perl.git
10 months agoPOSIX.xs - support C99 math for mingw
sisyphus [Thu, 25 Oct 2018 08:26:29 +0000 (19:26 +1100)] 
POSIX.xs - support C99 math for mingw

10 months agoperl.h - support C99 math for mingw
sisyphus [Thu, 25 Oct 2018 08:25:37 +0000 (19:25 +1100)] 
perl.h - support C99 math for mingw

10 months agoCall a hash %hash in the documentation
E. Choroba [Wed, 5 Dec 2018 22:49:57 +0000 (23:49 +0100)] 
Call a hash %hash in the documentation

Previously, %array was used which was confusing.

10 months ago(perl #133708) remove build-time probing for stack limits for Storable
Tony Cook [Tue, 4 Dec 2018 04:11:17 +0000 (15:11 +1100)] 
(perl #133708) remove build-time probing for stack limits for Storable

10 months agohandy.h: Fix definition of isPOWER_OF_2()
Karl Williamson [Sat, 8 Dec 2018 21:45:05 +0000 (14:45 -0700)] 
handy.h: Fix definition of isPOWER_OF_2()

Prior to this commit, it returned TRUE if the input was 0.

10 months agoregcomp.h: Fix typo in comment
Karl Williamson [Sat, 8 Dec 2018 21:46:48 +0000 (14:46 -0700)] 
regcomp.h: Fix typo in comment

10 months agoregexec.c: Use mnemonics instead of "256"
Karl Williamson [Thu, 6 Dec 2018 17:02:55 +0000 (10:02 -0700)] 
regexec.c: Use mnemonics instead of "256"

There are only three valid numbers in Computer Science: 0, 1, and "as
many as you like".  256 is not therefore a valid number, and its use
should be commented, or better, a mnemonic used instead.

These uses were spotted by lgtm as being unnecessary, and some people
were confused as to their purpose.  This commit changes the 256, to in
one case, a sizeof() that indicates its a guard against going outside
the array bounds of that data structure.  And in the 2nd case, it
verifies that it fits inside the space allotted to a function parameter.
A comment helps clarify that.

10 months agoregcomp.c: Allow more EXACTFish nodes to be trieable
Karl Williamson [Tue, 4 Dec 2018 16:58:13 +0000 (09:58 -0700)] 
regcomp.c: Allow more EXACTFish nodes to be trieable

The previous two commits fixed bugs where it would be possible during
optimization to join two EXACTFish nodes together, and the result would
not work properly with LATIN SMALL LETTER SHARP S.  But by doing so,
the commits caused all non-UTF-8 EXACTFU nodes that begin or end with
[Ss] from being trieable.

This commit changes things so that the only the ones that are
non-trieable are the ones that, when joined, have the sequence [Ss][Ss]
in them.  To do so, I created three new node types that indicate if the
node begins with [Ss] or ends with them, or both.  These preclude having
to examine the node contents at joining to determine this.  And since
there are plenty of node types available, it seemed the best choice.
But other options would be available should we run out of nodes.
Examining the first and final characters of a node is not expensive, for
example.

10 months agoregcomp.c: Make sure /di nodes begining in 's' are EXACTF
Karl Williamson [Fri, 30 Nov 2018 16:31:46 +0000 (09:31 -0700)] 
regcomp.c: Make sure /di nodes begining in 's' are EXACTF

This is defensive coding.  The previous commit changed things so under
/di a node ending in [Ss] doesn't get made an EXACTFU.  This commit does
the same for nodes that begin with [Ss].  This isn't actually necessary
as one needs two EXACTFU nodes in a row for the problem to occur, and
the previous commit appears to remove the possibility for the first node
being an EXACTFU.  But I'm leery of relying on this.  So this commit
makes sure that a node beginning with 'S' or 's' under /di remains
EXACTF

10 months agoregcomp.c: Make sure /di nodes ending in 's' are EXACTF
Karl Williamson [Fri, 30 Nov 2018 03:43:32 +0000 (20:43 -0700)] 
regcomp.c: Make sure /di nodes ending in 's' are EXACTF

Prior to this commit only nodes that filled fully were guaranteed not to
be upgraded to EXACTFU.

EXACTF nodes are used when /d rules are to be used unless the string
being matched against is in UTF-8.  EXACTFU nodes are used when the /u
rules are to be used always.  If the node contains only characters whose
interpretation is the same under /d and /u, both node types behave
identically, and so either can be used.  EXACTFU nodes are preferred
because they are trie-able, and otherwise faster at runtime due to not
having to check the UTF-8ness of the target string.  The pattern
compilation uses an EXACTFU node when possible, over an EXACTF node.

The sequences 'ss', 'SS', 'Ss', 'sS'  are very tricky, for several
reasons.  For this commit, the trickiness lies in the fact that they are
the only sequences where both the pattern and target string aren't in
UTF-8, and what matches under /ui rules differs from what matches under
/di.  (There are various single characters that match differently, but
this is the only sequence whose individual components match the same,
but the sequence as a whole matches differently.)

