Front page | perl.perl5.porters |
Postings from November 2000
perl@7892
From:
Jarkko Hietaniemi
Date:
November 27, 2000 07:21
Subject:
perl@7892
Message ID:
20001127092114.A7260@chaos.wustl.edu
http://www.iki.fi/jhi/perl@7892.tgz
http://www.hietaniemi.org/jarkko/perl@7892.tgz
ftp://ftp.funet.fi/pub/languages/perl/snap/perl@7892.tgz
Changes since the last:
____________________________________________________________________________
[ 7891] By: jhi on 2000/11/27 14:57:23
Log: Subject: Re: perlfaq style changes
From: sthoenna@efn.org (Yitzchak Scott-Thoennes)
Date: Sat, 25 Nov 2000 23:32:26 -0800
Message-ID: <KyLI6gzkgiRK092yn@efn.org>
plus a note from Ronald Kimball.
Branch: perl
! pod/perlfaq.pod pod/perlfaq1.pod pod/perlfaq6.pod
____________________________________________________________________________
[ 7890] By: jhi on 2000/11/27 14:53:14
Log: Subject: [PATCH perl@7825] Re: [ID 20001122.006] weird behaviour of $|
From: Benjamin Holzman <bah@ecnvantage.com>
Date: Sun, 26 Nov 2000 16:27:33 -0500
Message-ID: <20001126162733.J25040@ecnvantage.com>
Branch: perl
! gv.c
____________________________________________________________________________
[ 7889] By: jhi on 2000/11/27 14:43:11
Log: Subject: [ID 20001127.004] White space problem in perlamiga.pod
From: lvirden@cas.org
Date: Mon, 27 Nov 2000 09:32:06 -0500 (EST)
Message-Id: <200011271432.JAA09550@lwv26awu.cas.org>
Branch: perl
! README.amiga
____________________________________________________________________________
[ 7888] By: jhi on 2000/11/27 14:32:34
Log: Restore also the locale test to no-sprintf-taint state.
Fixes the bug 20001127.003.
Branch: perl
! sv.c t/pragma/locale.t
____________________________________________________________________________
[ 7887] By: gsar on 2000/11/27 14:13:05
Log: integrate changes#6613..6616,6620..6665 from mainline
VMS configure.com update continues.
Subject: Test fails / warnings with perl-current #6612
Subject: [PATCH] @+, @- readonly (was Re: @<punct> interpolating in "")
Subject: Re: [ID 20000807.003] [PATCH] Debugger treatment of condition "0"
For now remove the mail code.
Subject: Re: [PATCH] @+, @- readonly
Subject: warning: storage class after type is obsolescent
Subject: sfio2000
Subject: Re: File::Temp problems on VMS in bleedperl
README.os2 update.
Subject: Re: [PATCH perl-current] Make op/sprintf.t more comprehensive,
Make the user to give up his firstborn, err, to knowingly
verify installing an unstable developer release. Also bump
the release to 5.7.0, but leave a patch tag in the local
patches saying that this is not yet the real thing.
Update (kinda) to Test 1.14, from Joshua Pritikin.
make ok etc also for win32.
Subject: [ID 20000815.005] [PATCH] perldoc not looking in the right place for script pod
Don't blow limited stacks, a lower number is enough to
tickle the lookbehind limit.
Use -Dusedevel; regen Configure and the respective Porting stuff.
Subject: [PATCH] debugger exit code should reflect user exit code
Subject: [PATCH perl@6620] cygwin port
Missed a file from #6638.
Subject: [PATCH] for t/lib/peek.t (was Re: [ID 20000814.005] Not OK: perl v5.6.0 on i686-linux-64int 2.2.13)
magic callbacks all need to have same type signature
Subject: [ID 20000815.014] [PATCH] INSTALL doesn't mention 64 bit support.
Fix a dependency problem.
Subject: [PATCH: 6640] VMS Makefile.SH update
The numeric locale was reset to "C" by s?printf and never restored.
