This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Karl Williamson [Fri, 24 Sep 2010 02:47:03 +0000 (20:47 -0600)]
Add 256 word bit table of character classes
This patch adds a table for looking up character classes. It is 256
words long, in l1_char_class_tab.h, with each word corresponding to the
ordinal of a Latin1 character, and each word contains a bit map of all
the properties that character matches. Each property has a bit or two.
Ones named _CC_property_A are true only if the character is also in the
ASCII character set. Ones named CC_property_L1 do not have this
restriction. (L1 stands for Latin1.)
Also added is a script that generates the table. It is not anticipated
that this will need to be used often.
(This commit was changed from its original form by Steffen.)
Karl Williamson [Fri, 24 Sep 2010 02:42:40 +0000 (20:42 -0600)]
handy.h: EBCDIC should use native isalpha()
Karl Williamson [Fri, 24 Sep 2010 02:27:32 +0000 (20:27 -0600)]
handy.h: Add isFOO_A() macros for ASCII range matches
These macros return true only if the parameter is an ASCII character.
Karl Williamson [Thu, 23 Sep 2010 21:40:21 +0000 (15:40 -0600)]
handy.h: should use EBCDIC libc isdigit()
as is better optimized and suitable for the purpose.
Karl Williamson [Thu, 23 Sep 2010 21:08:06 +0000 (15:08 -0600)]
handy.h: move macro in file
Karl Williamson [Thu, 23 Sep 2010 20:40:42 +0000 (14:40 -0600)]
Subject: handy.h: Add isWORDCHAR() for clarity
The name isALNUM() is problematic, as it is very close to isALNUMC(),
and doesn't mean exactly what most people might think. I presume the C
in isALNUMC stands for C language or libc, but am not sure. Others
don't know either. But in any event, isALNUM is different from the C
isalnum(), in that it matches the Perl concept of \w, which differs from
the C definition in exactly one place. Perl includes the underscore
character, '_'.
So, I'm adding a isWORDCHAR() macro for future code to use to be more
clear. I thought also about isWORD(), but I think confusion can arise
from thinking that means a whole word. isWORDCHAR_L1() matches in the
Latin1 range, to be equivalent to isALNUMU(). The motivation for using
L1 instead of U will be explained in a commit message for the other L1
macros that are to be added.
Karl Williamson [Thu, 23 Sep 2010 20:30:54 +0000 (14:30 -0600)]
Add a comment; clarify another
Karl Williamson [Thu, 23 Sep 2010 20:27:57 +0000 (14:27 -0600)]
Indent a comment better
Karl Williamson [Thu, 23 Sep 2010 20:26:51 +0000 (14:26 -0600)]
Subject: handy.h: Reorder #defines alphabetically
The only change here is that I sorted these #defines within their
groups, to make it much easier to follow what's going on.
Karl Williamson [Thu, 23 Sep 2010 20:14:09 +0000 (14:14 -0600)]
handy.h: isSPACE() is wrong for EBCDIC
It didn't include the Latin1 space components.
Karl Williamson [Thu, 23 Sep 2010 19:57:51 +0000 (13:57 -0600)]
handy.h: EBCDIC isBLANK() is wrong
It doesn't include NBSP
Karl Williamson [Thu, 23 Sep 2010 19:45:58 +0000 (13:45 -0600)]
handy.h: isPSXSPC() is wrong for EBCDIC
The macro was using the ASCII definition, which doesn't include NEL nor
NBSP. But, libc contains the correct definition, which is usable on
EBCDIC since we don't worry about locales there.
Karl Williamson [Thu, 23 Sep 2010 19:41:52 +0000 (13:41 -0600)]
Subject: handy.h: Move defn's outside #ifndef EBCDIC
Commit
4125141464884619e852c7b0986a51eba8fe1636 improperly got rid of
EBCDIC handling, as it combined the ASCII and EBCDIC versions, but left
the result in the ASCII-only branch. Just move to the common code.
