This is a live mirror of the Perl 5 development currently hosted at
12 years agobump ExtUtils-CBuilder version for blead change
Zefram [Sun, 19 Dec 2010 11:36:15 +0000 (11:36 +0000)]
bump ExtUtils-CBuilder version for blead change

12 years agobump autodie version for blead test fix
Zefram [Sun, 19 Dec 2010 10:49:16 +0000 (10:49 +0000)]
bump autodie version for blead test fix

12 years agoperldelta up to ca88a729
Zefram [Sun, 19 Dec 2010 09:00:09 +0000 (09:00 +0000)]
perldelta up to ca88a729

12 years agoperldelta for 8e88cfee26d866223a6b3bfffce6270271de00db
Zefram [Sun, 19 Dec 2010 08:50:06 +0000 (08:50 +0000)]
perldelta for 8e88cfee26d866223a6b3bfffce6270271de00db

12 years agoperldelta for 9cef83062267e94311e1fd8744396e440642738e
Zefram [Sun, 19 Dec 2010 08:46:30 +0000 (08:46 +0000)]
perldelta for 9cef83062267e94311e1fd8744396e440642738e

12 years agoperldelta for bf5522a13a381257966e7ed6b731195a873b153e
Zefram [Sun, 19 Dec 2010 08:40:12 +0000 (08:40 +0000)]
perldelta for bf5522a13a381257966e7ed6b731195a873b153e

12 years agoperldelta for d9a4b459f94297889956ac3adc42707365f274c2
Zefram [Sun, 19 Dec 2010 08:36:00 +0000 (08:36 +0000)]
perldelta for d9a4b459f94297889956ac3adc42707365f274c2

12 years agoTest B::Concise’s -tree mode
Father Chrysostomos [Sun, 19 Dec 2010 01:21:41 +0000 (17:21 -0800)]
Test B::Concise’s -tree mode

12 years agoUpdate Archive-Tar to CPAN version 1.74
Chris 'BinGOs' Williams [Sat, 18 Dec 2010 22:36:50 +0000 (22:36 +0000)]
Update Archive-Tar to CPAN version 1.74


  * important changes in version 1.74 18/12/2010
  - Skip extracting pax extended headers, reported as
    RT #64038

12 years agoRe-order the tests to match the comment
Tony Cook [Sat, 18 Dec 2010 00:05:39 +0000 (11:05 +1100)]
Re-order the tests to match the comment

Win32 has been failing in re/re.t, maybe this will fix it.

12 years agore.t: Use portable C locale
Karl Williamson [Thu, 16 Dec 2010 15:34:35 +0000 (08:34 -0700)]
re.t: Use portable C locale

The locale en_US.UTF-8 has varying and wrong definitions depending on
manufacturer.  Using the portable C locale works just as well for the
purposes of these tests.

12 years agore.t: Avoid encoding issues by using hex chars
Karl Williamson [Thu, 16 Dec 2010 05:33:07 +0000 (22:33 -0700)]
re.t: Avoid encoding issues by using hex chars

re.t is getting encoded as utf8 in places, which confuses things.  So
use a hex notation to enter a non-ASCII char

12 years agoupdate warnings tests to cope with %z now being valid. use %y instead
Chip Salzenberg [Fri, 17 Dec 2010 19:21:54 +0000 (11:21 -0800)]
update warnings tests to cope with %z now being valid.  use %y instead

12 years agodocument new printf size modifiers
Chip Salzenberg [Fri, 17 Dec 2010 18:55:20 +0000 (10:55 -0800)]
document new printf size modifiers

12 years agosmoke tests for %hhd %zd %td
Chip Salzenberg [Fri, 17 Dec 2010 18:46:56 +0000 (10:46 -0800)]
smoke tests for %hhd %zd %td

12 years agoadd %jd to printf where C99 exists; tweak %zd and %td
Chip Salzenberg [Fri, 17 Dec 2010 18:13:43 +0000 (10:13 -0800)]
add %jd to printf where C99 exists; tweak %zd and %td

12 years agoRefactor ExtUtils::Constant's tests to pass cleanly on older versions.
Nicholas Clark [Fri, 17 Dec 2010 17:18:33 +0000 (17:18 +0000)]
Refactor ExtUtils::Constant's tests to pass cleanly on older versions.