Subject: [ID 20000809.003] setlocale(LC_NUMERIC...) produces different results in 5.005 and 5.6
No test since adding the failing example to locale.t
does not fail -- probably because the locale settings are so
thoroughly tweaked by that time. Running the example standalone
does fail, though. UPDATE: test case added at change #7540.
Subject: [ID 20000324.040] minor fix to perlhpux.pod
Update to CPAN 1.57.
Subject: [PATCH] Cwd::_backtick_pwd does not check return value
Change the perlbug address to perl.org since it's more forgiving.
Change the regx compilation error markers to use = instead of <
since pod makes using the latter quite messy. Reported in
ID 20000814.006 by Abigail and in
Subject: Unknown escape E<> ?
Update to perldebtut 1.9, from Richard Foley.
check that the number pseudo children doesn't exceed
MAXIMUM_WAIT_OBJECTS, which is currently 64 (avoids overflowing
the WaitForMultipleObjects() limit that would cause wait()
to crash)
wait() and waitpid() could potentially be rewritten to use
more than one thread to do the waiting to eliminate this
limitation
change#6328 could make close(SOCKET) return false on windows
when it shouldn't
pod nit seen in passing
on windows, the return values from wait() and waitpid() don't
match those of pseudo-pids
waitpid() now handles externally spawned pids correctly;
fixes for backtick/wait/waitpid failures on Windows 9x
these changes make the pid returned by process functions on
Windows 9x always positive by clearing the high bit (which
is always set on Win9x); pseudo-process PIDs are likewise
always negative now on Win9x (just as on NT/2000)
trailing new %ENV entries weren't being pushed into the real
environment of subprocesses on Windows
Tweak the regex compilation errors once more.
avoid warnings from dense compiler
add "ok" targets from change#6632 in makefile.mk
Branch: maint-5.6/perl
- lib/Pod/PlainText.pm vms/configure.com
!> (integrate 66 files)
____________________________________________________________________________
[ 7886] By: jhi on 2000/11/27 13:54:18
Log: __FUNCTION__ isn't portable and trying to emulate it leads
into practically spelling out the name of the function.
Takes care of the bug 20001127.001.
Branch: perl
! perlio.c
____________________________________________________________________________
[ 7885] By: gsar on 2000/11/27 13:53:18
Log: integrate changes#6540..6541,6546..6549,6552..6554,6557..6606,
6610..6611 from mainline
Make regular expression parse error messages easier to understand.
Subject: Re: enhanced(?) regex error messages
Tiny tidying on report_evil_fh().
Subject: Re: enhanced(?) regex error messages
plus Capitalize the error messages, plus perldiag them.
Subject: Patch against 5.6.0 to allow "-d:Module=arg,arg,arg"
Document here-doc better.
Subject: [ID 20000807.003] [PATCH] Debugger treatment of condition "0"
Subject: [PATCH] Re: [ID 20000807.008] Double reads considered evil? (deja vu)
Do away with array context, from Daniel Chetlin <daniel@chetlin.com>
(either perlbug or p5p ate the original), plus regen
perlapi and perltoc.
Regen global.sym.
Double check that we have a dirhandle.
Subject: Re: enhanced(?) regex error messages
(plus two small patches sent privately)
(this still seems to leave few test failures)
warn is a macro, avoid using at a variable to avoid warnings
in some configurations; readdir.t is too conservative in
estimating number of *.t's
Get back into sync with Jeffrey on the enhanced regex warnings.
Subject: [PATCH 5.6.0] cygwin port
Zero entries were skipped, fix from Adrian Goalby
<argoalby@yahoo.co.uk>
Subject: Remove dead entry in perldiag
Amend the description of Perl6.
Subject: [PATCH Perl-5.6.0] perlfaq1.pod
detypo
It's the 2ndO'ROSSC.
Revert the sv.c part of #6559, a better fix is needed.
Iterating perl6 description.
Update to Term::ANSIColor 1.03, from Russ Allbery.
Update to Getopt::Long 2.23_05, from Johan Vromans.
Small AUTHORS and MAINTAIN updates. Could do with big updates.
