This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
perl5.git
10 years agoclarify how $SIG{__DIE__} can return
David Mitchell [Sun, 8 Jan 2012 15:45:39 +0000 (15:45 +0000)]
clarify how $SIG{__DIE__} can return

It can return via 'goto &sub', but not via 'goto LABEL'.
The docs originally just said 'via goto'

See [perl #44367].

10 years agoperlfunc: spaces after dots
Father Chrysostomos [Sat, 7 Jan 2012 16:29:49 +0000 (08:29 -0800)]
perlfunc: spaces after dots

10 years agoMake localtime()' s documentation more succinct
Tom Hukins [Sat, 7 Jan 2012 15:37:38 +0000 (15:37 +0000)]
Make localtime()' s documentation more succinct

It's now twelve years since Y2K, so the documentation should not make
such a fuss about it.

10 years ago[perl #85670] Copy magic to ary elems properly
Father Chrysostomos [Sat, 7 Jan 2012 07:36:38 +0000 (23:36 -0800)]
[perl #85670] Copy magic to ary elems properly

On Tue Mar 08 07:26:35 2011, thospel wrote:
> #!/usr/bin/perl -l
> use Data::Dumper;
> use Scalar::Util qw(weaken);
> our @ISA;
>
> for (1..2) {
>     @ISA = qw(Foo);
>     weaken($a = \@ISA);
>     weaken($a = \$ISA[0]);
>     print STDERR Dumper(\@ISA);
> }
>
> This prints:
> $VAR1 = [
>           'Foo'
>         ];
> $VAR1 = [
>           'Foo',
>           \$VAR1->[0]
>         ];
>
> So the first time it's the expected @ISA, but the second time round it
>    automagically added a reference to to the first ISA element
>
> (bug also exists in blead)

Shorter:

#!/usr/bin/perl -l

use Scalar::Util qw(weaken);

weaken($a = \@ISA);
@ISA = qw(Foo);
use Devel::Peek; Dump \@ISA;
weaken($a = \$ISA[0]);
print scalar @ISA;   # prints 2

The dump shows the problem.  backref magic is being copied to the ele-
ment.  Put the magic in a different order, and everything is fine:

#!/usr/bin/perl -l

use Scalar::Util qw(weaken);

weaken($a = $b = []);
*ISA = $a;
@ISA = qw(Foo);
use Devel::Peek; Dump \@ISA;
weaken($a = \$ISA[0]);
print scalar @ISA;   # prints 2

This code in av_store is so wrong:

    if (SvSMAGICAL(av)) {
const MAGIC* const mg = SvMAGIC(av);
if (val != &PL_sv_undef) {
    sv_magic(val, MUTABLE_SV(av), toLOWER(mg->mg_type), 0, key);
}
if (PL_delaymagic && mg->mg_type == PERL_MAGIC_isa)
    PL_delaymagic |= DM_ARRAY_ISA;
else
   mg_set(MUTABLE_SV(av));
    }

It doesn’t follow the magic chain at all.  So anything magic could get
attached to the @ISA array, and that will be copied to the element
instead of isa magic.

Notice that MUTABLE_SV(av) is the second argument to sv_magic, so
mg->mg_obj for the element always points back to the array.

Since backref magic’s mg->mg_obj points to the backrefs array, @ISA
ends up being used as this element’s backrefs array.

What if arylen_p gets copied instead?  Let’s see:

$#ISA = -1;
@ISA = qw(Foo);
$ISA[0] = "Bar";

main->ber;

sub Bar::ber { warn "shave" }
__END__
Can't locate object method "ber" via package "main" at - line 7.

I’ve fixed this by making av_store walk the magic chain, copying any
magic for which toLOWER(mg->mg_type) != mg->mg_type.