Karl Williamson [Mon, 13 Sep 2010 04:19:54 +0000 (22:19 -0600)]
charnames.pm: reformat comments
Now that have less indent, don't need so many lines. The only changes
in this commit are several blocks of comments to occupy more of each
line. No wording changes are involved.
Karl Williamson [Mon, 13 Sep 2010 04:17:35 +0000 (22:17 -0600)]
charnames.pm: indent less to fit in 80 columns
This patch changes white space only. It lessens the indent of certain
lines that were made longer in an earlier commit, and now most of them
fit into 80 columns.
Karl Williamson [Mon, 13 Sep 2010 03:33:12 +0000 (21:33 -0600)]
Teach Perl about Unicode named character sequences
mktables is changed to process the Unicode named sequence file.
charnames.pm is changed to cache the looked-up values in utf8. A new
function, string_vianame is created that can handle named sequences, as
the interface for vianame cannot. The subroutine lookup_name() is
slightly refactored to do almost all of the common work for \N{} and the
vianame routines. It now understands named sequences as created my
mktables..
tests and documentation are added. In the randomized testing section,
half use vianame() and half string_vianame().
Karl Williamson [Sun, 12 Sep 2010 18:46:07 +0000 (12:46 -0600)]
charnames.pm: Nits in pod
Karl Williamson [Sun, 12 Sep 2010 16:47:56 +0000 (10:47 -0600)]
perlrecharclass.pod: Add caveat about multi-char sequences
Inside a bracketed character class, any \N{name} which expands to more
than one character will have only the first one considered. This
doesn't need named character sequences, as user-defined aliases have
long been able to be multi-char.
Karl Williamson [Sun, 12 Sep 2010 16:27:36 +0000 (10:27 -0600)]
charnames.pm: Clarify comments
Karl Williamson [Sun, 12 Sep 2010 15:50:15 +0000 (09:50 -0600)]
charnames.t: Add output message
Karl Williamson [Sat, 11 Sep 2010 19:35:23 +0000 (13:35 -0600)]
charnames.t: Clarify message
Karl Williamson [Sat, 11 Sep 2010 19:11:44 +0000 (13:11 -0600)]
charnames.t: Clarify value is hex
Karl Williamson [Sat, 11 Sep 2010 19:10:37 +0000 (13:10 -0600)]
charnames.t: Add tests for NameAliases
Karl Williamson [Sat, 11 Sep 2010 17:17:03 +0000 (11:17 -0600)]
Fix casing, wording
Karl Williamson [Sat, 11 Sep 2010 15:20:04 +0000 (09:20 -0600)]
Fix spelling
Karl Williamson [Fri, 10 Sep 2010 17:38:25 +0000 (11:38 -0600)]
charnames.pm: Change variable name
This is an intermediate commit in preparation for handling named
sequences
Karl Williamson [Fri, 10 Sep 2010 16:47:47 +0000 (10:47 -0600)]
charnames.t: Add code so can test 100% of names
If the percentage of characters to test is changed to 100%, add code to
make the block size 1. This guarantees each character gets tested in
spite of randomness
Karl Williamson [Fri, 10 Sep 2010 16:47:15 +0000 (10:47 -0600)]
charnames.t: clarify comments
Karl Williamson [Fri, 10 Sep 2010 00:06:22 +0000 (18:06 -0600)]
charnames.t: Don't call srand(undef)
srand(undef) is the same as srand(0). The code is trying to get random
seeds, not a fixed one.
Karl Williamson [Thu, 9 Sep 2010 23:50:16 +0000 (17:50 -0600)]
charnames: Remove unnecessary \t in Name.pl
The double \t\t is unnecessary, and so we can remove one of them,
shortening the table.
Karl Williamson [Thu, 9 Sep 2010 23:16:53 +0000 (17:16 -0600)]
charnames.pm: Small performance enhancements
mktables is changed to output 5 digit code points, which means that
charnames doesn't have to go looking for the boundaries, which gives a
slight performance enhancement.
