Father Chrysostomos [Mon, 2 Jan 2012 21:23:51 +0000 (13:23 -0800)]
perldelta for helem and hslice on null values
Father Chrysostomos [Mon, 2 Jan 2012 21:20:07 +0000 (13:20 -0800)]
Correct comment in APItest’s hash.t
Father Chrysostomos [Mon, 2 Jan 2012 21:18:51 +0000 (13:18 -0800)]
perldelta for newHVhv on tied hashes
Father Chrysostomos [Mon, 2 Jan 2012 21:17:14 +0000 (13:17 -0800)]
perldelta for deparsing features
Father Chrysostomos [Mon, 2 Jan 2012 20:57:46 +0000 (12:57 -0800)]
perldelta for feature revamp
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.
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.
Nicholas Clark [Mon, 2 Jan 2012 17:26:39 +0000 (18:26 +0100)]
Merge the refactoring of installperl that fixes Pod installation bugs.
Nicholas Clark [Mon, 2 Jan 2012 17:02:38 +0000 (18:02 +0100)]
Note the user-visible installperl changes in perldelta.
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.
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)
Nicholas Clark [Mon, 26 Dec 2011 13:05:13 +0000 (14:05 +0100)]
In installperl, don't chdir 'lib' before finding files in lib/
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.
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.
Nicholas Clark [Sat, 24 Dec 2011 12:40:24 +0000 (13:40 +0100)]
In installperl, simplify the strip/chmod code for libraries on Darwin.
Nicholas Clark [Sat, 24 Dec 2011 10:53:02 +0000 (11:53 +0100)]
Sanitise NetWare changes to installperl.
Commit
2986a63f7e513cf3's changes to installperl could have been cleaner.
Consistently pass $opts{verbose} to all invocations of mkpath().
The skip logic for paths matching /\.(?:nlp|nlm|bs)$/ should have been
earlier, and should not have created copy&paste code duplication of the
calls to copy_if_diff() and chmod.
Nicholas Clark [Sun, 1 Jan 2012 16:48:29 +0000 (17:48 +0100)]
Avoid t/re/regexp.t failing on miniperl when displaying TODO test output.
Change
1e9285c2ad54ae39 refactored Data::Dumper to load on miniperl.
t/re/regexp.t attempts to load Data::Dumper (in an eval) to display failure
output, including the failure of TODO tests. Hence Data::Dumper is now loaded
without error as part of minitest, so regexp.t then attempts to use
Data::Dumper to output better diagnostics. This fails (hard) because
Data::Dumper attempts to load Scalar::Util, which attempts to load B, which
bails out because this is miniperl.
It's not obvious that there's a 100% solution here that gets full-on
Data::Dumper functionality for miniperl.
Nicholas Clark [Sun, 1 Jan 2012 20:35:33 +0000 (21:35 +0100)]
In diagnostics.t, avoid passing a literal \n in the command to run.
Express the one-liner with qq to interpolating \n to a newline character at
execution time, instead of q quoting a literal newline character. Passing
a literal newline character is not portable to all platforms.
Father Chrysostomos [Sun, 1 Jan 2012 22:19:54 +0000 (14:19 -0800)]
Don’t double-free GVs in gv:S_maybe_add_coresub
Commit
8756617677 did not take into account that newSVOP steals a ref-
erence count.
Father Chrysostomos [Sun, 1 Jan 2012 21:57:06 +0000 (13:57 -0800)]
[perl #107366] Don’t clone GVs during thread join
unless they are orphaned.
This commit stops globs that still reside in their stashes from being
cloned during a join.
That way, a sub like sub{$::x++}, when cloned into a subthread and
returned from it, will still point to the same $::x.
This commit takes the conservative approach of copying on those globs
that can be found under their names in the original thread.
While this doesn’t work for all cases, it’s probably not possible to
make it work all the time.
Father Chrysostomos [Sun, 1 Jan 2012 21:06:27 +0000 (13:06 -0800)]
Simplify gv:S_maybe_add_coresub
It was working around the fact that newATTRSUB expects to be able to
look up the GV by name.
And for speed, it was going through hoops to try to avoid creating
extra SVs holding the name unnecessarily.
By tweaking newATTRSUB to accept a GV instead of a name, we can sim-
plify not only S_maybe_add_coresub but all its callers, too.