Work round bugs with the value of ~0 on 5.005 built on 64 bit systems.
Convert @INC to absolute paths on 5.6.x so that
ExtUtils::Constant::Aaargh56Hash will load from blib.
Remove $runperl, which is no longer needed.

12 years agoRefactor ExtUtils::Constant::Utils backwards compatibility code.
Nicholas Clark [Fri, 17 Dec 2010 17:16:46 +0000 (17:16 +0000)]
Refactor ExtUtils::Constant::Utils backwards compatibility code.

Avoid warnings from letting 5.005 even think about POSIX charclass constants
in regexps, and use compile-time constants to enable backcompat features.

12 years agoConvert ext/PerlIO-encoding/t/encoding.t to Test::More.
Nicholas Clark [Fri, 17 Dec 2010 15:16:33 +0000 (15:16 +0000)]
Convert ext/PerlIO-encoding/t/encoding.t to Test::More.

12 years agoConvert ext/Opcode/t/ops.t to Test::More
Nicholas Clark [Fri, 17 Dec 2010 11:07:32 +0000 (11:07 +0000)]
Convert ext/Opcode/t/ops.t to Test::More

12 years agoConvert ext/IPC-Open3/t/IPC-Open3.t to Test::More
Nicholas Clark [Fri, 17 Dec 2010 10:41:39 +0000 (10:41 +0000)]
Convert ext/IPC-Open3/t/IPC-Open3.t to Test::More

Unfortunately the gubbins of about 25% of its tests still rely on causing
subprocesses to emit the correct TAP, so part of it has to use a an explicit
test counter outside of Test::Builder.

12 years agoIn, change checkOutput(...) to 5 calls to is().
Nicholas Clark [Fri, 17 Dec 2010 09:31:11 +0000 (09:31 +0000)]
In, change checkOutput(...) to 5 calls to is().

Replacing a 5 armed boolean fed to ok() with 5 calls to is() gives much better
diagnostics on failure.

12 years agoperldelta for efef081 and 91fad77
Father Chrysostomos [Fri, 17 Dec 2010 06:31:16 +0000 (22:31 -0800)]
perldelta for efef081 and 91fad77

12 years agoperldelta for 20f15c4
Father Chrysostomos [Fri, 17 Dec 2010 06:28:22 +0000 (22:28 -0800)]
perldelta for 20f15c4

12 years agoperldelta for a42d024
Father Chrysostomos [Fri, 17 Dec 2010 06:23:42 +0000 (22:23 -0800)]
perldelta for a42d024

12 years agoperldelta for 8e720305
Father Chrysostomos [Fri, 17 Dec 2010 05:56:49 +0000 (21:56 -0800)]
perldelta for 8e720305

12 years agoIncrease B::Concise’s version
Father Chrysostomos [Fri, 17 Dec 2010 05:36:02 +0000 (21:36 -0800)]
Increase B::Concise’s version

12 years agoFix [perl #80632] -MO=Concise,-tree format
Reini Urban [Mon, 13 Dec 2010 19:06:33 +0000 (20:06 +0100)]
Fix [perl #80632] -MO=Concise,-tree format

Some time between 5.8.3 and 5.8.4, the -tree output
format started getting extra line breaks.

12 years ago[perl #80548] perldelta for DTrace package name change
David Leadbeater [Fri, 17 Dec 2010 04:15:57 +0000 (20:15 -0800)]
[perl #80548] perldelta for DTrace package name change

12 years ago[perl #80674] Fix compilation with very old versions of glibc
David Leadbeater [Fri, 17 Dec 2010 04:14:40 +0000 (20:14 -0800)]
[perl #80674] Fix compilation with very old versions of glibc

__priority_which_t does not exist on glibc 2.1.

sin6_scope is not present in the set of kernel headers my copy of glibc 2.1 is
using. (The presence of sin6_scope in sockaddr_in6 should maybe be a Configure

blead now compiles on a positively ancient box -- although the Socket tests

12 years agoSilence some data truncation compiler warnings
Jan Dubois [Thu, 16 Dec 2010 06:41:47 +0000 (22:41 -0800)]
Silence some data truncation compiler warnings

12 years extend EBCDIC functions to beyond 255
Karl Williamson [Thu, 16 Dec 2010 17:38:09 +0000 (10:38 -0700)] extend EBCDIC functions to beyond 255

Allow the functions to handle non-latin1 input.  This would only show up
on EBCDIC platforms.