Karl Williamson [Thu, 9 Sep 2010 22:34:55 +0000 (16:34 -0600)]
mktables: Remove stubbed out code
This commented out code will never be used, as a different solution was
done in charnames.
It was to automatically handle parenthesized character names. Unicode
is extremely unlikely to ever add new names like this, and all the
existing names are now hard-coded in charnames.pm
Nicholas Clark [Wed, 22 Sep 2010 15:31:16 +0000 (16:31 +0100)]
Use the same $func to Perl_$func macros independent of multiplicity.
The macros of the form #define foo(a, b) Perl_foo(aTHX_ a, b) will also work
under non-multiplicity. There's no issue with adding explicit arguments in
the non-multiplicity case, because it changes one form of compile-time error
into another. (Function Perl_foo passed the wrong number of arguments into
macro foo passed the wrong number of arguments.)
Father Chrysostomos [Sat, 25 Sep 2010 06:12:00 +0000 (23:12 -0700)]
Make the new sigtrap.t test more tolerant of OS differences
Father Chrysostomos [Sat, 25 Sep 2010 05:22:26 +0000 (22:22 -0700)]
define SvTRUE_nomg for compilers other than gcc
This time I *really* broke the Windows build!
Father Chrysostomos [Sat, 25 Sep 2010 03:42:54 +0000 (20:42 -0700)]
perldelta entries for the double-magic fixes [perl #76814]
Father Chrysostomos [Sat, 25 Sep 2010 03:33:47 +0000 (20:33 -0700)]
[perl #76814] FETCH called twice - y
This patch stops y from calling get-magic twice. (This has caused
double magick since as far back as 5.6.2.)
Father Chrysostomos [Sat, 25 Sep 2010 03:33:42 +0000 (20:33 -0700)]
[perl #76814] FETCH called twice - m and s
This fixes m and s. It modifies pp_regcomp to avoid extra magic. It
also corrects a bug in sv_catsv_flags, which would still call
mg_get(ssv) even without the SV_GMAGIC flag set.
Father Chrysostomos [Sat, 25 Sep 2010 03:33:03 +0000 (20:33 -0700)]
[perl #76814] FETCH called twice - !
This fixes ! by changing sv_2bool to sv_2bool_flags (with a macro
wrapper) and adding SvTRUE_nomg. It also corrects the docs that state
incorrectly that SvTRUE does not handle magic.
Father Chrysostomos [Sat, 25 Sep 2010 03:31:28 +0000 (20:31 -0700)]
[perl #76814] FETCH called twice - string comparison ops
This patch changes sv_eq, sv_cmp, sv_cmp_locale and sv_collxfrm
to _flags forms, with macros under the old names for sv_eq and
sv_collxfrm, but functions for sv_cmp* since pp_sort.c needs them.
Jesse Vincent [Fri, 24 Sep 2010 23:20:29 +0000 (19:20 -0400)]
Committing the current freeze and release plan for 5.14. khw++ for
noticing that I'd failed to do this after conference season.
gregor herrmann [Fri, 24 Sep 2010 20:17:34 +0000 (22:17 +0200)]
Fis various Math::BigInt spelling mistakes
Jerry D. Hedden [Fri, 24 Sep 2010 20:42:04 +0000 (21:42 +0100)]
[perl #78000] [PATCH] Upgrade to threads 1.79
Attached is a revised patch to upgrade to v1.79. I needed to fix
the tests in the newly added t/kill2.t file. The blead version will
be 1.79_01.
Father Chrysostomos [Fri, 24 Sep 2010 19:28:36 +0000 (12:28 -0700)]
Modify a utf8.t test in anticipation of a fix for [perl #74022]
In anticipation of a fix for [perl #74022], this patch corrects the
test for q·foo· in utf8.t to use ¡ instead of ·. I’ve also moved the
test outside of the eval, so it will still run (and fail) if the com-
pilation fails, instead of causing an invalid test count.