Update to Pod::Parser 1.17, from Brad Appleton.
Update to CPAN 1.56, from Andreas König.
Update to CGI 2.70, from Lincoln Stein.
Put back the std @INC thing.
Fixes to looking-like-number to keep behaviour as it was in 5.005_03.
Subject: Re: [ID 20000810.002] $a["1foo"] same as $a[0]
Document the IO::Select timeout.
sleep(1) does not necessarily return 1.
Subject: [PATCH bleadperl] op/lex_assign.t
Subject: debugger "d" command doesnt check line number
B::Deparse didn't do sub attributes.
Subject: B::Deparse was Re: [ID 20000808.005] refs to returned lvalues are lvalues??
Preprocessing and postprocessing for File::Find.
Subject: Patch to Find::File.pm to allow alphabetical results
Subject: Re: [ID 20000809.005] trouble with long string and /m modifier - uninitialized value
Subject: Re: [ID 20000809.006] Debugger lost the ability to see $1 et al
Subject: Re: [ID 20000730.003] utf8::length() bad
Subject: Getting perlio and threads to compile
(the Solaris version changes in Configure skipped)
Tests for #6589.
Subject: Re: B::Deparse was Re: [ID 20000808.005] refs to returned lvalues are lvalues??
Add Perl debugging tutorial, regen toc.
Subject: perldebtut.pod
Add a few missing files, update MANIFEST.
Rewrite of vms/subconfigure.com as configure.com,
from Peter Prymmer and the vmsperl crew.
Should have deleted this in #6603.
Fix the test for 5005threads.
Fix-n-skip the tests under 5005threads.
Subject: [PATCH] t/op/regmesg.t fails if REG_INFTY set
Upgrade to CGI 2.71, from Lincoln Stein.
Branch: maint-5.6/perl
+> lib/CGI/eg/make_links.pl lib/CGI/eg/wilogo.gif
+> lib/Pod/PlainText.pm pod/perldebtut.pod t/lib/gol-oo.t
+> t/op/regmesg.t t/pod/find.t vms/configure.com
- vms/subconfigure.com
! lib/lib.pm
!> (integrate 115 files)
____________________________________________________________________________
[ 7884] By: jhi on 2000/11/27 13:44:33
Log: Subject: [PATCH perl@7795 (and earlier)] VMS test cleanup
From: lane@DUPHY4.Physics.Drexel.Edu (Charles Lane)
Date: Mon, 27 Nov 2000 08:23:54 EST
Message-Id: <001127082318.6381c@DUPHY4.Physics.Drexel.Edu>
Branch: perl
! vms/descrip_mms.template vms/test.com
____________________________________________________________________________
[ 7883] By: gsar on 2000/11/27 11:50:46
Log: integrate changes#6469..6484,6486..6501,6504..6505,6507..6509,
6511..6513,6515..6523,6525..6536
The swallow_bom() saga continues. The #23 of require.t
(UTF16-LE) still fails (silently, no output) but the #22
(UTF16-BE) seems to be working now. The root of the
failure may be in sv_gets(): is it UTF-16LE-aware,
especially when it comes to line endings?
Document the problem with -P in HP-UX and its workaround.
Subject: [PATCH] allow non-variable as lhs of non-updating tr///
(aka ID 20000730.002)
Subject: fix and question re: waitpid() under win32
Make the safety catch for buggy gccs work with triple version
numbers like 2.95.2. Reported in
Subject: [ID 20000731.005] Perl 5.6.0 "Configure" fails to recognize gcc 2.95.2
In Digital UNIX warn if gcc explicitly chosen because even
2.95.2 is known to cause problems.
Make chr() for values >127 to create utf8 when under utf8.
various syntax errors and such (not fixed: comp/require.t#22 coredump
on Windows)
Stash away the largefiles flags and libswanted.
BOM patching from Simon Cozens.
If gccosandvers is equal to osname, clear gccosandvers.
Make p4desc to skip non-mainperl branches by default.