12 years agoVersion bumps for modules changed by a6d37805ca8a9ba8 ($Id$ removal).
Nicholas Clark [Thu, 16 Dec 2010 16:48:26 +0000 (16:48 +0000)]
Version bumps for modules changed by a6d37805ca8a9ba8 ($Id$ removal).

12 years agoRemove "dead" RCS $Id$ tags from files that we own.
Nicholas Clark [Thu, 16 Dec 2010 16:28:44 +0000 (16:28 +0000)]
Remove "dead" RCS $Id$ tags from files that we own.

All files have been modified more recently than their tag, rendering
information in the tag redundant.

12 years agoonly call amagic_deref_call() if we have to
David Mitchell [Thu, 16 Dec 2010 16:14:06 +0000 (16:14 +0000)]
only call amagic_deref_call() if we have to

12 years agouniprops.pod: restore properties counts
Karl Williamson [Thu, 16 Dec 2010 16:11:27 +0000 (09:11 -0700)]
uniprops.pod: restore properties counts

As a side effect of commit bd9ebcfd4a6439f22bcaf330cd6d936de78362c5,
which was to make smalle unicore tables, the counts of the number of
code points matched by the various properties in perluniprops.pod were
removed.  This restores them

12 years agoMove common code from ext/[GONS]DBM_File/t/[gons]dbm.t to t/lib/
Nicholas Clark [Thu, 16 Dec 2010 16:02:20 +0000 (16:02 +0000)]
Move common code from ext/[GONS]DBM_File/t/[gons]dbm.t to t/lib/

This eliminates 1445 lines, ie almost 500 lines duplicated fourfold.

12 years agoConverge ext/[GNOS]DBM_File/t/[gnos]dbm.t by parameterising the class name.
Nicholas Clark [Thu, 16 Dec 2010 14:09:42 +0000 (14:09 +0000)]
Converge ext/[GNOS]DBM_File/t/[gnos]dbm.t by parameterising the class name.

12 years agoConverge ext/[GNOS]DBM_File/t/[gnos]dbm.t further.
Nicholas Clark [Thu, 16 Dec 2010 13:25:05 +0000 (13:25 +0000)]
Converge ext/[GNOS]DBM_File/t/[gnos]dbm.t further.

Including Cross propagating some fixes:

grep in void context warning (f84167b37281b9fd c57cf257e9e58200), but improve
it by avoiding void context entirely, by actually testing the results :-)

"cleaner close on tests, take 2", d1e4d418969ad3c5

12 years agoConverge ext/[GNOS]DBM_File/t/[gnos]dbm.t by using the same filename.
Nicholas Clark [Thu, 16 Dec 2010 11:37:43 +0000 (11:37 +0000)]
Converge ext/[GNOS]DBM_File/t/[gnos]dbm.t by using the same filename.

Choose the 1 dot form used by sdbm.t, to keep VMS happy. Also, propagate into
ndbm.t the part of the test for 20001013.009 that cbc5248d01a71061 missed.
Move the exist tests from f4b9d8806d76b352 earlier in sdbm.t, to increase
consistency - the alternative attempts to have 2 DBM files open simultaneously,
which ODBM_File doesn't support. (Implied TODO: add an explicit test for this
to the other 3.)

12 years agoTweak gdbm.t to use GDBM_WRITER instead of GDBM_WRCREAT when re-opening a file.
Nicholas Clark [Thu, 16 Dec 2010 14:47:27 +0000 (15:47 +0100)]
Tweak gdbm.t to use GDBM_WRITER instead of GDBM_WRCREAT when re-opening a file.

The analagous tests for [nos]dbm.t don't use |O_CREAT for this case. gdbm.t
has been using GDBM_WRCREAT here since the file was added by Larry in 5.000

12 years agoFix typo in perlvar
Chris 'BinGOs' Williams [Thu, 16 Dec 2010 14:09:55 +0000 (14:09 +0000)]
Fix typo in perlvar

12 years agoregcomp.c: Optimize [cC] char class to EXACTF
Karl Williamson [Thu, 16 Dec 2010 01:11:44 +0000 (18:11 -0700)]
regcomp.c: Optimize [cC] char class to EXACTF

A two character character class where the two elements are the folds of
each other can be optimized into an EXACTF regnode.  This should not
change the speed of execution appreciably, except that EXACTF regnodes
are candidates for further optimization by combining with adjacent nodes
of the same type.