The code has long taken special care for these sequences, but overlooked
two fairly obscure cases where it matters.  This commit addresses one of
those cases; the next commit, the other.

Because these  are sequences, it might be possible for it to be split
across two EXACTFish nodes, so that the first 's' or 'S' is the last
thing in the first node, and the second 's' or 'S' is the first thing in
the second.  Should these nodes get joined during optimization, they
form the sequence.  The code has long recognized this possibility if the
first node gets filled up, necessitating a split, and it doesn't make
the first node EXACTFU if it ends in 's' or 'S'.  But we don't fill
those nodes totally completely, and optimization can join two together.

Future commits in the pipeline will join nodes in more cases during
optimization, and so, we need to not create an EXACTFU for trailing 's'
or 'S' always, not just if the first node fills up.  This commit moves
the code that accomplishes this so it always gets executed at node end
of /di nodes, instead of previously only getting executed when the node
fills.

10 months agoregcomp.c: Simplify a bit of code
Karl Williamson [Mon, 3 Dec 2018 00:39:05 +0000 (17:39 -0700)] 
regcomp.c: Simplify a bit of code

By using a macro with a slightly different API, we don't have to mess
with the parse pointer.

10 months agoregcomp.c: Can join certain EXACTish node types
Karl Williamson [Mon, 3 Dec 2018 18:33:49 +0000 (11:33 -0700)] 
regcomp.c: Can join certain EXACTish node types

The optimization phase of regular expression pattern compilation looks
for adjacent EXACTish nodes and joins them if they are the same flavor
of EXACT.  Commits a9f8c7ac75c364c3e05305718f38c5f8ccd935d8 and
f6b4b99d2e584fbcd85eeed475eea10b87858e54 introduced two new nodes
that are so close to existing flavors that they are joinable with their
respective flavor.  This commit does that.

10 months agoregcomp.c: Move clause of while() conditional into loop
Karl Williamson [Mon, 3 Dec 2018 00:45:06 +0000 (17:45 -0700)] 
regcomp.c: Move clause of while() conditional into loop

This is in preparation for making the conditional more complicated than
can be easily done in the condition.

10 months agoregcomp.c: Add assertion
Karl Williamson [Sun, 2 Dec 2018 19:27:46 +0000 (12:27 -0700)] 
regcomp.c: Add assertion

10 months agoRemove one use of static function
Karl Williamson [Sun, 2 Dec 2018 19:38:27 +0000 (12:38 -0700)] 
Remove one use of static function

Previous commits in 5.29 have removed all but two calls to this
function, and the remaining ones take radically different paths in it,
with very little common code.  It simplifies things if we expand each
call to the code that gets evaluated.  This commit does one call.
Changing the other usage is deferred until a later commit.

The need for an #ifdef is removed by adding a flag and setting it in an
existing #ifdef.

10 months agoNo need to convert int to long to int
H.Merijn Brand [Fri, 7 Dec 2018 13:43:37 +0000 (14:43 +0100)] 
No need to convert int to long to int

Beyond that, I agree with this branch

Signed-off-by: James E Keenan <jkeenan@cpan.org>
10 months agoNo need to include 'values.h'.
James E Keenan [Wed, 28 Nov 2018 03:15:22 +0000 (22:15 -0500)] 
No need to include 'values.h'.

timecheck.c does not appear to use any thing from this file.  Linux
documentation marks the interface as obsolete.  It's not found, e.g., on
FreeBSD.

10 months agoEliminate 4 build-time warnings in timecheck.c.
James E Keenan [Tue, 27 Nov 2018 21:12:45 +0000 (16:12 -0500)] 
Eliminate 4 build-time warnings in timecheck.c.

This commit suppresses output like this:

[Porting] 758 $ cc -O -o timecheck timecheck.c
timecheck.c: In function ‘gm_check’:
timecheck.c:37:36: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 5 has type ‘int’ [-Wformat=]
      fprintf (stderr, "%3d:%s: %12ld-%02d-%02d %02d:%02d:%02d\n",
                                ~~~~^
                                %12d
timecheck.c:39:3:
   tmp->tm_year + 1900, tmp->tm_mon + 1, tmp->tm_mday,
   ~~~~~~~~~~~~~~~~~~~
timecheck.c: In function ‘lt_check’:
timecheck.c:89:36: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 5 has type ‘int’ [-Wformat=]
      fprintf (stderr, "%3d:%s: %12ld-%02d-%02d %02d:%02d:%02d\n",
                                ~~~~^
                                %12d
timecheck.c:91:3:
   tmp->tm_year + 1900, tmp->tm_mon + 1, tmp->tm_mday,
   ~~~~~~~~~~~~~~~~~~~
timecheck.c: In function ‘main’:
timecheck.c:130:21: warning: implicit declaration of function ‘strcmp’ [-Wimplicit-function-declaration]
     if (argc > 1 && strcmp (argv[1], "-v") == 0) opt_v++;
                     ^~~~~~
timecheck.c:139:32: warning: format ‘%ld’ expects argument of type ‘long int’, but argument 2 has type ‘int’ [-Wformat=]
     printf ("Sizeof time_t = %ld\n", (i = sizeof (time_t)));
                              ~~^     ~~~~~~~~~~~~~~~~~~~~~
                              %d