Craig A. Berry [Sun, 1 Jan 2012 14:40:54 +0000 (08:40 -0600)]
sigsystem.t depends on fork().
So don't run it if the function doesn't exist.
Father Chrysostomos [Sun, 1 Jan 2012 08:16:33 +0000 (00:16 -0800)]
Suppress void warning in t/op/sprintf.t
Father Chrysostomos [Sun, 1 Jan 2012 07:52:14 +0000 (23:52 -0800)]
Correct comment in op.c
‘Backward compatibility mode’ is poppycock, I’m afraid. When part of
an argument list a sub call it may or may not be an lvalue. That can-
not be determined until the value passed is (or is not) assigned to.
It has little to do with backward compatibility.
Father Chrysostomos [Sun, 1 Jan 2012 07:49:01 +0000 (23:49 -0800)]
[perl #103492] Make %n printf format work with Unicode
It was using the internal byte count instead of the number of charac-
ters. The iatter is documented. The former is useless, even for C
code calling this, as later arguments could cause the current buffer
to be upgraded to utf8, throwing off any offsets returned.
Father Chrysostomos [Sun, 1 Jan 2012 07:24:57 +0000 (23:24 -0800)]
[perl #103492] Give lvalue cx to (s)printf args
Or potential lvalue context, like function calls.
The %n format code’s existence renders these two very much like func-
tion calls, as they can modify their arguments.
This allows sprintf("...%n", substr ...) to work.
Father Chrysostomos [Sun, 1 Jan 2012 03:30:04 +0000 (19:30 -0800)]
[perl #103492] Correct docs for %n format in perlfunc
Father Chrysostomos [Sun, 1 Jan 2012 03:20:17 +0000 (19:20 -0800)]
hv.c: Consistent spaces after dots in apidocs
Father Chrysostomos [Sun, 1 Jan 2012 03:12:30 +0000 (19:12 -0800)]
Remove method cache reset from hv.c:hv_free_ent
It hasn’t been necessary since commit
f50383f5. Before that it wasn’t
sufficient. See commit 5743f2a for details.
If a hash element is being deleted, S_hv_delete_common takes care of
this. If a hash is being freed or cleared, hv_undef or hv_clear takes
care of it.
Father Chrysostomos [Sun, 1 Jan 2012 02:58:29 +0000 (18:58 -0800)]
squash some code in hv.c:S_hv_delete_common
Changes four commits ago made this possible.
Father Chrysostomos [Sun, 1 Jan 2012 02:54:14 +0000 (18:54 -0800)]
[perl #100340] Free hash entries before values on delete
When a hash element is deleted in void context, if the value is freed
before the hash entry, it is possible for a destructor to see the hash
in an inconsistent state--inconsistent in that it contains entries
that are about to be freed, with nothing to indicate that. So the
destructor itself could free the very same hash entry (e.g., by
freeing the hash), resulting in a double free, panic, or other
unpleasantness.
Father Chrysostomos [Sun, 1 Jan 2012 02:46:37 +0000 (18:46 -0800)]
mro.c: Consistent use of spaces after dots in apidocs
Father Chrysostomos [Sun, 1 Jan 2012 02:44:23 +0000 (18:44 -0800)]
mro.c: Use F<> for file names
Father Chrysostomos [Sun, 1 Jan 2012 02:18:57 +0000 (18:18 -0800)]
Update method caches for non-void stash elem deletions
Commit 71be2cb (inseparable changes from patch from perl5.003_13 to
perl5.003_14) added code to hv_free_ent to reset method caches if
necessary.
Later, a bug fix in the deletion code (
f50383f5) made it necessary for
the value in the HE to be set to &PL_sv_placeholder, so it wouldn’t
free the sv just yet. So the method cache code (which by then had
changed from using PL_sub_generation to using mro_method_changed_in)
got repeated in S_hv_delete_common.
The only problem with all this is that hv_free_ent was the wrong place
to put it to begin with. If delete is called in non-void context,
the sv in it is not freed just yet, but returned; so hv_free_ent was
already being called with a HE pointing to &PL_sv_placeholder.
Commit
f50383f5 only added the mro code for the void case, to avoid
changing existing behaviour when rearranging the code.
It turns out it needs to be done in S_hv_delete_common uncon-
ditionally.
Changing a test in t/op/method.t to use ()=delete instead of delete
makes it fail, because method caches end up going stale. See the test
in the diff.