This commit brings the optimization level up to somewhat better than
when I started mucking around with ANYOF nodes.

12 years agoregcomp.c: More work on ANYOF_CLASS
Karl Williamson [Wed, 15 Dec 2010 23:33:02 +0000 (16:33 -0700)]
regcomp.c: More work on ANYOF_CLASS

I overlooked two cases in a previous commit where it would be advisable
to make changes in case the ANYOF_CLASS bit needs to be combined with

12 years agol1_char_class_tab.h: include multi-char folds
Karl Williamson [Wed, 15 Dec 2010 23:31:02 +0000 (16:31 -0700)]
l1_char_class_tab.h: include multi-char folds

This patch is the result of running

12 years extend for multi-char folds
Karl Williamson [Wed, 15 Dec 2010 21:44:45 +0000 (14:44 -0700)] extend for multi-char folds

12 years agoregexec.c: white space changes due to prev. commit
Karl Williamson [Wed, 15 Dec 2010 22:34:48 +0000 (15:34 -0700)]
regexec.c: white space changes due to prev. commit

This is essentially a white-space patch.  Commit
dbaae6dafbe2bb8bf9fcd1e83874b9a683e0290f effectively removed two nests
of blocks.  This finishes that removal, outdents the previous innter
blocks, and reformats the comments for the extra space.

One declaration was moved as a result of removing a block.

12 years agoregex: Multi-char /i shouldnt match single char []
Karl Williamson [Wed, 15 Dec 2010 22:22:38 +0000 (15:22 -0700)]
regex: Multi-char /i shouldnt match single char []

":\N{LATIN SMALL LIGATURE ST}:" !~ /:[_st]:/i

because it is two character, but previously it did match.  The cause of
this is that the code was using swash_fetch() to test if the fold
matches.  But swash_fetch() only checks if the first character matches,
not all characters, so it was falsely returning true.

This is an intermediate commit, with some cleanup of blocks, comments,
and accompanying indentation of regexec.c to follow immediately.

12 years agoRevamp t/uni/fold.t
Karl Williamson [Wed, 15 Dec 2010 16:57:25 +0000 (09:57 -0700)]
Revamp t/uni/fold.t

This patch revamps fold.t but using essentially the same tests on
essentially the same character set.  It:
    1) Works on EBCDIC
    2) Uses
    3) Separates out the 8 tests per character that previously were all
        combined into a single test per character
    4) Outputs on each line the actual test performed
    5) Corrects and hardens some tests on multi-character folding

To expand on point 5):  Previously, the wrong behavior was tested for;
correct behavior failed.  For example,
    ":\N{LATIN SMALL LIGATURE ST}:" =~ /:[_st]:/i
previously passed.  But the fold of the string is two characters, and so
should not match a one-character long character class.  Instead it
should match:
    ":\N{LATIN SMALL LIGATURE ST}:" =~ /:[_st]{2}:/i
The new test includes TODO tests for both of them.
    ":\N{LATIN SMALL LIGATURE ST}:" !~ /:[_st]:/i
    ":\N{LATIN SMALL LIGATURE ST}:" =~ /:[_st]{2}:/i

Also the inverse relation
    ":st:" =~ /:[_\N{LATIN SMALL LIGATURE ST}]:/i
passes, semi-coincidentally, for some.  By changing the test to
    ":ST:" =~ /:[_\N{LATIN SMALL LIGATURE ST}]:/i
they all fail, (and are made TODO's).

12 years agodon't upgrade overload IV return values to NV
David Mitchell [Wed, 15 Dec 2010 19:38:17 +0000 (19:38 +0000)]
don't upgrade overload IV return values to NV

(if we can avoid it).

Fix for RT #77456. Basically it extends the usage of the AMGf_numeric flag
to the remaining overloadable numeric ops that behave differently with IV
and NV.