10 months agoUse more plausible argument for 'cd' in example.
James E Keenan [Tue, 27 Nov 2018 21:07:35 +0000 (16:07 -0500)] 
Use more plausible argument for 'cd' in example.

In our documentation we generally assume that the user is navigating
from the top-level of the core distribution.  If so, then to compile a C
program found in Porting/ the user would simply 'cd Porting' rather than
'cd perl/Porting'.

10 months agoregen/mk_invlists.pl: Add new table
Karl Williamson [Thu, 6 Dec 2018 23:57:17 +0000 (16:57 -0700)] 
regen/mk_invlists.pl: Add new table

This table contains all the code points that are in any multi-character
fold (not the folded-from character, but what that character folds to).

It will be used in a future commit.

10 months agoregen/mk_invlists.pl: Rmv no longer used array
Karl Williamson [Thu, 6 Dec 2018 23:53:23 +0000 (16:53 -0700)] 
regen/mk_invlists.pl: Rmv no longer used array

10 months agot/harness: Catch incorrect serial directory specification
Karl Williamson [Fri, 7 Dec 2018 17:42:21 +0000 (10:42 -0700)] 
t/harness: Catch incorrect serial directory specification

The previous commit fixed a bug.  This commit detects if someone creates
a new instance of that bug.

10 months agot/harness: Actually run IO::Zlib tests sequentially
Karl Williamson [Fri, 7 Dec 2018 17:32:35 +0000 (10:32 -0700)] 
t/harness: Actually run IO::Zlib tests sequentially

Commit bc55cddf915d70f806feda38506788f4c61a2574 tried to do this, but it
turns out it didn't actually, because it didn't precisely name the
directory.

10 months agoMake ext/B/t/strict.t test alike to other ones.
Petr Písař [Fri, 7 Dec 2018 14:25:53 +0000 (09:25 -0500)] 
Make ext/B/t/strict.t test alike to other ones.

Signed-off-by: Petr Písař <ppisar@redhat.com>
For: RT 133713

10 months agoCorrect error message in t/taint.t.
James E Keenan [Fri, 7 Dec 2018 02:42:51 +0000 (02:42 +0000)] 
Correct error message in t/taint.t.

In this test file, our temporary directory is named 'for_find_taint'.
Make sure the error method reflects that.  (This was overlooked in
f09dd912ca4 2018-11-19.)

10 months agoregcomp.c: Use simpler variable name as long as possible
Karl Williamson [Sun, 2 Dec 2018 20:05:47 +0000 (13:05 -0700)] 
regcomp.c: Use simpler variable name as long as possible

This just extends the use of a variable name a little longer, as it's
easier to read than the nested macro calls that eventually have to be
used.

10 months agoregcomp.c: Prefer one of similarly named vars
Karl Williamson [Sun, 2 Dec 2018 19:22:39 +0000 (12:22 -0700)] 
regcomp.c: Prefer one of similarly named vars

These two variables are similarly named, but have slightly different
purposes.  Comment out one of them, and convert to using the other
consistently

10 months agot/re/anyof.t: Remove duplicate test case
Karl Williamson [Sun, 2 Dec 2018 18:46:07 +0000 (11:46 -0700)] 
t/re/anyof.t: Remove duplicate test case

10 months agoUse consistent spelling in qr// dumping
Karl Williamson [Sun, 2 Dec 2018 18:17:26 +0000 (11:17 -0700)] 
Use consistent spelling in qr// dumping

Under -Dr (or use re 'Debug') the compiled regex engine program is
displayed.  I noticed that it used two different spellings for
'infinity'.  This commit changes so only one is used, the one that has
been in the field the longest.

10 months agoregcomp.c: Clarify comment
Karl Williamson [Wed, 5 Dec 2018 01:22:05 +0000 (18:22 -0700)] 
regcomp.c: Clarify comment

10 months agoFix t/porting/manifest.t failures when run in a foreign git checkout
Niko Tyni [Sun, 2 Dec 2018 09:57:01 +0000 (11:57 +0200)] 
Fix t/porting/manifest.t failures when run in a foreign git checkout

The change at ba6733216202523a95b0b7ee2e534b8e30b6d7df didn't work
correctly: find_git_or_skip() in t/test.pl looks at PERL_BUILD_PACKAGING
too late, so it doesn't change the logic of skipping, only the explanation
of why tests get skipped if they do. This was originally reported at
<https://bugs.debian.org/914962>.