Craig A. Berry [Sun, 1 Jan 2012 00:28:46 +0000 (18:28 -0600)]
'S' case already exists with ALTERNATE_SHEBANG.
Follow-up to
b7e077d0225eac833ce2eb6fe9e1807f77d0f848.
Craig A. Berry [Sat, 31 Dec 2011 23:36:02 +0000 (17:36 -0600)]
Current VMS can handle filenames with spaces.
If they are escaped and on an ODS-5 volume, that is. We
then have to clean up all versions of the files created
by a test we weren't running before.
Father Chrysostomos [Sat, 31 Dec 2011 19:02:50 +0000 (11:02 -0800)]
sigsystem.t: put miniperl skip earlier
Under minitest, extensions like constant.pm and Data::Dumper
haven’t been processed yet, so they won’t be available.
Father Chrysostomos [Sat, 31 Dec 2011 18:55:36 +0000 (10:55 -0800)]
Correct skip count in sigsystem.t
Father Chrysostomos [Sat, 31 Dec 2011 18:53:10 +0000 (10:53 -0800)]
Allow Data::Dumper to load on miniperl
For a long time it has been doing eval { require XSLoader }, but
XSLoader will load on miniperl, it’s just XSLoader::load() that fails.
Father Chrysostomos [Sat, 31 Dec 2011 18:49:54 +0000 (10:49 -0800)]
Increase $Data::Dumper::VERSION to 2.135_03
I’m about to make a change, and 2.135_02 has already been used by a
CPAN dev release.
Leon Timmermans [Thu, 8 Dec 2011 23:32:10 +0000 (00:32 +0100)]
Block SIGCHLD during system() call (per POSIX)
Leon Timmermans [Fri, 30 Dec 2011 18:02:07 +0000 (20:02 +0200)]
Added tests for SIGCHLD blocking during system()
Craig A. Berry [Sat, 31 Dec 2011 18:31:49 +0000 (12:31 -0600)]
strlcpy fix-ups in vms/vms.c following
a35dcc95dd24.
In three places I was using the intended string length rather than
the buffer length as the length argument, and in one place I was
assuming the return value was the number of non-null bytes copied,
which is not true when you are intentionally copying only part of
the source string. If strlcpy is supposed to be idiot-proof, they
clearly didn't anticipate the superior form of idiot I've proven
myself to be.
Chris 'BinGOs' Williams [Sat, 31 Dec 2011 01:11:58 +0000 (01:11 +0000)]
Update Version-Requirements to CPAN version 0.101021
[DELTA]
0.101021 2011-12-30 15:39:56 America/New_York
DEPRECATED DEPRECATED DEPRECATED DEPRECATED
Version::Requirements is now DEPRECATED
Use CPAN::Meta::Requirements, which is a drop-in replacement.
Father Chrysostomos [Sat, 31 Dec 2011 05:49:42 +0000 (21:49 -0800)]
perldelta up to a65c7ff
Father Chrysostomos [Sat, 31 Dec 2011 05:35:30 +0000 (21:35 -0800)]
perldelta for deparsing "foo::]"->...
Father Chrysostomos [Sat, 31 Dec 2011 05:32:50 +0000 (21:32 -0800)]
perldelta for tying empty %^H (
cb1f05e8f)
Father Chrysostomos [Sat, 31 Dec 2011 05:30:41 +0000 (21:30 -0800)]
perldelta for deparsing "string"->[0]
Father Chrysostomos [Sat, 31 Dec 2011 05:27:37 +0000 (21:27 -0800)]
perldelta for deparsing open("string")
Father Chrysostomos [Sat, 31 Dec 2011 04:44:57 +0000 (20:44 -0800)]
perldelta for safe %^H cloning
Father Chrysostomos [Sat, 31 Dec 2011 04:28:43 +0000 (20:28 -0800)]
Test ‘Too late for -[Mm]’
Father Chrysostomos [Sat, 31 Dec 2011 04:13:29 +0000 (20:13 -0800)]
[perl #104288] Die with ‘Unrecognized switch’ on #! line
If a switch such as -x that cannot occur on the shebang line is used
there, perl dies with ‘Can’t emulate...’.
If an unrecognised switch is used on the command line, perl dies with
‘Unrecognized switch...’.
It just happens to use the former error message for unrecognized
switches on the shebang line, which can be confusing:
$ perl -e '#!perl -G'
Can't emulate -G on #! line at -e line 1.