12 years agoUpgrade Devel::SelfStubber from 1.04 to 1.05
Florian Ragwitz [Wed, 15 Dec 2010 19:14:31 +0000 (20:14 +0100)]
Upgrade Devel::SelfStubber from 1.04 to 1.05

This is necessary so the CPAN version can fix the INSTALLDIRS parameter it uses
on older perls.

12 years agoUpgrade Dumpvalue from version 1.14 to 1.15
Florian Ragwitz [Wed, 15 Dec 2010 19:14:31 +0000 (20:14 +0100)]
Upgrade Dumpvalue from version 1.14 to 1.15

This is necessary so the CPAN version can fix the INSTALLDIRS parameter it uses
on older perls.

12 years agoUpgrade Env from version 1.01 to 1.02
Florian Ragwitz [Wed, 15 Dec 2010 19:14:31 +0000 (20:14 +0100)]
Upgrade Env from version 1.01 to 1.02

This is necessary so the CPAN version can fix the INSTALLDIRS parameter it uses
on older perls.

12 years agoConvert ext/SDBM_File/t/sdbm.t to Test::More.
Nicholas Clark [Wed, 15 Dec 2010 16:48:48 +0000 (16:48 +0000)]
Convert ext/SDBM_File/t/sdbm.t to Test::More.

12 years agoConvert ext/ODBM_File/t/odbm.t to Test::More.
Nicholas Clark [Wed, 15 Dec 2010 16:20:39 +0000 (16:20 +0000)]
Convert ext/ODBM_File/t/odbm.t to Test::More.

12 years agoConvert ext/NDBM_File/t/ndbm.t to Test::More.
Nicholas Clark [Wed, 15 Dec 2010 16:01:55 +0000 (16:01 +0000)]
Convert ext/NDBM_File/t/ndbm.t to Test::More.

12 years agoConvert ext/GDBM_File/t/gdbm.t to Test::More.
Nicholas Clark [Wed, 15 Dec 2010 15:25:33 +0000 (15:25 +0000)]
Convert ext/GDBM_File/t/gdbm.t to Test::More.

12 years agoRemove redundant use strict,warnings,?DBM_File from ext/?DBM_File/t/?dbm.t.
Nicholas Clark [Wed, 15 Dec 2010 15:08:04 +0000 (15:08 +0000)]
Remove redundant use strict,warnings,?DBM_File from ext/?DBM_File/t/?dbm.t.

12 years agoAdd another e-mail address for Renee Baecker's to
Nicholas Clark [Wed, 15 Dec 2010 16:28:54 +0000 (16:28 +0000)]
Add another e-mail address for Renee Baecker's to

12 years agoChange ExtUtils::CBuilder upstream to blead
David Golden [Wed, 15 Dec 2010 15:47:14 +0000 (10:47 -0500)]
Change ExtUtils::CBuilder upstream to blead

ExtUtils::CBuilder now has blead as its upstream repository.  The
distribution has been moved to the dist/ directory consistent with
this change.

12 years agoadd some tests for the regex flags u, d, and l
OtrsUser [Tue, 14 Dec 2010 11:11:48 +0000 (03:11 -0800)]
add some tests for the regex flags u, d, and l

12 years agoConvert all File::Glob tests to Test::More.
Nicholas Clark [Wed, 15 Dec 2010 11:47:19 +0000 (11:47 +0000)]
Convert all File::Glob tests to Test::More.

(Apart from basic.t, which was already using Test::More)

12 years agoregcomp.c: Fix VC6 compiler warnings
Karl Williamson [Wed, 15 Dec 2010 03:34:02 +0000 (20:34 -0700)]
regcomp.c: Fix VC6 compiler warnings

The number passed here is never larger than 255.

12 years agoskip_all() is a separate function in
Jan Dubois [Wed, 15 Dec 2010 00:04:00 +0000 (16:04 -0800)]
skip_all() is a separate function in

and not just an argument to plan().

12 years agoDefine our own sockaddr_in6 for VC6
Jan Dubois [Tue, 14 Dec 2010 23:48:10 +0000 (15:48 -0800)]
Define our own sockaddr_in6 for VC6

The version defined by the VC6 headers is missing the sin6_scope_id
field, used by Socket.xs.  Ideally VC6 should be used with later
versions of the header files from a Windows Platform SDK, but this
patch at least keeps it compilable with plain VC6.  I have not verified
that the IPv6 code will actually works with this configuration.