Committer: update commit message for upstream context

10 months agoregcomp.sym: Clarify descriptions of EXACTish regnodes
Karl Williamson [Wed, 5 Dec 2018 00:59:31 +0000 (17:59 -0700)] 
regcomp.sym: Clarify descriptions of EXACTish regnodes

10 months agoinline.h: Clarify comment
Karl Williamson [Tue, 4 Dec 2018 19:05:59 +0000 (12:05 -0700)] 
inline.h: Clarify comment

This function works as well as possible on non-UTF-8 inputs.  No need to
list the UTF-8-only failures, since those aren't legal inputs.

10 months agobump $ExtUtils::ExtUtils::VERSION to 3.40
Tony Cook [Wed, 5 Dec 2018 04:35:19 +0000 (15:35 +1100)] 
bump $ExtUtils::ExtUtils::VERSION to 3.40

10 months ago(perl #133654) don't include OUTLIST parameters in the prototype
Tony Cook [Wed, 5 Dec 2018 04:22:52 +0000 (15:22 +1100)] 
(perl #133654) don't include OUTLIST parameters in the prototype

The generated prototype (with PROTOTYPES: ENABLE) would include
OUTLIST parameters, but these aren't arguments to the perl function.

10 months ago(perl #133706) remove exploit code from Storable
Tony Cook [Mon, 3 Dec 2018 05:15:52 +0000 (16:15 +1100)] 
(perl #133706) remove exploit code from Storable

Storable packaged the metasploit framework code for CVE-2015-1592,
which triggered virus scanners.

To prevent that remove the packaged exploit code and test for the
underlying structure we trigger the warning on.

10 months agoAdd header-guards to 2 additional files
James E Keenan [Tue, 4 Dec 2018 00:25:51 +0000 (19:25 -0500)] 
Add header-guards to 2 additional files

Two header files in ext/SDBM_File appear to be blead-upstream, so we can
add header guards here as well.

For: RT 133699

10 months agoProvide header guards to prevent re-inclusion
James E Keenan [Thu, 29 Nov 2018 03:50:29 +0000 (22:50 -0500)] 
Provide header guards to prevent re-inclusion

Per LGTM analysis: https://lgtm.com/projects/g/Perl/perl5/alerts/?mode=tree&ruleFocus=2163210746

and LGTM recommendation: https://lgtm.com/rules/2163210746/

For: RT 133699

10 months agoRevert "regcomp.c: Use a weird value in a place where ignored"
Karl Williamson [Sun, 2 Dec 2018 08:30:39 +0000 (01:30 -0700)] 
Revert "regcomp.c: Use a weird value in a place where ignored"

This reverts commit 5a4bc50b395654abde21718cbce2e296049f470d.
It turns out I was wrong, and the value isn't ignored.  The commit
caused the pattern to be optimized differently, but still be valid,
so no tests failed.

10 months agoSync Test::Simple with CPAN version 1.302141
James E Keenan [Sat, 1 Dec 2018 17:58:41 +0000 (12:58 -0500)] 
Sync Test::Simple with CPAN version 1.302141

Committer:  Update perldelta

10 months agoAdd instruction to configure to SYNOPSIS
James E Keenan [Sat, 1 Dec 2018 17:39:27 +0000 (12:39 -0500)] 
Add instruction to configure to SYNOPSIS

Because I always to forget to ./Configure before trying to bring in an
updated upstream module from CPAN.

10 months agoCorrect typo
James E Keenan [Fri, 30 Nov 2018 03:12:10 +0000 (22:12 -0500)] 
Correct typo

10 months agoDrop sentence which is no longer valid as of 5.28
James E Keenan [Fri, 30 Nov 2018 03:09:41 +0000 (22:09 -0500)] 
Drop sentence which is no longer valid as of 5.28

Per recommendation of Dan Book.

For: RT 133705

10 months agoTick off 5.26.3 and 5.28.1
Steve Hay [Thu, 29 Nov 2018 21:06:04 +0000 (21:06 +0000)] 
Tick off 5.26.3 and 5.28.1

10 months agoperlhist for 5.26.3 and 5.28.1
Steve Hay [Thu, 29 Nov 2018 21:03:00 +0000 (21:03 +0000)] 
perlhist for 5.26.3 and 5.28.1

These were applied to maint-5.26 and maint-5.28 in the following commits:

maint-5.26:
b9de8c10f90b3a2559572b35083d7004790ff6aa
1c6b38711fd6e33e1305d39081a7ed6977e2dadb

maint-5.28:
d7dd16fb7dd9064a236e9430754052ab496e3b4c
63afdf6c0f65af480aa5bb9ccba9f46dae52f6fc

