot after %s literal is concatenation"

This reverts commit 6fb472bab4fadd0ae2ca9624b74596afab4fb8cb.

Zefram asked me to revert this as he's going to be doing something more

11 years agoRevert "Fix tests and add one more test for the deprecation warning added in last...
Jesse Vincent [Wed, 5 May 2010 18:40:45 +0000 (14:40 -0400)]
Revert "Fix tests and add one more test for the deprecation warning added in last change"

This reverts commit c4deb7365787eb01845a9d0e371e343169530659.

Zefram asked me to revert this as he's going to be doing something more

11 years agoRevert "Deprecation warnings should always be mandatory since 5.12.0"
Jesse Vincent [Wed, 5 May 2010 18:40:26 +0000 (14:40 -0400)]
Revert "Deprecation warnings should always be mandatory since 5.12.0"

This reverts commit a7e260e62a5e47961e908363da32ef16f41301b2.

Zefram asked me to revert this as he's going to be doing something more

11 years agoRevert "tweak "0x123.456" deprecation"
Jesse Vincent [Wed, 5 May 2010 18:35:25 +0000 (14:35 -0400)]
Revert "tweak "0x123.456" deprecation"

This reverts commit 1183a10042af0734ee65e252f15bd820b7bbe686.

Zefram asked me to revert this as he's going to be doing something more

11 years agoWhen saving ints, if the value is small enough save it with the type.
Nicholas Clark [Sat, 20 Feb 2010 19:39:37 +0000 (19:39 +0000)]
When saving ints, if the value is small enough save it with the type.

This uses a new type, SAVEt_INT_SMALL.

11 years agoWhen saving I32s, if the value is small enough save it with the type.
Nicholas Clark [Sat, 20 Feb 2010 19:20:09 +0000 (19:20 +0000)]
When saving I32s, if the value is small enough save it with the type.

This uses a new type, SAVEt_I32_SMALL.

11 years agoIf we're going to introduce an @@ array, we'll want to be able to parse $#@ too
Rafael Garcia-Suarez [Fri, 22 May 2009 14:47:22 +0000 (16:47 +0200)]
If we're going to introduce an @@ array, we'll want to be able to parse $#@ too