David Golden [Fri, 24 Sep 2010 18:03:49 +0000 (18:03 +0000)]
fix typo in perl5135delta
Karl Williamson [Fri, 24 Sep 2010 14:46:12 +0000 (08:46 -0600)]
perldelta: Note Perl version of regexp_pattern()
perldelta suggests using this, but it didn't come into being until
5.9.5, so cpan modules may not be able to use it.
Nicholas Clark [Fri, 24 Sep 2010 10:46:41 +0000 (11:46 +0100)]
Add a warning to test.pl if quotes are used in one-liners.
Fix all the tests that were using quotes, and (by implication) working
because barewords are treated as strings, and one-liners aren't run under
strict.
Nicholas Clark [Fri, 24 Sep 2010 10:22:24 +0000 (11:22 +0100)]
Avoid using double quotes in a one-liner runperl.
test.pl uses "" for command line quoting on Win32, VMS and NetWare, '' on *nix.
Hence what "works on my machine" on *nix may not work elsewhere.
Father Chrysostomos [Fri, 24 Sep 2010 07:45:35 +0000 (00:45 -0700)]
sigtrap version bump + perldelta
Father Chrysostomos [Fri, 24 Sep 2010 07:42:32 +0000 (00:42 -0700)]
Missing link in perldelta
Father Chrysostomos [Fri, 24 Sep 2010 07:34:04 +0000 (00:34 -0700)]
Add tests for [perl #72340] sigtrap attempts to modify read-only thing
Joshua N Pritikin [Fri, 24 Sep 2010 07:23:46 +0000 (00:23 -0700)]
[perl #72340] sigtrap attempts to modify read-only thingy
No much to say. sigtrap otherwise attempts to modify a
read-only thingy. Patch below solves it for me.
Father Chrysostomos [Fri, 24 Sep 2010 07:22:11 +0000 (00:22 -0700)]
Add Joshua Pritikin’s other e-mail address to checkAUTHORS.pl
Alex Davies [Fri, 24 Sep 2010 07:12:27 +0000 (00:12 -0700)]
The rest of [perl #71710] fixes for File::Find
The code is from Alex Davies. The commit message is from the committer.
Somehow I only ended up with half the patch applied. No tests failed
for me, because these code paths are only reached on Windows, which,
thankfully, I am not using.
Hopefully this will eliminate the smoke that is already pouring in....
(It’s *my* turn to break blead for Windows. :-)
Father Chrysostomos [Fri, 24 Sep 2010 07:03:23 +0000 (00:03 -0700)]
Oops. I broke sigtrap.t’s test output
Karl Williamson [Fri, 24 Sep 2010 06:24:18 +0000 (23:24 -0700)]
Rename isALNUM_L1 to isWORDCHAR_L1
Father Chrysostomos [Fri, 24 Sep 2010 06:08:18 +0000 (23:08 -0700)]
Convert sigtrap.t to test.pl
Karl Williamson [Thu, 23 Sep 2010 18:16:12 +0000 (12:16 -0600)]
handy.h: Add isALNUM_L1() macro
This is a synonym for isALNUMU
Karl Williamson [Thu, 23 Sep 2010 17:32:13 +0000 (11:32 -0600)]
Subject: handy.h: Add isSPACE_L1 with Unicode semantics
Karl Williamson [Fri, 24 Sep 2010 01:50:34 +0000 (19:50 -0600)]
perldelta: Give better instructions about (?^...)
Avar pointed out that there is an existing API wwhich helps in coping
with this
Father Chrysostomos [Fri, 24 Sep 2010 05:15:02 +0000 (22:15 -0700)]
perldelta entry for File::DosGlob
Alex Davies [Fri, 24 Sep 2010 05:08:28 +0000 (22:08 -0700)]
[perl #71712] fixes for File::DosGlob
The changes are
1. Allow for parentheses in glob pattern.
2. Strip redundant "./" from drive relative glob patterns results.
brian d foy [Fri, 24 Sep 2010 01:43:00 +0000 (20:43 -0500)]
* Fixed output for lex-like scanners example in perlop
This should probably be given-when instead of a naked block
and a redo
brian d foy [Fri, 24 Sep 2010 01:34:08 +0000 (20:34 -0500)]
perlfaa4: fix a one-off mistake
+ How do I perform an operation on a series of integers?