10 months agoImport perl5263delta.pod and perl5281delta.pod
Steve Hay [Thu, 29 Nov 2018 20:49:00 +0000 (20:49 +0000)] 
Import perl5263delta.pod and perl5281delta.pod

10 months agoUpdate Module-CoreList with data for 5.26.3 and 5.28.1
Steve Hay [Thu, 29 Nov 2018 20:11:17 +0000 (20:11 +0000)] 
Update Module-CoreList with data for 5.26.3 and 5.28.1

10 months agoEpigraphs for 5.26.3 and 5.28.1
Steve Hay [Thu, 29 Nov 2018 19:49:30 +0000 (19:49 +0000)] 
Epigraphs for 5.26.3 and 5.28.1

10 months agoPATCH: [perl #133423]
Karl Williamson [Mon, 24 Sep 2018 17:16:14 +0000 (11:16 -0600)] 
PATCH: [perl #133423]

10 months agolocale.c: Failure to build if not allowing LC_COLLATE
Karl Williamson [Thu, 29 Nov 2018 18:55:09 +0000 (11:55 -0700)] 
locale.c: Failure to build if not allowing LC_COLLATE

This is part of [perl #133696].  A typo was causing a macro to be
defined in terms of itself, hence an illegal recursive definition.

10 months agolocale.c: Don't use numeric unless LC_NUMERIC
Karl Williamson [Thu, 29 Nov 2018 18:53:58 +0000 (11:53 -0700)] 
locale.c: Don't use numeric unless LC_NUMERIC

This commit #ifdef's a usage of a variable that isn't valid unless the
system has LC_NUMERIC

10 months agolocale.c: Fix wrong scope of #if's
Karl Williamson [Thu, 29 Nov 2018 18:50:58 +0000 (11:50 -0700)] 
locale.c: Fix wrong scope of #if's

The function print_bytes_for_locale() should be defined if DEBUGGING;
prior to this commit it didn't get defined unless LC_COLLATE was
defined on the platform.

10 months agoMore removals of $a,$b in perldata for [#perl #133700]
Eugen Konkov [Thu, 29 Nov 2018 17:56:07 +0000 (10:56 -0700)] 
More removals of $a,$b in perldata for [#perl #133700]

10 months agoPATCH: [perl #133700] avoid use of $a and $b in perldata
Eugen Konkov [Thu, 29 Nov 2018 17:21:20 +0000 (10:21 -0700)] 
PATCH: [perl #133700] avoid use of $a and $b in perldata

10 months agomake boot_Win32CORE extern "C" for C++ builds tonyc/cygwin-fixes
Tony Cook [Thu, 29 Nov 2018 03:16:03 +0000 (14:16 +1100)] 
make boot_Win32CORE extern "C" for C++ builds

10 months agostdio.h on Cygwin doesn't expose cuserid() with _GNU_SOURCE
Tony Cook [Thu, 29 Nov 2018 03:12:01 +0000 (14:12 +1100)] 
stdio.h on Cygwin doesn't expose cuserid() with _GNU_SOURCE

It's probably possible to expose it by setting _XOPEN_SOURCE to
some specific value, but this appears to be a bug.

https://cygwin.com/ml/cygwin/2018-11/msg00230.html

10 months agoPass a UV to a format expecting a UV
Tony Cook [Wed, 28 Nov 2018 23:50:19 +0000 (23:50 +0000)] 
Pass a UV to a format expecting a UV

MAX_LEGAL_CP can end up as int depending on the ranges of the types
involved, causing a type mismatch on the format in cp_above_legal_max.

By adding the cast to the macro definition we both prevent the type
mismatch on the format, but also may allow some static analysis tool to
detect comparisons against signed types, which is likely an error.

10 months agoUpgrade to threads::shared 1.59
jdhedden [Wed, 28 Nov 2018 03:53:40 +0000 (22:53 -0500)] 
Upgrade to threads::shared 1.59

Committer: perldelta entry

10 months agoAdd USE_THREAD_SAFE_LOCALE to non-bin-compat options list
Karl Williamson [Tue, 27 Nov 2018 17:07:12 +0000 (10:07 -0700)] 
Add USE_THREAD_SAFE_LOCALE to non-bin-compat options list

Spotted by Tux

10 months agoregcomp.c: White-space only
Karl Williamson [Tue, 27 Nov 2018 16:43:01 +0000 (09:43 -0700)] 
regcomp.c: White-space only

Vertical align continuation chars in a macro

10 months agoAdd regnode EXACTFU_ONLY8
Karl Williamson [Tue, 27 Nov 2018 16:42:45 +0000 (09:42 -0700)] 
Add regnode EXACTFU_ONLY8

This is a regnode that otherwise would be an EXACTFU except that it
contains a code point that requires UTF-8 to match, including all the
possible folds involving it.  Hence if the target string isn't UTF-8, we
know it can't possibly match, without needing to try.