Subject: [Proposed PATCH] Let Perl define QUAD_MIN and _MAX itself
The test from this
Subject: Re: [ID 20000411.002] qw() gives different results in 5.6 to previous versions
In new BSDs changes to argv[] do not show up in ps(1) output,
instead one must use setproctitle(). This was already addressed
by change #6457, but the below has a new variant for FreeBSD 4.0
or later, and the matter is also documented more.
FreeBSD 3.* updates from
Subject: [ID 20000801.007] setting $0 on FreeBSD 4.x does not get reflected in /bin/ps
regen_headers, regen perltoc.
Document in one place the memory abstractions used in Perl core.
memcpy has n o in it, as pinted ut by Sarathy.
Remove the extraneous "main::" prefix from all the
"opened only for", "on closed", and "never opened" warnings.
The name of a filehandle does not have <these>.
The tr utf8 patching continues.
The new setproctitle() feature is available only in
bleeding edge FreeBSD. From Paul Saab.
Subject: [PATCH bleadperl] [ID 20000731.010] regex error
Dump UVs as UVs in Data::Dumper.
detypo #6494
Document the IVdf UVuf UVof UVxf.
require.t needs binmode() to work on windows
Generate OP_IS_SOCKET() and OP_IS_FILETEST() macros
that are hopefully soon put into use.
Allow "no Module;" even if there is no 'unimport'.
Better skip message for the test; one of the two problems in
Subject: [ID 20000224.003] Not OK: perl v5.5.660 on i86pc-solaris 2.7
The subtest 4 may fail also on VOBS, as pointed out
by Nick Ing-Simmons in November 1999, bug id 19991124.003
(but the failure in that bug report isn't the subtest 4).
Be more informative on what is skipped and why,
also repeat the list at the end.
Add a URL for FSF.
Subject: [PATCH] sv.h documentation - SvLEN
Subject: [PATCH bleadperl] [ID 20000803.001] further regexp counting problems
Subject: [PATCH perl-current] Comings and goings in op/sprintf.t
Subject: [PATCH] bad cppsymbols on os2 + Configure question
Subject: [ID 20000802.002] [PATCH] memory pseudo-leak in sv_dump
Subject: [ID 20000802.004] Tests op/grent.t and op/pwent.t fail unnecessarily
mention the idea of @( and @)
This is 6512. Really.
Subject: [ID 19990721.004] Documentation bug in perlfunc
Subject: Minor tweak to perlvar.pod
In the warnings call filehandles consistently so;
add "unopened" warning for stat().
After the #6519 a warning about stat() is just that,
not about a filetest, which now have their own warning.
Subject: [ID 20000804.002] configure.gnu and arguments with whitespace characters
Subject: Re: Array vs. List context
Subject: New perlcc, take 2
Weed buglets pointed out by
Subject: Re: [ID 20000803.005] miniperl aborts during Perl make
gcc versions might have (parentheses) in them.
Subject: [ID 20000724.004] Perl interpreter segfault when using built-in flock
Essential prototype changes were missing from #6527.
Also make report_evil_fh() more bomb-proof.
Zap lib/Sys directory when cleaning up.
Change the Policy policy: now -Dprefix= with an existing
Policy.sh and prefix == siteprefix == vendorprefix, then all
of them follow along the new prefix.
Subject: Re: [ID 20000508.002] -Dprefix completely broken [PATCH]
Continue fixing the io warnings. This also
sort of fixes bug ID 20000802.003: the core dump
is no more. Whether the current behaviour is correct
(giving a warning: "Not a format reference"), is another matter.
Have symbols for the IoTYPEs.
Subject: [PATCH] perlfunc.pod use documentation (5.6.0)
Document a bit that UDP is not what you might think.
Subject: Re: IO::Socket::INET bug sending large UDP packets/fragmentation
tr memory corruption fix from Simon Cozens.
Plug the security hole described in the Aug 05 2000 bugtraq message
"sperl 5.00503 (and newer ;) exploit" by Michal Zalewski.
The security hole exists only in suidperls, which isn't
installed or even built by default.