12 years ago#define some symbols for VC6 compatibility
Jan Dubois [Tue, 14 Dec 2010 23:40:17 +0000 (15:40 -0800)]
#define some symbols for VC6 compatibility

HWND_MESSAGE and WC_NO_BEST_FIT_CHARS are not defined by the VC6
headers unless WINVER is set to 0x0500 or higher, but then the
headers will generate a lot of warnings because Windows 2000
was still in beta when those headers where released.
PROCESSOR_ARCHITCTURE_AMD64 is not defined at all in the
VC6 headers.

12 years agoDual-life Dumpvalue
Florian Ragwitz [Wed, 15 Dec 2010 00:00:30 +0000 (01:00 +0100)]
Dual-life Dumpvalue

12 years agoUpdate Digest-SHA to CPAN version 5.50
Chris 'BinGOs' Williams [Tue, 14 Dec 2010 18:56:42 +0000 (18:56 +0000)]
Update Digest-SHA to CPAN version 5.50


  5.50  Tue Dec 14 06:20:08 MST 2010
        - adopted convention that '-' always means STDIN
                -- actual filename '-' accessed as './-'
                -- accords with behavior of sha1sum/md5sum
        - corrected undefined subroutine oversight in shasum
                -- inadvertent migration of _bail() from

12 years agoConvert ext/Fcntl/t/syslfs.t to Test::More and t/op/lfs.t to
Nicholas Clark [Tue, 14 Dec 2010 16:59:11 +0000 (16:59 +0000)]
Convert ext/Fcntl/t/syslfs.t to Test::More and t/op/lfs.t to

12 years agoAdd $Tests_Are_Passing to, analogous to Test::Builder's is_passing().
Nicholas Clark [Tue, 14 Dec 2010 16:58:10 +0000 (16:58 +0000)]
Add $Tests_Are_Passing to, analogous to Test::Builder's is_passing().

12 years agoIn Fcntl's syslfs.t and t/op/lfs.t, eliminate bye().
Nicholas Clark [Tue, 14 Dec 2010 16:12:39 +0000 (16:12 +0000)]
In Fcntl's syslfs.t and t/op/lfs.t, eliminate bye().

Its cleanup actions are implicit in the END block, so replace C<warn ...; bye>;
with C<die ...> and other calls of C<bye;> with C<exit 0;>

Also, remove the newlines from the strings passed to die, to make the
diagnostics more useful for locating failures.

12 years agoIn Fcntl's syslfs.t and t/op/lfs.t, eliminate zap().
Nicholas Clark [Tue, 14 Dec 2010 15:43:02 +0000 (15:43 +0000)]
In Fcntl's syslfs.t and t/op/lfs.t, eliminate zap().

Now that we're using tempfiles, it no longer contains code to unlink the test
files, only a close. Inline the C<close BIG> into bye(), and remove the other
call to zap(), which was immediately after an explicit close of BIG.

12 years agoRefactor syslfs.t and lfs.t to call bye() directly from explain() when skipping.
Nicholas Clark [Mon, 13 Dec 2010 17:20:44 +0000 (17:20 +0000)]
Refactor syslfs.t and lfs.t to call bye() directly from explain() when skipping.

This will make it easier to refactoring to use Test::More/

12 years agoConvert ext/Fcntl/t/syslfs.t to File::Temp
Nicholas Clark [Mon, 13 Dec 2010 17:01:09 +0000 (17:01 +0000)]
Convert ext/Fcntl/t/syslfs.t to File::Temp

This reduces the differences between ext/Fcntl/t/syslfs.t and t/op/lfs.t

12 years agoReduce inconsistencies between ext/Fcntl/t/syslfs.t and t/op/lfs.t
Nicholas Clark [Mon, 13 Dec 2010 14:42:32 +0000 (14:42 +0000)]
Reduce inconsistencies between ext/Fcntl/t/syslfs.t and t/op/lfs.t

The two are testing the same functionality, and comments in each reference the
other. However, the two have diverged, sometimes in the same commit, sometimes
when corrections have been applied to only one. (eg 972720f939262dd0)

12 years agoCheck return values in the test programs run by syslfs.t and lfs.t
Nicholas Clark [Mon, 13 Dec 2010 16:08:38 +0000 (16:08 +0000)]
Check return values in the test programs run by syslfs.t and lfs.t

Also fix a bug introduced in 1c25d394345c1b97, which accidentally neutered the
test program of t/op/lfs.t, causing it to attempt to open an empty file, and
hence do nothing and then always exit with 0.