Father Chrysostomos [Fri, 24 Sep 2010 00:45:59 +0000 (17:45 -0700)]
perldelta entry for File::Find
Alex Davies [Fri, 24 Sep 2010 00:23:49 +0000 (17:23 -0700)]
[perl #71710] fixes for File::Find
Please find attached patches for File::Find and its test file.
These changes ensure that paths passed to File::Find::find() on Win32
which have a trailing *back*slash are neatly handled. That is, the
change ensures paths such as c:\dir\/file are no longer generated.
Chris 'BinGOs' Williams [Thu, 23 Sep 2010 16:46:43 +0000 (17:46 +0100)]
Updated Unicode-Collate to CPAN version 0.60
[DELTA]
0.60 Thu Sep 23 21:37:36 2010
- bug fix: index() [and its friends including gmatch()] didn't remove
ignorable characters in the substring correctly.
Thanks for the bug report:
http://www.xray.mpe.mpg.de/mailing-lists/perl-unicode/2010-09/msg00014.html
- U::C::Locale newly supports locales: de__phonebook, nso, om, tn, vi.
- precomposites of a-breve, a-circ, e-circ, o-circ are tailored as well.
(affected locales: ro, sk, sv)
Andy Dougherty [Thu, 23 Sep 2010 16:06:10 +0000 (12:06 -0400)]
Add default safe guesses for static_inline to plan9 files.
Andy Dougherty [Thu, 23 Sep 2010 15:59:47 +0000 (11:59 -0400)]
Add static inline values to canned Netware/config_H.wc file.
Andy Dougherty [Thu, 23 Sep 2010 15:55:28 +0000 (11:55 -0400)]
Update canned win32/config_H.* files with static inline values
to match the canned win32/config.* settings. For .bc and
.ce, we still have a plain 'static'. Better settings from
knowledgeable users about those platforms would be welcome.
Andy Dougherty [Thu, 23 Sep 2010 15:53:16 +0000 (11:53 -0400)]
Update static_inline guesses in win32/config.[gv]c*
For gcc, assume all variants used on Win32 will support
static __inline__. VC already had static __inline.
Update VC64 to match.
Nicholas Clark [Thu, 23 Sep 2010 14:16:05 +0000 (15:16 +0100)]
Some tests for Perl_grok_number().
Not yet comprehensive - only tests the integer conversion code paths.
Andy Dougherty [Thu, 23 Sep 2010 14:07:41 +0000 (10:07 -0400)]
Set default static inline for Netware/config.wc.
If anyone knows how to support static inline on Netware,
a patch would be appreciated.
Andy Dougherty [Thu, 23 Sep 2010 13:56:11 +0000 (09:56 -0400)]
Add default setting of perl_static_inline='static' to all win32 configs.
Win32 users who know better are invited to change the definitions
as appropriate, but this should at least allow it to build.
Andy Dougherty [Thu, 23 Sep 2010 13:53:43 +0000 (09:53 -0400)]
Set d_static_inline for win32/config.vc, since VC supports static __inline.
Tony Cook [Thu, 23 Sep 2010 10:44:37 +0000 (20:44 +1000)]
Make dquote_static.c available to ext/re/
Under Win32 the main perl source directory isn't in the C include
path, so as we do with the re source files, copy dquote_static.c to
the ext/re directory.
Tony Cook [Thu, 23 Sep 2010 10:43:44 +0000 (20:43 +1000)]
Configure inline support for 32-bit VC++
VC supports static inline via "static __inline".