Branch: maint-5.6/perl
!> (integrate 71 files)
____________________________________________________________________________
[ 7882] By: gsar on 2000/11/27 10:25:36
Log: integrate changes#6439..6444,6446..6453,6455..6457,6460..6465,6467..6468
from mainline
Make the unimplemented POSIX regex features [[.cc.]] and [[=c=]]
to be fatal errors (instead of by default ignoring them, and
ignoring with a bug: even though -w gave an error, the opening [
was left in) Reported in:
Subject: [PATCH: perl@6409] bug fix for munchconfig (turned up by CXX)
Subject: [PATCH] split /^/
MacOS nits from Matthias Neeracher.
More split() doc and test patches from Mike Guy.
Allow "sub AUTOLOAD;" to stop AUTOLOAD inheritance,
from Graham Barr in the module list.
docfix from Peter Scott <Peter@PSDT.com>.
File::Temp patches for VMS and OS/2 from Tim Jenness.
open() wariness in perlbug.
Subject: [PATCH] minor doc change - perlguts
Subject: Minor doc patch: handy.h
Be wary of close()s, too.
Further File::Temp patches from Yitzchak Scott-Thoennes
and Craig A. Berry.
Subject: [PATCH] fixes bug 20000508.004
Subject: [ID 19990709.002] [DOCUMENTATION PATCH] perldiag
Allow "no AutoLoader;", based on change #6444,
suggested by Graham Barr.
Use setproctitle() if available to modify $0.
Warn if the version of the operating system used to compile gcc
differs from the current version of the operating system.
Also display the gcc compilation os and version in myconfig.
Inspiration from
Tiny fixes for #6460.
The problem described in this
Subject: [ID 20000322.018] named chars aren't magical enough
has been fixed in perl 5.6.0 but just in case added a test
to keep it away. (The report from Joseph Hall.)
Tune the comments and hopefully stop a memory leak.
Subject: UTF8 concat
(with a memory leak fixed, plus a few casts added)
This also seems to help for
Subject: [ID 20000716.015] join UTF8 weirdness
Do not upgrade SVs into utf8 just because they participate
in eq or cmp. Reported and fix suggested in
Subject: [ID 20000720.009] sv_eq UTF8 bug
Fix the HALF_UPGRADE() macro introduced in #6263.
Find green threads before native threads.
Subject: Re: Patch to jpl/JNI/Makefile.PL
Branch: maint-5.6/perl
!> (integrate 30 files)
____________________________________________________________________________
[ 7881] By: jhi on 2000/11/27 05:48:41
Log: It seems that *both* the unused submatch loop cleanup
codes are needed.
Branch: perl
! regexec.c
____________________________________________________________________________
[ 7880] By: jhi on 2000/11/27 04:50:07
Log: Clean .exists deeper.
Branch: perl
! Makefile.SH
____________________________________________________________________________
[ 7879] By: jhi on 2000/11/27 02:31:35
Log: The code in regcppop() (see #7878) contains the correct lower
limit for the unused submatch 'cleanup' loop so that under
"use utf8" the following code wouldn't dump core:
"," =~ /([^,]*,)*/ With the the wrong lower limit (>=1)
the cleanup loop in regtry() stomped beyond allocated area
in the startp[] array. Therefore, copied the correct lower
loop limit (*PL_reglastparen) to regtry(). Note: something
may still not be quite right: why was the _higher_ loop limit
(prog->nparens) different in the utf8 case?
After this patch "./perl -Ilib -Mutf8 t/op/regexp.t" works
without core dumps, there were about 17 of them before
the patch (with us since Perl 5.7.0). Two failures, still:
496 and 505 (though these may not be severe).
Patch #7881 is also needed since both the cleanup loops
seem to be needed.
Also, the t/op/pat#44 seems to core dump under utf8.
Plus a couple of failures. UGH-8.
Branch: perl
! regexec.c
____________________________________________________________________________
[ 7878] By: jhi on 2000/11/27 02:21:17
Log: The unused submatch cleanup code in regtry() seems to be more crucial,
the code in regcppop() seems to be redundant for the test suite --
but it contains a germ of truth, and it needed for the build
process itself: see #7879 and #7881.