10 years ago[perl #107440] Save av/hv on mortals stack when clearing
Father Chrysostomos [Fri, 6 Jan 2012 21:50:35 +0000 (13:50 -0800)]
[perl #107440] Save av/hv on mortals stack when clearing

In pp_undef and pp_aassign, we should put the av or hv that is being
cleared on the mortals stack (with an increased refcount), so that
destructors fired during the clearing do not free the av or hv.

I was going to put this in av_undef, etc., but pp_aassign also needs
to access the aggregate after clearing it.  We still get a crash with
that approach.

Putting the aggregate on the mortals stack in av_undef, av_clear and
h_freeentries would work, too, but might cause the aggregate to leak
too far.  That may cause problems, e.g., if it is %^H, because it may
last until the end of the current compilation unit.

Directly inside a runloop (in a pp function), it should be OK to use
the mortals stack, as it *will* be cleared ‘soon’.  This seems the
least intrusive approach.

10 years ago[perl #90632] perlfunc: Rewrite `split'
Michael Witten [Fri, 6 Jan 2012 21:11:37 +0000 (13:11 -0800)]
[perl #90632] perlfunc: Rewrite `split'

I couldn't stand the way the documenation for `split' was written;
it felt like a kludge of broken English dumped into a messy pile by
several people, each of whom was unaware of the other's work.

This variation completes sentences, adds new ones, rearranges ideas,
expands on ideas, simplifies and unifies examples, and includes more
cross references.

While the original text seemed to be written in a way that touched upon
the arguments in reverse order (which did have a hint of elegance), this
version attempts to provide the reader with the most useful information
upfront.

Thanks to Brad Baxter and Thomas R. Sibley for their constructive
criticism.

[Modified by the committer to incorporate suggestions from Aristotle
Pagaltzis and Tom Christiansen.]

10 years agodocument the upgrade of Perldoc
Ricardo Signes [Fri, 6 Jan 2012 21:07:45 +0000 (16:07 -0500)]
document the upgrade of Perldoc

10 years agoPod-Perldoc now includes test documents
Ricardo Signes [Fri, 6 Jan 2012 20:58:12 +0000 (15:58 -0500)]
Pod-Perldoc now includes test documents

do not test them as if they were Pod we ship

10 years agoUpgrade Pod-Perldoc to CPAN version 3.15_15
Ricardo Signes [Fri, 6 Jan 2012 12:58:33 +0000 (07:58 -0500)]
Upgrade Pod-Perldoc to CPAN version 3.15_15

10 years agoUncomment evals in sort.t
Father Chrysostomos [Fri, 6 Jan 2012 19:20:18 +0000 (11:20 -0800)]
Uncomment evals in sort.t

These were commented out temporarily during development.
I forgot to uncomment them before committing.

10 years agoPerlIO::scalar: allow writing to SvIOK SVs
Father Chrysostomos [Fri, 6 Jan 2012 07:55:32 +0000 (23:55 -0800)]
PerlIO::scalar: allow writing to SvIOK SVs

It used to crash if the PVX buffer happened to be null.
If the PVX buffer happened to be left over from before,
it would use that instead of the numeric value, even for
!SvPOK scalars.

10 years agoIn PerlIO::Scalar’s write, stringify refs
Father Chrysostomos [Fri, 6 Jan 2012 07:48:16 +0000 (23:48 -0800)]
In PerlIO::Scalar’s write, stringify refs

Otherwise, it won’t work with an overloaded object.

10 years agoperlsyn: spaces after dots
Father Chrysostomos [Fri, 6 Jan 2012 07:13:59 +0000 (23:13 -0800)]
perlsyn: spaces after dots

10 years agoregen pod issues
Father Chrysostomos [Fri, 6 Jan 2012 07:10:15 +0000 (23:10 -0800)]
regen pod issues

10 years agoperlsyn: wrap long verbatim line
Father Chrysostomos [Fri, 6 Jan 2012 07:10:04 +0000 (23:10 -0800)]
perlsyn: wrap long verbatim line