For completeness, there could also be an EXACTFAA_ONLY8 and an
EXACTFL_ONLY8 created, but I think these are unlikely to actually appear
in the wild, since using /aa is mainly about ASCII, and /l mostly will
involve characters that don't require UTF-8.

10 months agoAdd regnode EXACT_ONLY8
Karl Williamson [Sat, 17 Nov 2018 22:51:19 +0000 (15:51 -0700)] 
Add regnode EXACT_ONLY8

This is a regnode that otherwise would be an EXACT except that it
contains a code point that requires UTF-8 to represent.  Hence if the
target string isn't UTF-8, we know it can't possibly match, without
needing to try.

10 months agoregcomp.c: Use common code instead of duplicating it
Karl Williamson [Sat, 17 Nov 2018 22:47:02 +0000 (15:47 -0700)] 
regcomp.c: Use common code instead of duplicating it

The common code is about to get more complicated, so use it instead of a
copy.

10 months agoperlreref.pod: disambiguate "code"
David Mitchell [Tue, 27 Nov 2018 16:36:35 +0000 (16:36 +0000)] 
perlreref.pod: disambiguate "code"

It says:

   (?(cond)yes)      Conditional expression, where "cond" can be:
                     (?=pat)   lookahead
                     ...

A strict reading of that is that there must be two pairs of parens
in each conditional construct, e.g. (?((?=pat))yes).

Make the text clearer.

10 months agohandle /(?(?{code}))/ mixed compile-and runtime
David Mitchell [Tue, 27 Nov 2018 13:26:39 +0000 (13:26 +0000)] 
handle /(?(?{code}))/ mixed compile-and runtime

Where a runtime pattern contains both compile-time and run-time code
blocks, e.g.:

    $re = '(?{ RRR })';
    / $re X(?{ CCC })Y/

The compile-time code-block CCC is parsed at the same time as the
surrounding text. The runtime code RRR is parsed at runtime by
constructing a fake pattern and re-parsing it, but with any compile-time
code-blocks blanked out (so they don't get compiled twice). The compiled
regex is then thrown away, but any optrees just created for the runtime
code blocks are kept.

For example at runtime, the re-parsed pattern looks like:

    / (?{ RRR }) X__________Y/

Unfortunately this was failing for the conditional pattern, e.g.

    / $re X(?(?{ CCC }))Y/

which was getting blanked as

    / (?{ RRR }) X(?_______)Y/

which isn't valid syntax.

This commit blanks (?{...}) into (?=====) instead which is always legal.

10 months agoRely on C89 "const"
Aaron Crane [Tue, 27 Nov 2018 11:32:11 +0000 (11:32 +0000)] 
Rely on C89 "const"

The metaconfig probe for <db.h> previously relied on the d_const symbol set
by the "const" probe, so generating Configure here has been done against
metaconfig commit 1204d4627a06b11f16620188f3fa83159ed35fd9 which changes
that.

Thanks to khw++ for pointing out this oversight in my attempt last year to
make the codebase rely on C89.

10 months agoregexec.c: Use ANYOF bitmap lookup in more cases
Karl Williamson [Sun, 18 Nov 2018 22:46:07 +0000 (15:46 -0700)] 
regexec.c: Use ANYOF bitmap lookup in more cases

ANYOFish nodes have a bitmap.  If we know the value is in the bitmap
range, then flags that apply to out-of-range values are irrelevant.
Other flags being set indicate that the desired answer is more
complicated than just using a bitmap lookup.  But exclude this
irrelevant flag from that calculation when we know the value is in the
bitmap.

There are other flags that it is possible to exclude, but not without
further conditionals, or unsharing code, and are either rarely set or
are for node types that we don't worry so much about optimal
performance, like /l and /d.  The changes introduced by this commit
are determined at .c compile time except for a runtime mask, and hence
don't introduce new branches that may destroy the instruction cache
pipeline.

10 months agoregexec.c: Refactor expanded macro from prev. commit
Karl Williamson [Sun, 18 Nov 2018 22:24:45 +0000 (15:24 -0700)] 
regexec.c: Refactor expanded macro from prev. commit

10 months agoregexec.c: Expand out macro in only remaining use
Karl Williamson [Sun, 18 Nov 2018 22:12:02 +0000 (15:12 -0700)] 
regexec.c: Expand out macro in only remaining use

There is only one call to this macro.  It's easier to understand if
expanded out instead of the call.  The next commit will refactor.

10 months agoregcomp.c: Clarify comment
Karl Williamson [Wed, 21 Nov 2018 03:59:18 +0000 (20:59 -0700)] 
regcomp.c: Clarify comment

10 months agoregcomp.c: Initialize a variable more conservatively
Karl Williamson [Sun, 18 Nov 2018 21:52:15 +0000 (14:52 -0700)] 
regcomp.c: Initialize a variable more conservatively

It doesn't matter currently, but thes variable shouldn't be TRUE unless
/i is in effect.