Branch: perl
! regexec.c
____________________________________________________________________________
[ 7877] By: jhi on 2000/11/27 01:53:09
Log: Comment on comment.
Branch: perl
! regexec.c
____________________________________________________________________________
[ 7876] By: jhi on 2000/11/27 00:49:59
Log: Adjust the docs to agree with #7875.
Branch: perl
! pod/perllocale.pod
____________________________________________________________________________
[ 7875] By: jhi on 2000/11/27 00:49:27
Log: Retract #7863. It makes more sense not to taint formatted output,
not even by sprintf().
Branch: perl
! sv.c
____________________________________________________________________________
[ 7874] By: jhi on 2000/11/26 23:36:02
Log: Debug dump of ANYOFUTF8 was garbage (data from ANYOF).
Not really fixed (should really dump the UTF-8 charclass),
but stopped displaying the garbage.
Also add a note on the (missing) Unicode PSXSPC and BLANK.
Branch: perl
! regcomp.c
____________________________________________________________________________
[ 7873] By: jhi on 2000/11/26 21:31:13
Log: BOUND regex opcodes (\b, \B) could try to scan zero length UTF-8.
Branch: perl
! regexec.c
____________________________________________________________________________
[ 7872] By: jhi on 2000/11/26 21:06:04
Log: No need to scan till infinity, 13 is enough.
Branch: perl
! handy.h utf8.c
____________________________________________________________________________
[ 7871] By: jhi on 2000/11/26 20:20:32
Log: Test line numbers are different with utf8.
Branch: perl
! t/op/re_tests
____________________________________________________________________________
[ 7870] By: jhi on 2000/11/26 20:10:12
Log: Message nit.
Branch: perl
! regcomp.c
____________________________________________________________________________
[ 7869] By: jhi on 2000/11/26 19:01:05
Log: Make utf8_length() and utf8_distance() (the latter of which
is unused at the moment) to be less forgiving about bad UTF-8.
Branch: perl
! embed.h embed.pl objXSUB.h proto.h utf8.c
____________________________________________________________________________
[ 7868] By: jhi on 2000/11/26 18:45:56
Log: Subject: Re: [ID 20001125.004] OK: perl v5.7.0 +DEVEL7825 on i686-linux 2.2.17 (UNINSTALLED)
From: Benjamin Holzman <bah@ecnvantage.com>
Date: Sat, 25 Nov 2000 20:40:20 -0500
Message-ID: <20001125204020.A25040@ecnvantage.com>
Debian allows installing /usr/bin/locale without installing
any locales, an error message will ensue.
Branch: perl
! t/pragma/locale.t
____________________________________________________________________________
[ 7867] By: jhi on 2000/11/26 18:31:12
Log: Subject: Re: [PATCH: perl@7825] SvTEMP-ness on rhs of aassign can wreak havoc
From: Benjamin Holzman <bah@ecnvantage.com>
Date: Sun, 26 Nov 2000 02:42:34 -0500
Message-ID: <20001126024234.G25040@ecnvantage.com>
Patch for the bug 20000212.002.
Branch: perl
! sv.c t/op/array.t
____________________________________________________________________________
[ 7866] By: jhi on 2000/11/26 18:28:09
Log: Subject: PATCH: File::Temp fix on WindowsNT/VMS
From: Tim Jenness <t.jenness@jach.hawaii.edu>
Date: Fri, 24 Nov 2000 18:38:34 -1000 (HST)
Message-ID: <Pine.LNX.4.21.0011241833230.18423-100000@lapaki.jach.hawaii.edu>
File::Temp 0.11.
Branch: perl
! lib/File/Temp.pm t/lib/ftmp-tempfile.t
____________________________________________________________________________
[ 7865] By: jhi on 2000/11/26 18:24:22
Log: The metaconfig unit change for #7864.