10 years agoIncrease $PerlIO::scalar::VERSION to 0.13
Father Chrysostomos [Fri, 6 Jan 2012 07:07:42 +0000 (23:07 -0800)]
Increase $PerlIO::scalar::VERSION to 0.13

10 years agoperlsyn: Correct ... example
Father Chrysostomos [Fri, 6 Jan 2012 07:05:16 +0000 (23:05 -0800)]
perlsyn: Correct ... example

10 years ago[perl #92706] In PerlIO::Scalar::seek, don’t assume SvPOKp
Father Chrysostomos [Fri, 6 Jan 2012 06:55:45 +0000 (22:55 -0800)]
[perl #92706] In PerlIO::Scalar::seek, don’t assume SvPOKp

Otherwise we get assertion failures.

In fact, since seeking might be just for reading, we can’t coerce and
SvGROW either.

In fact, since the scalar might be modified between seek and write,
there is no *point* in SvGROW during seek, even for SvPOK scalars.

PerlIO::scalar assumes in too many places that the scalar it is using
is its own private scalar that nothing else can modify.  Nothing could
be farther from the truth.

This commit moves the zero-fill that usually happens when seeking past
the end from seek to write.  During a write, if the current position
is past the end of the string, the intervening bytes are zero-filled
at that point, since the seek hasn’t done it.

10 years agoperlsyn: add triple-dot index entries and alias
Father Chrysostomos [Fri, 6 Jan 2012 04:50:55 +0000 (20:50 -0800)]
perlsyn: add triple-dot index entries and alias

This adds the index entries to perlsyn that were removed in
the previous commit, and mentions in perlsyn that the ellipsis
is also called a triple-dot.

10 years agoperlop: remove triple-dot
Father Chrysostomos [Fri, 6 Jan 2012 04:48:49 +0000 (20:48 -0800)]
perlop: remove triple-dot

This has been superseded by c2f1e229, which adds it
to perlsyn.

10 years ago[perl #90064] warn once for dbmopen with undef 3rd arg
Father Chrysostomos [Fri, 6 Jan 2012 04:41:08 +0000 (20:41 -0800)]
[perl #90064] warn once for dbmopen with undef 3rd arg

10 years agoIncrease $overload::VERSION to 1.17
Father Chrysostomos [Fri, 6 Jan 2012 04:24:32 +0000 (20:24 -0800)]
Increase $overload::VERSION to 1.17

10 years agoregen pod issues
Father Chrysostomos [Fri, 6 Jan 2012 04:23:52 +0000 (20:23 -0800)]
regen pod issues

10 years agoCorrect links to perlsyn and perlop
Father Chrysostomos [Fri, 6 Jan 2012 04:18:04 +0000 (20:18 -0800)]
Correct links to perlsyn and perlop

10 years ago[perl #90926] Corrections to the previous patch
Tom Christiansen [Fri, 6 Jan 2012 04:11:15 +0000 (20:11 -0800)]
[perl #90926] Corrections to the previous patch

Here is a patch against the second patch,
fixing typos reported to me.

10 years ago[perl #90926] smartmatch PATCH 2 of 2: perlsyn.pod
Tom Christiansen [Fri, 6 Jan 2012 04:02:59 +0000 (20:02 -0800)]
[perl #90926] smartmatch PATCH 2 of 2: perlsyn.pod

As previously explained, this patch against perlsyn
is part and parcel of the previous one against perlop
(two commits ago; perl #90906).

10 years ago[perl #90906] Corrections to the previous patch
Tom Christiansen [Fri, 6 Jan 2012 02:09:58 +0000 (18:09 -0800)]
[perl #90906] Corrections to the previous patch

Here is a patch against the first patch,
fixing typos reported to me.

10 years ago[perl #90906] smartmatch PATCH 1 of 2: perlop.pod
Tom Christiansen [Fri, 6 Jan 2012 02:07:48 +0000 (18:07 -0800)]
[perl #90906] smartmatch PATCH 1 of 2: perlop.pod

The thrust of this patch is to move the description of the ~~
operator into perlop where it properly belongs; given and when
remain relegated to perlsyn. This is also (nearly) the first-ever
set of examples for the smartmatch operator. Staggerment.