12 years agoBe more consistent with 'the Perl core'
Abigail [Tue, 14 Dec 2010 16:35:18 +0000 (17:35 +0100)]
Be more consistent with 'the Perl core'

12 years agopermit labels to be stacked
Zefram [Sun, 12 Dec 2010 11:41:55 +0000 (11:41 +0000)]
permit labels to be stacked

Liberalise label syntax a little more, by allowing multiple adjacent
labels with no intervening statements, as in "foo: bar: baz:".

12 years agomark blockhooks API as experimental
Zefram [Mon, 13 Dec 2010 20:53:19 +0000 (20:53 +0000)]
mark blockhooks API as experimental

12 years agoExplicitly close filehandle in ExtUtils::CBuilder::have_compiler().
Craig A. Berry [Mon, 13 Dec 2010 13:21:28 +0000 (07:21 -0600)]
Explicitly close filehandle in ExtUtils::CBuilder::have_compiler().

A recent change removed the lexical scope that would have caused
the lexical filehandle on the temporary source file to get closed
automatically.  If the filehandle is still open for write, you
can't compile it or delete unless it's open for shared access,
which is not the default on VMS (and deleting an open fle might
be a problem on Windows as well).

Without deleting the .c files, a subsequent make (even after a make
realclean) will identify CBuilder as an XS extension and fail when
it can't compile the non-existent CBuilder.c, which breaks the build
for second and subsequent runs of a smoke test.

12 years agoperldelta tweaks
Father Chrysostomos [Mon, 13 Dec 2010 06:38:24 +0000 (22:38 -0800)]
perldelta tweaks

12 years agoperldelta for a4256772
Father Chrysostomos [Mon, 13 Dec 2010 06:32:22 +0000 (22:32 -0800)]
perldelta for a4256772

12 years agoperldelta for ad021bf/[perl #76026]
Father Chrysostomos [Mon, 13 Dec 2010 04:45:47 +0000 (20:45 -0800)]
perldelta for ad021bf/[perl #76026]

12 years agoperldelta for 78cdf10
Father Chrysostomos [Mon, 13 Dec 2010 04:44:26 +0000 (20:44 -0800)]
perldelta for 78cdf10

12 years agoperldelta for 88b5a87
Father Chrysostomos [Mon, 13 Dec 2010 02:12:37 +0000 (18:12 -0800)]
perldelta for 88b5a87

12 years agoReduce redundant macros and prototypes in vmsish.h.
Craig A. Berry [Sun, 12 Dec 2010 23:43:32 +0000 (17:43 -0600)]
Reduce redundant macros and prototypes in vmsish.h.

In imitation of e64ca59f2852bc5d0450eea5ed412609a046829d, we don't
really need different prototypes or macros for MULTIPLICITY and
non-MULTIPLICITY.  For now, leave drop-in replacements of CRTL
functions alone as there may be more going on there (arguably
these really have no business passing around Perl thread context).

12 years agoUpdate ExtUtils-CBuilder to CPAN version 0.2802
Chris 'BinGOs' Williams [Sun, 12 Dec 2010 22:08:49 +0000 (22:08 +0000)]
Update ExtUtils-CBuilder to CPAN version 0.2802


  0.2802 - Sun Dec 12 07:22:43 EST 2010


    - Incorporated another t/04-base.t fix from bleadperl
      [Chris Williams]

12 years agoUpdate Digest-SHA to CPAN version 5.49
Chris 'BinGOs' Williams [Sun, 12 Dec 2010 21:58:41 +0000 (21:58 +0000)]
Update Digest-SHA to CPAN version 5.49


  5.49  Sun Dec 12 07:22:04 MST 2010
        - modified Addfile to accept all POSIX filenames
                -- standard allows all characters except NUL and '/'
        - updated shasum to more closely mimic sha1sum/md5sum
                -- added "backslash processing" to handle newlines
                        and backslashes in filenames
                -- now accepts all POSIX filenames via Addfile
                -- thanks to Sean Burke for identifying edge cases

12 years agoMake entertry a LOGOP, not BASEOP/UNOP
Reini Urban [Sun, 12 Dec 2010 15:20:33 +0000 (16:20 +0100)]
Make entertry a LOGOP, not BASEOP/UNOP

The initial parse-time UNOP is upgraded at ck_eval to an LOGOP
- op_other holding the ptr to leavetry -
which causes problems at run-time B optype inspection.
The opclass info was always wrong, the Bytecode compiler broke with 5.12.