10 months agoregcomp.c: Use a weird value in a place where ignored
Karl Williamson [Sun, 18 Nov 2018 20:39:49 +0000 (13:39 -0700)] 
regcomp.c: Use a weird value in a place where ignored

This way, it doesn't confuse that it is legal, and should it stop being
ignored in the called function, it will show up as a problem much
sooner.

10 months agoregexec.c: Add comment
Karl Williamson [Sat, 17 Nov 2018 23:02:31 +0000 (16:02 -0700)] 
regexec.c: Add comment

10 months agoregexec.c: Rmv unused macros
Karl Williamson [Sat, 17 Nov 2018 22:54:34 +0000 (15:54 -0700)] 
regexec.c: Rmv unused macros

10 months agoregcomp.h: Clarify comments
Karl Williamson [Sun, 18 Nov 2018 23:06:50 +0000 (16:06 -0700)] 
regcomp.h: Clarify comments

10 months agoregcomp.c: Consolidate duplicated code into 1 place
Karl Williamson [Thu, 15 Nov 2018 18:09:35 +0000 (11:09 -0700)] 
regcomp.c: Consolidate duplicated code into 1 place

10 months agoregcomp.c: Use better method for setting debug offsets
Karl Williamson [Wed, 21 Nov 2018 04:21:44 +0000 (21:21 -0700)] 
regcomp.c: Use better method for setting debug offsets

This was changing the parse pointer around some code and restoring it
afterwards.  The purpose must have been to get the debug offsets
correct.  But there is a better way to do that, which doesn't take up
cycles on a non-Debugging build, and that is to set the offsets
directly.

10 months agoregcomp.c: Remove another sizing pass relict
Karl Williamson [Wed, 21 Nov 2018 04:18:51 +0000 (21:18 -0700)] 
regcomp.c: Remove another sizing pass relict

We no longer play with the emit ptr in this function, so no need to save
and restore it.

10 months agoMove isPOWER_OF_2() macro to handy.h
Karl Williamson [Sun, 25 Nov 2018 19:14:49 +0000 (12:14 -0700)] 
Move isPOWER_OF_2() macro to handy.h

This is in preparation for it to be used outside of the file which
previously defined it.

10 months agoregen/mk_invlists.pl: Generate a new value
Karl Williamson [Sun, 25 Nov 2018 19:12:19 +0000 (12:12 -0700)] 
regen/mk_invlists.pl: Generate a new value

The new value is the maximum number of code points that fold to any
single code point.  It will be used in a future commit.

10 months agoRename local variable to prevent confusion with global
James E Keenan [Mon, 26 Nov 2018 17:29:03 +0000 (12:29 -0500)] 
Rename local variable to prevent confusion with global

Per:  https://lgtm.com/projects/g/Perl/perl5/alerts/?mode=tree&ruleFocus=2157860312

For: RT # 133686 (partial)

10 months agoRename global variable to prevent confusion with local
James E Keenan [Mon, 26 Nov 2018 17:14:51 +0000 (12:14 -0500)] 
Rename global variable to prevent confusion with local

Per:  https://lgtm.com/projects/g/Perl/perl5/alerts/?mode=tree&ruleFocus=2157860312

10 months agoperldelta for 404395d24bc8, 640e129d0fc4 and 85d2f7cacba4
Tony Cook [Mon, 26 Nov 2018 03:21:13 +0000 (14:21 +1100)] 
perldelta for 404395d24bc8640e129d0fc4 and 85d2f7cacba4

10 months ago(perl #133659) make an in-place edit successful if the exit status is zero
Tony Cook [Tue, 20 Nov 2018 23:05:27 +0000 (10:05 +1100)] 
(perl #133659) make an in-place edit successful if the exit status is zero

during global destruction.

This means that code like:

  perl -i -ne '...; last'

will replace the input file with the in-place edit output of the file,
but:

  perl -i -ne '...; die'

or

  perl -i -ne '...; exit 1'

won't.

10 months ago(perl #133659) tests for global destruction handling of inplace editing
Tony Cook [Tue, 20 Nov 2018 05:43:43 +0000 (16:43 +1100)] 
(perl #133659) tests for global destruction handling of inplace editing

10 months ago(perl #133659) move argvout cleanup to a new function
Tony Cook [Tue, 20 Nov 2018 04:30:20 +0000 (15:30 +1100)] 
(perl #133659) move argvout cleanup to a new function

10 months agoRemove 1 comparison whose result is always the same.
James E Keenan [Fri, 23 Nov 2018 16:49:03 +0000 (11:49 -0500)] 
Remove 1 comparison whose result is always the same.