10 years agoUpdate CPANPLUS to CPAN version 0.9116
Chris 'BinGOs' Williams [Fri, 6 Jan 2012 00:16:06 +0000 (00:16 +0000)]
Update CPANPLUS to CPAN version 0.9116

  [DELTA]

  Changes for 0.9116      Thu Jan  5 22:45:06 2012
  ================================================
  * add NAME headings in modules with POD, Debian
    Lintian fixes, http://bugs.debian.org/65045
  * Implement reload command in the shell

10 years ago[perl #91850] utf8::decode: croak for ro scalars
Father Chrysostomos [Fri, 6 Jan 2012 00:47:54 +0000 (16:47 -0800)]
[perl #91850] utf8::decode: croak for ro scalars

10 years ago[perl #90064] perlfunc: Document special 0 mode for dbmopen
Father Chrysostomos [Thu, 5 Jan 2012 22:36:15 +0000 (14:36 -0800)]
[perl #90064] perlfunc: Document special 0 mode for dbmopen

10 years ago[perl #90648] perlop: There is no low-prec //
H.Merijn Brand [Thu, 5 Jan 2012 22:26:18 +0000 (14:26 -0800)]
[perl #90648] perlop: There is no low-prec //

10 years ago[perl #90648] perlop: There ain’t no C-style //
Father Chrysostomos [Thu, 5 Jan 2012 22:25:01 +0000 (14:25 -0800)]
[perl #90648] perlop: There ain’t no C-style //

10 years agoregen pod issues
Father Chrysostomos [Thu, 5 Jan 2012 20:29:34 +0000 (12:29 -0800)]
regen pod issues

10 years agoperlfunc: spaces after dots
Father Chrysostomos [Thu, 5 Jan 2012 17:43:31 +0000 (09:43 -0800)]
perlfunc: spaces after dots

10 years agocflags does not exists in Config.pm; use ccflags
ambs [Tue, 3 Jan 2012 22:01:27 +0000 (22:01 +0000)]
cflags does not exists in Config.pm; use ccflags

[dagolden bumped $VERSION and added ambs to the AUTHORS file]

Signed-off-by: David Golden <dagolden@cpan.org>
10 years ago[perl #90030] sort with no arguments
Father Chrysostomos [Thu, 5 Jan 2012 16:48:35 +0000 (08:48 -0800)]
[perl #90030] sort with no arguments

This eliminates the error ‘sort is now a reserved word’, which was
added for the sake of code that does close(sort) at about the time
that sort became a keyword.

As Tom Christiansen pointed out, it has been long enough.

This error only occurred with ) or ; after the keyword.  In other
cases with no arguments, like {sort} and (sort,0), it was treated as
an empty list.

So this commit makes it follow that.

Since the tests triggered it, this commit also fixes a crash caused by
commit 540dd770, which, when seeing whether it can optimise something
like ‘@a = sort @a’, assumes that there is something after the sort,
and crashes when there isn’t, as in {@a = sort}.

10 years ago[perl #89758] important perlfunc version guards
Tom Christiansen [Thu, 5 Jan 2012 07:53:33 +0000 (23:53 -0800)]
[perl #89758] important perlfunc version guards

This reminds users to put version guards on their neologisms.
We're changing Perl's basic sytnax a great deal, and users need
to understand that to use those syntactic changes will cause
weird errors if they don't put this sort of thing in.  This
sort of thing really should have gone in all along.  Let's
please continue what I have begun.