Father Chrysostomos [Thu, 23 Sep 2010 08:38:10 +0000 (01:38 -0700)]
[perl #71806] perldb does not setup %dbline with the shebang option -d
The first time gv_fetchfile is called for a particular file, it
creates the glob and, if debugging is on, creates an AV. If the glob
already exists (i.e., in subsequent calls), the AV is not created. The
attached patch moves the check for debugging mode and the creation of
the AV outside the if-block that checks whether the glob exists.
This bug seems to have existed for a very long time and has been
intermittent. It seems that many different things can change the order
in which #!perl -d and gv_fetchfile occur. Whether compilation options
affect it I do not know. I can reproduce it in 5.6.2, 5.8.[123456]
(non-threaded) and 5.11.3 (both threaded and non-threaded), but not
5.8.[789] or 5.10.[01] (threaded).
Father Chrysostomos [Thu, 23 Sep 2010 08:01:20 +0000 (01:01 -0700)]
perldelta for bug #70615 + tweaks
Steve Hay [Thu, 23 Sep 2010 07:28:32 +0000 (08:28 +0100)]
Fix code-before-declaration problem in universal.c
Father Chrysostomos [Thu, 23 Sep 2010 06:47:11 +0000 (23:47 -0700)]
[perl #70614] Seg. fault with eval/use/UNITCHECK/DESTROY/eval combination
This case can be reduced to:
eval "UNITCHECK{ eval 0 }"
(eval 0 triggers the bug more reliably than eval ''.)
The inner eval clobbers PL_eval_start, so when the UNITCHECK finishes,
the outer eval tries to call the inner eval’s code, which has been
freed already.
This commit saves PL_eval_start and restores it after UNITCHECK.
Jerry D. Hedden [Wed, 22 Sep 2010 20:36:14 +0000 (16:36 -0400)]
Remove unneeded 'use' from ext/XS-APItest/t/peep.t Devel::Peek is not used by ext/XS-APItest/t/peep.t, so remove it from the test code.
Karl Williamson [Tue, 21 Sep 2010 21:09:12 +0000 (15:09 -0600)]
regcomp.c: Convert some things to use cBOOL()
Karl Williamson [Tue, 21 Sep 2010 01:16:24 +0000 (19:16 -0600)]
re.pm: Change comment to use new (?^...)
Karl Williamson [Tue, 21 Sep 2010 00:57:24 +0000 (18:57 -0600)]
Add /d, /l, /u (infixed) regex modifiers
This patch adds recognition of these modifiers, with appropriate action
for d and l. u does nothing useful yet. This allows for the
interpolation of a regex into another one without losing the character
set semantics that it was compiled with, as for the first time, the
semantics is now specified in the stringification as one of these
modifiers.
To this end, it allocates an unused bit in the structures. The off-
sets change so as to not disturb other bits.
Karl Williamson [Tue, 21 Sep 2010 00:31:00 +0000 (18:31 -0600)]
Change .t to use new (?^...)
There is a line in this .t which uses the old regex stringification.
The contents are not currently tested for, but for cleanliness, change
to the new.
Karl Williamson [Tue, 21 Sep 2010 00:29:59 +0000 (18:29 -0600)]
Change comments, documentation for new (?^...)
I overlooked these earlier in adding the caret notation.
Karl Williamson [Tue, 21 Sep 2010 00:26:33 +0000 (18:26 -0600)]
Change to use mnemonic instead of char constant
The new '^' in (?^...) should really be a macro.
Karl Williamson [Sun, 19 Sep 2010 18:29:02 +0000 (12:29 -0600)]
handy.h: isASCII() extend to work on > 8 bit values
Prior to this patch, if isASCII() is called with something like '256',
it would return true.
For some reason unknown to me, U64 is defined only inside the perl core.
However, the equivalent U64TYPE is known everywhere, so in the macro
that can be called outside of core, use that instead.
The commit log doesn't give a reason for not defining U64 outside of
core, and no tests in the suite fail when it is defined outside core.
But out of caution, I'm just doing this workaround instead of exposing
U64.