Per: https://lgtm.com/projects/g/Perl/perl5/alerts/?mode=tree&ruleFocus=2154840804

For: RT 133686 (partial)

10 months agoEliminate empty conditional branch
James E Keenan [Fri, 23 Nov 2018 18:45:25 +0000 (13:45 -0500)] 
Eliminate empty conditional branch

Per:  https://lgtm.com/projects/g/Perl/perl5/alerts/?mode=tree&ruleFocus=2154840803

For: RT 133686 (partial)

10 months agoAltaVista is no more.
James E Keenan [Sat, 24 Nov 2018 00:17:38 +0000 (19:17 -0500)] 
AltaVista is no more.

Provide a non-Mountain View alternative.

For: RT # 133684

10 months agoSplit NAME line on multiple whitespaces
James E Keenan [Fri, 23 Nov 2018 22:13:59 +0000 (17:13 -0500)] 
Split NAME line on multiple whitespaces

For:  RT # 133683

pod/perlmodlib.pod is a file generated by pod/perlmodlib.PL, which is
run by 'miniperl' during 'make'.  That program parses the 'NAME' header
of .pod files and fragments of POD found in 'regen/opcode.pl'.  The POD
for B::Op_private is one such fragment.  Correcting a superfluous
whitespace in that fragment did not suffice to prevent the downstream
formatting error reported in the RT -- an error visible with 'pod2text'
and 'pod2html' as well.  We also had to make the regex which
perlmodlib.PL uses to parse the 'NAME' header more flexible.

10 months agoperlfunc: clarify reset EXPR behaviour
Dagfinn Ilmari Mannsåker [Thu, 22 Nov 2018 12:14:38 +0000 (12:14 +0000)] 
perlfunc: clarify reset EXPR behaviour

Mention that it only affects variables in the current package, and
that it affects scalars, arrays and hashes ("variables and arrays"
is just confus(ed|ing)).

10 months agooops, typo
Karen Etheridge [Wed, 21 Nov 2018 17:40:19 +0000 (09:40 -0800)] 
oops, typo

10 months agooptimize IV -> UV conversions
Tomasz Konojacki [Wed, 21 Nov 2018 08:26:31 +0000 (09:26 +0100)] 
optimize IV -> UV conversions

This commit replaces all instances of code that looks like this:

  uv = (iv == IV_MIN) ? (UV)iv : (UV)(-iv)

with simpler and more optimal:

  uv = -(UV)iv

While -iv indeed results in an undefined behaviour when iv == IV_MIN,
-(UV)iv is perfectly well defined and does the right thing.

C standard guarantees that the result of (UV)iv (for negative iv) is
equal to iv + UV_MAX + 1 (see 6.3.1.3, paragraph 2 in C11). It also
guarantees that the result of -uv is UV_MAX - uv + 1 (6.2.5,
paragraph 9).

That means that the result of -(UV)iv is UV_MAX - (iv + UV_MAX + 1) + 1
which is equal to -iv for *all* possible negative values of iv.

[perl #133677]

10 months agoS_hv_delete_common(): avoid undefined behaviour
David Mitchell [Wed, 21 Nov 2018 12:09:45 +0000 (12:09 +0000)] 
S_hv_delete_common(): avoid undefined behaviour

ASAN -fsanitize-undefined was tripping on the second of these two lines:

    svp = AvARRAY(isa);
    end = svp + AvFILLp(isa)+1;

In the case where svp is NULL and AvFILLp(isa) is -1, the first addition
is undefined behaviour. Add the 1 first, so that it becomes
svp + (-1+1), which is safe.

10 months agolgtm.yml: fix erroneous inclusion
Dominic Hargreaves [Wed, 21 Nov 2018 10:49:39 +0000 (10:49 +0000)] 
lgtm.yml: fix erroneous inclusion

10 months agoadd entries for Module-CoreList 5.0181220
Karen Etheridge [Tue, 20 Nov 2018 22:46:24 +0000 (14:46 -0800)] 
add entries for Module-CoreList 5.0181220

I did this manually; I am not sure if 'perl -Ilib Porting/corelist.pl
cpan' would add these entries on the next blead-point release day.

10 months agobump version of released Module-CoreList
Karen Etheridge [Tue, 20 Nov 2018 22:46:05 +0000 (14:46 -0800)] 
bump version of released Module-CoreList

10 months agoBump the perl version in various places for 5.29.6
Karen Etheridge [Tue, 20 Nov 2018 22:24:49 +0000 (14:24 -0800)] 
Bump the perl version in various places for 5.29.6

10 months agonew perldelta for 5.29.6
Karen Etheridge [Tue, 20 Nov 2018 22:18:38 +0000 (14:18 -0800)] 
new perldelta for 5.29.6

10 months agotick off 5.29.5 release
Karen Etheridge [Tue, 20 Nov 2018 22:17:22 +0000 (14:17 -0800)] 
tick off 5.29.5 release