I also fixed the order: untie does not precede unshift. :(

10 years ago[perl #97464] Document actual case when @DB::args is set
Father Chrysostomos [Thu, 5 Jan 2012 07:33:23 +0000 (23:33 -0800)]
[perl #97464] Document actual case when @DB::args is set

10 years ago[perl #95548] Returned magical temps are not copied
Father Chrysostomos [Thu, 5 Jan 2012 07:28:54 +0000 (23:28 -0800)]
[perl #95548] Returned magical temps are not copied

return and leavesub, for speed, were not copying temp variables with a
refcount of 1, which is fine as long as the fact that it was not cop-
ied is not observable.

With magical variables, that *can* be observed, so we have to forego
the optimisation and copy the variable if it’s magical.

This obviously applies only to rvalue subs.

10 years agoDon’t iterate through magic with local $_
Father Chrysostomos [Wed, 4 Jan 2012 06:59:32 +0000 (22:59 -0800)]
Don’t iterate through magic with local $_

If we are going to skip all set-magic when restoring a localised tied
$_, there’s no point in looping through it.

10 years agoRestrict $[ comp warning to constants
Father Chrysostomos [Wed, 4 Jan 2012 06:56:37 +0000 (22:56 -0800)]
Restrict $[ comp warning to constants

$#a > $[ is a legitimate use of $[ with >.  So warning in that case
is not nice.  Most version comparisons are done with constants, like
5.006, so warn only for constants.

10 years agominiperl can no longer run installperl.
Craig A. Berry [Thu, 5 Jan 2012 01:07:59 +0000 (19:07 -0600)]
miniperl can no longer run installperl.

installperl now requires Porting/pod_lib.pl, which uses Digest::MD5,
which means we need something that can do dynamic loading (or has
extensions statically linked in).  But it doesn't really matter
because presumably we wouldn't be installing Perl if we hadn't
built it, so use the perl we've built rather than miniperl.

10 years agohv.c pod: 'Perl_sv_placeholder' should be 'PL_sv_placeholder'
Karl Williamson [Wed, 4 Jan 2012 18:57:12 +0000 (11:57 -0700)]
hv.c pod: 'Perl_sv_placeholder' should be 'PL_sv_placeholder'

10 years agoRemove some dead x2p-related code in vmsish.h.
Nicholas Clark [Wed, 4 Jan 2012 03:20:50 +0000 (21:20 -0600)]
Remove some dead x2p-related code in vmsish.h.

10 years ago[perl #105912] local $_ should not FETCH
Father Chrysostomos [Wed, 4 Jan 2012 06:44:49 +0000 (22:44 -0800)]
[perl #105912] local $_ should not FETCH

This commit finishes the work of 658a9f3 by skipping FETCH as well as
STORE during local($_).

10 years agoCorrect bug-report email in IO docs
Father Chrysostomos [Wed, 4 Jan 2012 03:47:12 +0000 (19:47 -0800)]
Correct bug-report email in IO docs

This was mentioned in ticket #75156.

10 years agoadd a missing apostrophe
Ricardo Signes [Wed, 4 Jan 2012 01:28:55 +0000 (20:28 -0500)]
add a missing apostrophe

thanks, Jim Keenan

10 years agoUpdate Module-Pluggable to CPAN version 4.0
Chris 'BinGOs' Williams [Tue, 3 Jan 2012 21:01:46 +0000 (21:01 +0000)]
Update Module-Pluggable to CPAN version 4.0

  [DELTA]

  No change log found for 4.0 release

10 years agoSync Maintainers.pl with CPAN for ExtUtils-ParseXS
Chris 'BinGOs' Williams [Tue, 3 Jan 2012 20:46:22 +0000 (20:46 +0000)]
Sync Maintainers.pl with CPAN for ExtUtils-ParseXS

10 years agoop.c Remove redundant continue
Father Chrysostomos [Tue, 3 Jan 2012 07:25:27 +0000 (23:25 -0800)]
op.c Remove redundant continue

10 years agoPacklist.t: add bug number
Father Chrysostomos [Tue, 3 Jan 2012 07:06:14 +0000 (23:06 -0800)]
Packlist.t: add bug number

10 years agoChange wording of \E warning
Father Chrysostomos [Tue, 3 Jan 2012 06:34:51 +0000 (22:34 -0800)]
Change wording of \E warning

following Karl Williamson’s suggestion.