Karl Williamson [Sun, 19 Sep 2010 17:57:04 +0000 (11:57 -0600)]
handy.h: Don't use isascii() as not in all libc's
EBCDIC platforms use isascii(), but is not in all libc's so better to
use our own.
Karl Williamson [Sun, 19 Sep 2010 17:04:34 +0000 (11:04 -0600)]
handy.h: Fix-up documentation
Previous documentation was wrong for EBCDIC platforms. This fixes that
and adds some more explanation.
Karl Williamson [Sun, 19 Sep 2010 16:38:14 +0000 (10:38 -0600)]
handy.h: toUPPER is not a char class fcn
toUPPER() and toLOWER() were grouped with the character class functions
(in perlapi), to which they are related, but aren't the same. Create a
new heading for these.
Karl Williamson [Sun, 19 Sep 2010 16:20:58 +0000 (10:20 -0600)]
autodoc.pl: Allow heading level documentation
This patch changes autodoc.pl to accept text that is to come immediately
after headings, so that it applies to across the whole section, and not
just to an individual function within that section.
Father Chrysostomos [Wed, 22 Sep 2010 20:10:19 +0000 (13:10 -0700)]
Add lib to @INC in t/base/rs.t
t/base/rs.t needs to @INClude the lib directory before requiring
PerlIO::scalar, even if the require is in an eval, because it will
otherwise pick up a previously-installed copy.
Since
87b9e1600 (Eliminate PL_* accessor functions under ithreads), a
binary-incompatible change, I’ve been getting these errors:
t/base/rs......................................................dyld: lazy symbol binding failed: Symbol not found: _Perl_Istack_sp_ptr
Referenced from: /usr/local/lib/perl5/5.13.5/darwin-thread-multi-2level/auto/PerlIO/scalar/scalar.bundle
Expected in: dynamic lookup
dyld: Symbol not found: _Perl_Istack_sp_ptr
Referenced from: /usr/local/lib/perl5/5.13.5/darwin-thread-multi-2level/auto/PerlIO/scalar/scalar.bundle
Expected in: dynamic lookup
FAILED--no leader found
Failed a basic test (base/rs.t) -- cannot continue.
make: *** [test] Error 255
Andy Dougherty [Wed, 22 Sep 2010 17:45:47 +0000 (13:45 -0400)]
Run perl regen.pl after removing regcurly from embed.fnc.
Since regcurly is now a static inline function, it no longer
needs to appear in embed.fnc. embed.pl doesn't quite have the
right flags to deal with static inline functions, so I just
removed regcurly entirely. It's not for embedding or exporting
anyway.
Andy Dougherty [Wed, 22 Sep 2010 17:44:36 +0000 (13:44 -0400)]
Extract regcurly as a static inline function.
This patch extracts regcurly from regcomp.c and converts it
to a static inline function in a new file dquote_static.c
that is now #included by regcomp.c and toke.c. This change
will require 'make regen'.
Nicholas Clark [Wed, 22 Sep 2010 16:32:15 +0000 (17:32 +0100)]
Remove @extvars from embed.pl
It's been unused since PERL_POLLUTE was removed in
cfef31b27d1a56cd.
Jerry D. Hedden [Wed, 22 Sep 2010 13:17:14 +0000 (09:17 -0400)]
Add missing newline to macro_to_ifndef in ExtUtils::Constant::Base
Nicholas Clark [Wed, 22 Sep 2010 15:05:40 +0000 (16:05 +0100)]
In embed.pl, replace C<$ret .=> with C<$ret => when we know that $ret is "".
Remove one unnecessary initialisation of $ret to "", when all code paths beyond
assign to it. Remove one $ret completely (from &write_global_sym).
Nicholas Clark [Wed, 22 Sep 2010 14:25:17 +0000 (15:25 +0100)]
Remove &munge_c_files from embed.pl, as it has never been used.
It was added, with the only call to it commented out, in
cea2e8a9dd23747f.
This means that walk_table's support for undef meaning 'no output' can
removed.