This commit changes it to output the same message as command-line
switches for those that are unrecognised wherever they are used.
Father Chrysostomos [Sat, 31 Dec 2011 00:04:57 +0000 (16:04 -0800)]
Fix bad reference in sv.h’s docs
SvPVx should be referring to SvPV as the faster version, not SvPVX.
Chris 'BinGOs' Williams [Fri, 30 Dec 2011 22:38:24 +0000 (22:38 +0000)]
Update Maintainers.pl for the CGI.pm update.
Chris 'BinGOs' Williams [Fri, 30 Dec 2011 22:37:19 +0000 (22:37 +0000)]
Update CPAN-Meta to CPAN version 2.113640
[DELTA]
2.113640 2011-12-30 15:19:46 America/New_York
- Version::Requirements has now been merged as CPAN::Meta::Requirements,
rendering Version::Requirements itself redundant
Father Chrysostomos [Fri, 30 Dec 2011 22:40:36 +0000 (14:40 -0800)]
[perl #107296] Document File::Glob’s default flags
Father Chrysostomos [Fri, 30 Dec 2011 22:05:23 +0000 (14:05 -0800)]
Increase $File::Glob::VERSION to 1.16
Ricardo Signes [Fri, 30 Dec 2011 18:23:11 +0000 (13:23 -0500)]
include a package POSIX::SigSet
this is mostly for the benefit of the PAUSE indexer, as described
in the code comment
Karl Williamson [Fri, 30 Dec 2011 16:49:36 +0000 (09:49 -0700)]
mktables: Fix bug for early Unicode versions
mktables should be able to be run on any Unicode version. Prior to this
patch, it was trying to do something unconditionally with the Name_Alias
property, whereas it should only do that on versions that have this
property.
Karl Williamson [Fri, 30 Dec 2011 16:11:23 +0000 (09:11 -0700)]
mktables: Change function name
This is in preparation for Unicode 6.1 for which the new name will make
more sense.
Karl Williamson [Fri, 30 Dec 2011 16:02:48 +0000 (09:02 -0700)]
mktables: Unicode 6.1 table will be empty
So, change mktables so it is expecting that.
Karl Williamson [Fri, 30 Dec 2011 16:02:02 +0000 (09:02 -0700)]
mktables: Add comment
Karl Williamson [Fri, 30 Dec 2011 15:55:49 +0000 (08:55 -0700)]
Unicode:UCD: Add comments
Karl Williamson [Fri, 30 Dec 2011 15:51:34 +0000 (08:51 -0700)]
Unicode::UCD: Extend prop_invmap() to handle multi-entries per code point
This is in preparation for Unicode 6.1, in which the Name_Alias property
has more than one entry in its table per code point.
Karl Williamson [Fri, 30 Dec 2011 15:47:48 +0000 (08:47 -0700)]
Unicode::UCD: white-space only
Karl Williamson [Fri, 30 Dec 2011 15:46:01 +0000 (08:46 -0700)]
Unicode::UCD: Allow prop_invlist() to work on non-compact tables
This accumulates adjacent ranges in case the table is not stored
compactly.
Karl Williamson [Fri, 30 Dec 2011 15:37:04 +0000 (08:37 -0700)]
Unicode::UCD: Special case prop_aliases() 'indic'
This is in preparation for Unicode 6.1 which adds a property whose name
begins with 'indic', as in Indic languages. This creates a conflict
with Perl's longstanding use of 'in' as a prefix in a property name.
"\p{infoo}" is a synonym for "\p{blk=foo}". The conflict is resolved
here with a special case.
Karl Williamson [Fri, 30 Dec 2011 04:54:42 +0000 (21:54 -0700)]
Unicode::UCD: Allow prop_invmap() to work on non-compact binary properties
It may be that the tables input to prop_invmap() are not in the most
compact form. There was a problem coping with these on binary
properties; which this commit fixes by realizing that a binary property
omits its 'Y' property value.
The .t does not pass these tests when the input is not compact, though.
Karl Williamson [Fri, 30 Dec 2011 04:48:35 +0000 (21:48 -0700)]
Unicode::UCD: Correct pod
The example is wrong; now the full script names are output, not the
short ones.
Shlomi Fish [Sat, 24 Dec 2011 20:03:58 +0000 (22:03 +0200)]
Add a test for "n EXPR()".