10 years agotoke.c: Unpaired '\E's in double-quotish contexts should warn.
Brian Fraser [Fri, 30 Dec 2011 04:01:21 +0000 (01:01 -0300)]
toke.c: Unpaired '\E's in double-quotish contexts should warn.

Until now, an unpaired \E (that is, one lacking a \U, \L or \Q)
would just pass through unnoticed. This commit changes that
behavior to issue a warning instead.

10 years agoExtUtils::Packlist: Restore 5.005 compatibility
Father Chrysostomos [Tue, 3 Jan 2012 05:28:32 +0000 (21:28 -0800)]
ExtUtils::Packlist: Restore 5.005 compatibility

I think.  I haven’t actually tested it.

10 years ago[perl #107410] Tests for used-once warnings from ExtUtils::Packlist
Father Chrysostomos [Tue, 3 Jan 2012 05:26:52 +0000 (21:26 -0800)]
[perl #107410] Tests for used-once warnings from ExtUtils::Packlist

10 years agoAdd Joel Berger to AUTHORS
Father Chrysostomos [Tue, 3 Jan 2012 05:26:49 +0000 (21:26 -0800)]
Add Joel Berger to AUTHORS

10 years agoRemoved 'once' warnings in ExtUtils::Packlist::mkfh.
Joel Berger [Mon, 2 Jan 2012 18:55:52 +0000 (12:55 -0600)]
Removed 'once' warnings in ExtUtils::Packlist::mkfh.

Fixes cpan bug #50315.

10 years agoIncrease $ExtUtils::Packlist::VERSION to 1.46
Father Chrysostomos [Tue, 3 Jan 2012 04:33:30 +0000 (20:33 -0800)]
Increase $ExtUtils::Packlist::VERSION to 1.46

10 years agoCorrect test count in regen.t
Father Chrysostomos [Tue, 3 Jan 2012 00:43:59 +0000 (16:43 -0800)]
Correct test count in regen.t

10 years agoExplain perlguts changes in mg_vtable.pl comment
Father Chrysostomos [Tue, 3 Jan 2012 00:43:20 +0000 (16:43 -0800)]
Explain perlguts changes in mg_vtable.pl comment

10 years agoRegenerate perlguts’ mg table automatically
Father Chrysostomos [Tue, 3 Jan 2012 00:40:38 +0000 (16:40 -0800)]
Regenerate perlguts’ mg table automatically

regen/mg_vtable.pl was modified a while ago to generate the table
for copying and pasting, but at least twice since then it has not
been updated properly; once by me and once by the author of that
part of regen/mg_vtable.pl.

10 years agoperldelta for GV non-cloning
Father Chrysostomos [Mon, 2 Jan 2012 22:38:21 +0000 (14:38 -0800)]
perldelta for GV non-cloning

10 years agoperldelta for %n
Father Chrysostomos [Mon, 2 Jan 2012 22:36:41 +0000 (14:36 -0800)]
perldelta for %n

10 years agoperldelta for hash element deletion and destructors
Father Chrysostomos [Mon, 2 Jan 2012 22:34:22 +0000 (14:34 -0800)]
perldelta for hash element deletion and destructors

10 years agoperldelta for method deletion
Father Chrysostomos [Mon, 2 Jan 2012 22:32:15 +0000 (14:32 -0800)]
perldelta for method deletion

10 years agoperldelta for system SIGCHLD interference
Father Chrysostomos [Mon, 2 Jan 2012 22:30:30 +0000 (14:30 -0800)]
perldelta for system SIGCHLD interference

10 years agoperldelta for #! line errors
Father Chrysostomos [Mon, 2 Jan 2012 22:28:05 +0000 (14:28 -0800)]
perldelta for #! line errors