Branch: metaconfig/U/perl
! need_va_copy.U
____________________________________________________________________________
[ 7864] By: jhi on 2000/11/26 18:23:41
Log: Make the va_copy() test not to be so talkative.
Branch: perl
! Configure config_h.SH
____________________________________________________________________________
[ 7863] By: jhi on 2000/11/26 18:11:02
Log: Fix locale inconsistencies unearthed by Hugo's work.
Now the floating point sprintf really does taint the result
string as perllocale promises (has promised for a long time)
if "use locale" is in the lexical scope.
Branch: perl
! sv.c t/pragma/locale.t
____________________________________________________________________________
[ 7862] By: jhi on 2000/11/26 00:35:15
Log: Subject: [PATCH] Re: pp_add -> pp_i_add efficiency hack?
From: Nicholas Clark <nick@ccl4.org>
Date: Sun, 26 Nov 2000 00:07:50 +0000
Message-ID: <20001126000750.A22446@plum.flirble.org>
Infinit.
Branch: perl
! sv.c
____________________________________________________________________________
[ 7861] By: jhi on 2000/11/26 00:01:26
Log: Integrate perlio.
Branch: perl
!> doio.c perlio.c perliol.h pp_sys.c win32/makefile.mk
____________________________________________________________________________
[ 7860] By: nick on 2000/11/25 22:04:17
Log: Correct relative path from new ext\Filter\Util\Call location
back to miniperl.
Branch: perlio
! win32/makefile.mk
____________________________________________________________________________
[ 7859] By: nick on 2000/11/25 21:26:54
Log: Integrate mainline.
Branch: perlio
+> ext/Filter/Util/Call/Call.pm ext/Filter/Util/Call/Call.xs
+> ext/Filter/Util/Call/Makefile.PL t/lib/filter-util.pl
+> t/lib/filter-util.t
- ext/Filter/Util/Call.pm ext/Filter/Util/Call.xs
- ext/Filter/Util/Makefile.PL t/lib/filt-util.pl
- t/lib/filt-util.t
!> AUTHORS Configure MANIFEST README.amiga configure.com doio.c
!> embed.h embed.pl ext/B/B.pm ext/B/B/Lint.pm ext/B/B/Terse.pm
!> hints/solaris_2.sh hints/uwin.sh hints/vmesa.sh
!> lib/ExtUtils/MM_Unix.pm makedef.pl perl.h perlio.c perlio.h
!> perlsdio.h proto.h t/lib/io_sock.t win32/Makefile
!> win32/makefile.mk
____________________________________________________________________________
[ 7858] By: gsar on 2000/11/25 20:52:17
Log: C<foreach my $x ...> in pseudo-fork()ed process may diddle
parent's memory; fix it by keeping track of the actual pad
offset rather than a raw pointer (this change is probably also
relevant to non-ithreads case to avoid fallout from reallocs of
the pad array, but is currently only enabled for the ithreads
case in the interests of minimal disruption to existing "well
tested" code)
Branch: perl
! embed.h embed.pl global.sym objXSUB.h perlapi.c pp_ctl.c
! proto.h scope.c scope.h sv.c t/op/fork.t
____________________________________________________________________________
[ 7857] By: nick on 2000/11/25 20:01:28
Log: All tests pass on Win32/gcc/USE_PERLIO.
- when crlf layer is pushed make unix-level binary.
- remove :crlf injection in pp_backtick in pp_sys.c
- (Change dependacy of extension .dll's to perldll.def rather than perl.exe)
Branch: perlio
! doio.c perlio.c pp_sys.c win32/makefile.mk
____________________________________________________________________________
[ 7856] By: jhi on 2000/11/25 18:38:30
Log: Update the metaconfig units for #7855.
Branch: metaconfig/U/perl
! useperlio.U usesocks.U
____________________________________________________________________________
[ 7855] By: jhi on 2000/11/25 18:37:54
Log: Undo the SOCKS workarounds, instead start using PerlIO
if SOCKS is selected.