12 years agoblead breaks Attribute::Constant
Karl Williamson [Sun, 12 Dec 2010 18:28:20 +0000 (11:28 -0700)]
blead breaks Attribute::Constant

The problem is that I confused FOLD with ANYOF_FOLD, and as a result,
emitted a locale regnode, which is tainted.  Any tests that required
non-tainting started failing

12 years agorename some grammar productions for clarity
Zefram [Sat, 11 Dec 2010 20:38:16 +0000 (20:38 +0000)]
rename some grammar productions for clarity

Some of the names for types of expression were misleading.
Rename "listexpr"->"optlistexpr", "listexprcom"->"optexpr", and
"argexpr"->"listexpr" to make them more accurately descriptive and
more consistent.

12 years agoFix test count in t/re/pat.t
Father Chrysostomos [Sun, 12 Dec 2010 07:31:44 +0000 (23:31 -0800)]
Fix test count in t/re/pat.t

12 years ago[perl #76026] match variables persist between calls to a sort sub
Father Chrysostomos [Sun, 12 Dec 2010 02:50:49 +0000 (18:50 -0800)]
[perl #76026] match variables persist between calls to a sort sub

Since, for speed’s sake, pp_sort does not call PUSH/POPBLOCK for
every invocation of a sort subroutine, it fails to restore PL_curpm
after each call (POPBLOCK usually handles that). So the new values of
match vars like $1 when the sub returns are what it sees at the next

This commit fixes this by resetting PL_curpm after each call to the
subroutine. There are actually three different functions for this
(S_sortcv*) so they all need modification.

12 years agoperldelta for [perl #68712]
Father Chrysostomos [Sun, 12 Dec 2010 00:28:59 +0000 (16:28 -0800)]
perldelta for [perl #68712]

12 years agorecursive-descent expression parsing
Zefram [Sat, 11 Dec 2010 01:31:03 +0000 (01:31 +0000)]
recursive-descent expression parsing

New API functions parse_fullexpr(), parse_listexpr(), parse_termexpr(),
and parse_arithexpr(), to parse an expression at various precedence

12 years agore/pat.t: Add todo test for #38133
Karl Williamson [Sun, 12 Dec 2010 00:20:18 +0000 (17:20 -0700)]
re/pat.t: Add todo test for #38133

12 years agoperldelta: mention need to update unicode pods
Karl Williamson [Thu, 9 Dec 2010 03:54:26 +0000 (20:54 -0700)]
perldelta: mention need to update unicode pods

12 years agounicode_strings.t: Revise tests for EBCDIC, clarity
Karl Williamson [Thu, 9 Dec 2010 03:31:43 +0000 (20:31 -0700)]
unicode_strings.t: Revise tests for EBCDIC, clarity

This patch addresses some of Hugo's concerns.

12 years agoregcomp.c: Clean up optimization for 1-char []
Karl Williamson [Wed, 8 Dec 2010 05:51:34 +0000 (22:51 -0700)]
regcomp.c: Clean up optimization for 1-char []

A single character character class can be optimized into an EXACT node.
The changes elsewhere allow this to no longer be constrained to
ASCII-only when the pattern isn't UTF-8.  Also, the optimization
shouldn't have happened for FOLDED characters, as explained in the
comments, when they participate in multi-char folds; so that is removed.

Also, a locale node with folded characters can be optimized.

12 years agoregcomp: Allow freeing up bit in ANYOF flags
Karl Williamson [Wed, 8 Dec 2010 00:04:02 +0000 (17:04 -0700)]
regcomp: Allow freeing up bit in ANYOF flags

The flags field is fully used, and until the ANYOF node is split later
in development, the CLASS bit will need to be freed up to give the space
for other uses.  This patch allows for this to easily be toggled.