10 years agoperldelta for POSIX sigsuspend
Father Chrysostomos [Mon, 2 Jan 2012 22:24:20 +0000 (14:24 -0800)]
perldelta for POSIX sigsuspend

10 years agoperldelta: update status comment
Father Chrysostomos [Mon, 2 Jan 2012 22:18:34 +0000 (14:18 -0800)]
perldelta: update status comment

10 years agoperldelta for reifying tied arrays
Father Chrysostomos [Mon, 2 Jan 2012 22:16:48 +0000 (14:16 -0800)]
perldelta for reifying tied arrays

10 years agoperldelta for tied @DB::args
Father Chrysostomos [Mon, 2 Jan 2012 22:10:26 +0000 (14:10 -0800)]
perldelta for tied @DB::args

10 years agoperldelta for splain’s multiline support
Father Chrysostomos [Mon, 2 Jan 2012 22:06:21 +0000 (14:06 -0800)]
perldelta for splain’s multiline support

10 years agoperldelta for splain’s backtrace formatting
Father Chrysostomos [Mon, 2 Jan 2012 22:04:42 +0000 (14:04 -0800)]
perldelta for splain’s backtrace formatting

10 years agoperldelta for lvalue fixes
Father Chrysostomos [Mon, 2 Jan 2012 21:59:39 +0000 (13:59 -0800)]
perldelta for lvalue fixes

10 years agoperldelta for Data::Dumper and *{""}
Father Chrysostomos [Mon, 2 Jan 2012 21:48:15 +0000 (13:48 -0800)]
perldelta for Data::Dumper and *{""}

10 years agoperldelta for Data::Dumper and vstrings
Father Chrysostomos [Mon, 2 Jan 2012 21:47:09 +0000 (13:47 -0800)]
perldelta for Data::Dumper and vstrings

10 years agoperldelta for PerlIO::get_layers and numbers
Father Chrysostomos [Mon, 2 Jan 2012 21:44:24 +0000 (13:44 -0800)]
perldelta for PerlIO::get_layers and numbers

10 years agoperldelta for PerlIO::get_layers and get-magic
Father Chrysostomos [Mon, 2 Jan 2012 21:42:04 +0000 (13:42 -0800)]
perldelta for PerlIO::get_layers and get-magic

10 years agoperldelta for select fixes
Father Chrysostomos [Mon, 2 Jan 2012 21:40:29 +0000 (13:40 -0800)]
perldelta for select fixes

10 years agoperldelta for deparsing $( $| $) in regexps
Father Chrysostomos [Mon, 2 Jan 2012 21:36:05 +0000 (13:36 -0800)]
perldelta for deparsing $( $| $) in regexps

10 years agoperldelta for /$#a/ deparsing
Father Chrysostomos [Mon, 2 Jan 2012 21:34:47 +0000 (13:34 -0800)]
perldelta for /$#a/ deparsing

10 years agoperldelta for open(foo::bar) warning
Father Chrysostomos [Mon, 2 Jan 2012 21:32:52 +0000 (13:32 -0800)]
perldelta for open(foo::bar) warning

10 years agoperldelta for deparsing strict [perl #24027]
Father Chrysostomos [Mon, 2 Jan 2012 21:28:21 +0000 (13:28 -0800)]
perldelta for deparsing strict [perl #24027]

10 years agoperldelta for B::Deparse ambient_pragmas fix
Father Chrysostomos [Mon, 2 Jan 2012 21:26:35 +0000 (13:26 -0800)]
perldelta for B::Deparse ambient_pragmas fix

10 years agoperldelta for helem and hslice on null values
Father Chrysostomos [Mon, 2 Jan 2012 21:23:51 +0000 (13:23 -0800)]
perldelta for helem and hslice on null values

10 years agoCorrect comment in APItest’s hash.t
Father Chrysostomos [Mon, 2 Jan 2012 21:20:07 +0000 (13:20 -0800)]
Correct comment in APItest’s hash.t