Shlomi Fish [Tue, 13 Dec 2011 17:03:06 +0000 (19:03 +0200)]
perl -d: add a test for s EXPR().
This patch involves addding another "perl -d" input script that contains a
subroutine that isn't called by default.
Shlomi Fish [Tue, 13 Dec 2011 16:41:32 +0000 (18:41 +0200)]
perl -d: add a test to the 's' command.
Conflicts:
lib/perl5db.t
Shlomi Fish [Sun, 11 Dec 2011 20:41:35 +0000 (22:41 +0200)]
Add a test for s.
Shlomi Fish [Sun, 11 Dec 2011 19:49:29 +0000 (21:49 +0200)]
perl -d: test for 'T'.
Shlomi Fish [Sat, 10 Dec 2011 15:23:04 +0000 (17:23 +0200)]
perl -d: add test for the x command.
Shlomi Fish [Sat, 10 Dec 2011 14:17:04 +0000 (16:17 +0200)]
Add a test for the debugger p command.
Chris 'BinGOs' Williams [Fri, 30 Dec 2011 15:35:54 +0000 (15:35 +0000)]
Update CGI to CPAN version 3.59
[DELTA]
Version 3.59 Dec 29th, 2011
[BUG FIXES]
- We no longer read from STDIN when the Content-Length is not set, preventing
requests with no Content-Length from freezing in some cases. This is consistent
with the CGI RFC 3875, and is also consistent with CGI::Simple. However, the old
behavior may have been expected by some command-line uses of CGI.pm.
Thanks to Philip Potter and Yanick Champoux. See RT#52469 for details:
https://rt.cpan.org/Public/Bug/Display.html?id=52469
[INTERNALS]
- remove tmpdirs more aggressively. Thanks to rjbs (RT#73288)
- use Text::ParseWords instead of ancient shellwords.pl. Thanks to AlexBio.
- remove use of define(@arr). Thanks to rjbs.
- spelling fixes. Thanks to Gregor Herrmann and Alessandro Ghedini.
- fix test count and warning in t/fast.t. Thanks to Yanick.
Eric Brine [Fri, 30 Dec 2011 05:55:40 +0000 (21:55 -0800)]
Clarify that \Q, \U, \L don't require \E
Signed-off-by: Abigail <abigail@abigail.be>
Nicholas Clark [Fri, 30 Dec 2011 13:23:53 +0000 (14:23 +0100)]
Note POSIX changes in perldelta.
Nicholas Clark [Fri, 9 Dec 2011 12:09:07 +0000 (13:09 +0100)]
Provide the correct POSIX return value for POSIX::dup2() on Win32.
Microsoft, in their wisdom, chose to ignore the POSIX spec when implementing
their dup2(), and have theirs return 0 on success, instead of the file
descriptor. It seems that no other vendor is this, um, "special", so code
the exception directly, as we don't run Configure on Win32, so there's little
point probing for this.
This resolves RT #98912.
Nicholas Clark [Thu, 8 Dec 2011 15:46:21 +0000 (16:46 +0100)]
Convert POSIX::sleep to an XS wrapper for PerlProc_sleep().
Previously it was a Perl wrapper for CORE::sleep, converting CORE::sleep's
return value of elapsed time slept into the POSIX return value of seconds
remaining. However, that approach could sometimes return a negative result
if CORE::sleep had slept for more than a second longer than the requested
time.
Karl Williamson [Fri, 30 Dec 2011 04:52:45 +0000 (21:52 -0700)]
perldelta for Unicode::UCD change
Karl Williamson [Fri, 30 Dec 2011 02:14:07 +0000 (19:14 -0700)]
Unicode::UCD: Use 6.1 format for Name_Alias
The Name_Alias property is changing significantly in Unicode 6.1. It
will have an extra value for each alias (and there will be a lot more
aliases). That extra value is considered informative, and gives the
type of alias this is. prop_invmap() should return both the alias and
its type. This commit changes it so that even in 6.0, it will return
what it is going to return in 6.1 (for the 6.0 aliases).