10 years agoperldelta for newHVhv on tied hashes
Father Chrysostomos [Mon, 2 Jan 2012 21:18:51 +0000 (13:18 -0800)]
perldelta for newHVhv on tied hashes

10 years agoperldelta for deparsing features
Father Chrysostomos [Mon, 2 Jan 2012 21:17:14 +0000 (13:17 -0800)]
perldelta for deparsing features

10 years agoperldelta for feature revamp
Father Chrysostomos [Mon, 2 Jan 2012 20:57:46 +0000 (12:57 -0800)]
perldelta for feature revamp

10 years agoutf8_heavy.pl: Skip unnecessary work for official properties
Karl Williamson [Mon, 2 Jan 2012 23:12:21 +0000 (16:12 -0700)]
utf8_heavy.pl: Skip unnecessary work for official properties

The tables that mktables generates are well behaved, and so the checks
and sorting that are done for user-defined properties may be skipped.

tainting needs to be preserved because $list can be passed in already
tainted.

This is also in preparation for Unicode 6.1, in which one table will
legitimately have duplicate entries that the old code removed.

10 years agoutf8_heavy.pl: Comment and white-space only
Karl Williamson [Mon, 2 Jan 2012 22:44:37 +0000 (15:44 -0700)]
utf8_heavy.pl: Comment and white-space only

This code is indented in preparation to be placed in an enclosing
conditional.

10 years agoMerge the refactoring of installperl that fixes Pod installation bugs.
Nicholas Clark [Mon, 2 Jan 2012 17:26:39 +0000 (18:26 +0100)]
Merge the refactoring of installperl that fixes Pod installation bugs.

10 years agoNote the user-visible installperl changes in perldelta.
Nicholas Clark [Mon, 2 Jan 2012 17:02:38 +0000 (18:02 +0100)]
Note the user-visible installperl changes in perldelta.

10 years agoinstallperl now uses get_pod_metadata() to install dual-life pods correctly.
Nicholas Clark [Tue, 27 Dec 2011 18:44:27 +0000 (19:44 +0100)]
installperl now uses get_pod_metadata() to install dual-life pods correctly.

This ensures that perlfaq*, perlglossary, perlxs, perlxstut and perldoc's
Pod files are installed in the same directory as perl.pod, perlintro.pod etc,
as they were in 5.14.0 and earlier releases. This commit is analogous to
commit d6a39ee219352c8f, which corrected the install location for man pages.

10 years agoIn installperl, convert the File::Find::find(..., 'lib/') to no_chdir.
Nicholas Clark [Tue, 27 Dec 2011 15:48:53 +0000 (16:48 +0100)]
In installperl, convert the File::Find::find(..., 'lib/') to no_chdir.

This eliminates the localized variable $::depth, which really only existed
as a bodge to get plausible pathnames in the dry run output when run with
-n. (Note, not for the output when running for real without -n, nor for
the verbose output)

10 years agoIn installperl, don't chdir 'lib' before finding files in lib/
Nicholas Clark [Mon, 26 Dec 2011 13:05:13 +0000 (14:05 +0100)]
In installperl, don't chdir 'lib' before finding files in lib/

10 years agoIn installperl, add the trailing / to the value held in $::depth.
Nicholas Clark [Sat, 24 Dec 2011 13:52:50 +0000 (14:52 +0100)]
In installperl, add the trailing / to the value held in $::depth.

This should make future refactoring simpler.

10 years agoIn installperl, initialise $do_install{arch,priv}lib at their declarations.
Nicholas Clark [Sat, 24 Dec 2011 13:26:43 +0000 (14:26 +0100)]
In installperl, initialise $do_install{arch,priv}lib at their declarations.

Also, add an editor block.

10 years agoIn installperl, simplify the strip/chmod code for libraries on Darwin.
Nicholas Clark [Sat, 24 Dec 2011 12:40:24 +0000 (13:40 +0100)]
In installperl, simplify the strip/chmod code for libraries on Darwin.