This function has not been released in a stable Perl version. Thus,
this gets the ultimate format into Perl before anyone has come to depend
on the earlier one.
squish
Karl Williamson [Fri, 23 Dec 2011 00:58:20 +0000 (17:58 -0700)]
regcomp.c: Silence valgrind warning
This happens only in doing debug output. Initialize these two debugging
variables
Karl Williamson [Tue, 27 Dec 2011 20:47:36 +0000 (13:47 -0700)]
mktables: Add comment
Karl Williamson [Tue, 27 Dec 2011 18:12:47 +0000 (11:12 -0700)]
mktables: Arrange to handle 6.1 ScriptExtensions.txt
Unlike 6.0, this property is listed in 6.1's PropValueAliases.txt, so is
automatically created in 6.1. Thus this needs to be checked for and the
initialization moved out of the constructor.
Karl Williamson [Tue, 27 Dec 2011 17:51:58 +0000 (10:51 -0700)]
mktables: Accommodate new @missings in Unicode 6.1
The Unicode 6.1 release candidate includes a new @missings line for the
ScriptExtenstions.txt file which we need to know about and deal with.
Karl Williamson [Tue, 27 Dec 2011 17:46:54 +0000 (10:46 -0700)]
mktables: Make protected method available
This is needed in 6.1 for a special case. It still should be used only
in limited circumstances so is still protected (meaning only that its
name begins with an underscore).
Karl Williamson [Tue, 27 Dec 2011 17:12:26 +0000 (10:12 -0700)]
mktables: Add capability to add dup FIFO
This is in preparation for Unicode 6.1. This adds the capability for a
duplicate code point to be added to a table after all existing
duplicates. Previously, new duplicates were added before any existing
ones.
Karl Williamson [Tue, 27 Dec 2011 16:57:53 +0000 (09:57 -0700)]
mktables: Rename variable
This is in preparation for Unicode 6.1. The variable means LIFO, and in
6.1, a FIFO mode will also be required, so indicate that the current
variable means LIFO.
Karl Williamson [Tue, 27 Dec 2011 02:30:45 +0000 (19:30 -0700)]
charnames.t: Skip null name test
In versions of Unicode earlier than 6.1, there was no possibility of a
name being empty here; but 6.1 will make that happen, so guard against
it.
Karl Williamson [Tue, 27 Dec 2011 02:28:03 +0000 (19:28 -0700)]
charnames.t: Add test names
This adds test names to nearly all the ones that were missing. Most
were done via a global substituted in a text editor.
Karl Williamson [Tue, 27 Dec 2011 02:23:53 +0000 (19:23 -0700)]
charnames.t: Fix test that passed whether or not it should
This test was calling grep, then the comma operator, then the non-empty
string after the comment operator caused it to always succeed.
Karl Williamson [Tue, 27 Dec 2011 00:38:44 +0000 (17:38 -0700)]
mktables: Comments only
Karl Williamson [Mon, 26 Dec 2011 20:05:10 +0000 (13:05 -0700)]
Unicode/UCD.t: Avoid early prototype messages
Add an ampersand before the calls to silence these warnings
Karl Williamson [Mon, 26 Dec 2011 19:45:17 +0000 (12:45 -0700)]
mktables: Ignore @missing lines in certain source files
This is in preparation for Unicode 6.1. The release candidate version of these files
changes the syntax of this line in several of its files, so that
mktables sees them as a problem. And, they aren't used anyway.
Karl Williamson [Fri, 23 Dec 2011 23:58:31 +0000 (16:58 -0700)]
perlunicode: nit
Karl Williamson [Mon, 26 Dec 2011 19:17:10 +0000 (12:17 -0700)]
mktables: Account for 2 provisional files
This is in preparation for Unicode 6.1. These two provisional
properties are available starting in 6.0, but the rest of the data base
did not refer to them, so we could safely ignore them. But in the beta
6.1, they are referred to elsewhere, which makes mktables generate hooks
for them, and then complains when those hooks are unused.
This patch should work no matter how 6.1 final comes out, and allows
someone who really wants to use these provisional properties to
recompile Perl without having to delve too far into the mktables code to
get it to work.
Karl Williamson [Mon, 26 Dec 2011 19:08:47 +0000 (12:08 -0700)]
mktables: Need to have two classes of unused files
The current code works only when skipped files aren't actually present.
Otherwise, it deletes them from a hash, which causes a wrong warning to
be output. By creating a separate hash for skipped files, we avoid
that.
Karl Williamson [Mon, 26 Dec 2011 18:51:21 +0000 (11:51 -0700)]
mktables: Consistently store lc of key in hash
In most places, the keys to this hash are stored in lowercase, but not
formerly this one. This enables not having to lc in other places.