Subject: perl@7847, [ID 20001030.005], close-patch, perlio - The big cleanup
From: Jens Hamisch <jens@Strawberry.COM>
Date: Fri, 24 Nov 2000 18:31:30 +0100
Message-ID: <20001124183130.E28337@Strawberry.COM>
Subject: Re: perl@7847, [ID 20001030.005], close-patch, perlio - Patch the patch ...
From: Jens Hamisch <jens@Strawberry.COM>
Date: Fri, 24 Nov 2000 19:11:51 +0100
Message-ID: <20001124191151.A28753@Strawberry.COM>
Branch: perl
! Configure doio.c embed.h embed.pl makedef.pl perl.h perlio.c
! perlio.h perlsdio.h proto.h t/lib/io_sock.t
____________________________________________________________________________
[ 7854] By: jhi on 2000/11/25 17:39:08
Log: Undo #7848. Some of the code seems to use walkoptree(),
some walkoptree_slow(). An unfinished renaming?
Now the sub is walkoptree() (which is @EXPORT_OK),
the walkoptree_slow() is a typeglob alias to walkoptree.
This makes the tests to pass, at least.
Branch: perl
! ext/B/B.pm ext/B/B/Lint.pm ext/B/B/Terse.pm
____________________________________________________________________________
[ 7853] By: nick on 2000/11/24 22:24:33
Log: Re-arrange crlf vs binary for platforms that care.
crlf layer is now "the" buffer layer and can turn its behaviour on/off.
Lip-service to making stdio layer work on such platfroms (untested).
Now fails 3 tests rather than one, checkin to see if I can debug
the issue under linux. (Do not merge.)
Branch: perlio
! doio.c perlio.c perliol.h
____________________________________________________________________________
[ 7852] By: jhi on 2000/11/24 18:57:39
Log: AUTHORS edits.
Branch: perl
! AUTHORS
____________________________________________________________________________
[ 7851] By: jhi on 2000/11/24 17:34:56
Log: Add Jan-Erik Karlsson.
Branch: perl
! AUTHORS
____________________________________________________________________________
[ 7850] By: jhi on 2000/11/24 17:27:56
Log: Configure should find stdchar on its own.
Branch: perl
! hints/solaris_2.sh
____________________________________________________________________________
[ 7849] By: jhi on 2000/11/24 17:24:55
Log: Subject: PATCH over 7848: Filter test flawed
From: andreas.koenig@anima.de (Andreas J. Koenig)
Date: 24 Nov 2000 10:28:29 +0100
Message-ID: <m3k89tbtxe.fsf@ak-71.mind.de>
Branch: perl
! t/lib/filter-util.t
____________________________________________________________________________
[ 7848] By: jhi on 2000/11/24 17:22:15
Log: Subject: [ID 20001124.001] B exports "walkoptree" but does not define it
From: Simon Cozens <simon@pembro4.pmb.ox.ac.uk>
Date: Fri, 24 Nov 2000 16:12:58 +0000
Message-Id: <E13zLT8-0005FN-00@pembro4.pmb.ox.ac.uk>
Branch: perl
! ext/B/B.pm
____________________________________________________________________________
[ 7847] By: jhi on 2000/11/24 03:07:01
Log: Integrate perlio.
Branch: perl
!> doio.c perlio.c perlio.h pp_sys.c t/lib/io_tell.t
!> win32/makefile.mk
____________________________________________________________________________
[ 7846] By: gsar on 2000/11/24 00:55:57
Log: integrate changes#6415..6418,6420..6438 from mainline
Fix the bitvector ops for utf8 (tricky since past 7 bits
the utf8 'characters' can be more than one octet).
MPE/ix updates for perl 5.6.0 from Mark Bixby.
Subject: SDBM_File documentation
Detypo.
Decutandpasto.
Send all installperl messages to STDERR and be -w clean.
Out-of-date note removed.
Protect against "wild next"s, that is, callbacks doing "next"
instead of "return".
Use STDOUT consistently.
The output might have been produced in the wrong order.
A missing 'break' after the [[:space:]] switch case.
Add tests for
[ID 19991110.003] another matching finding by pcre author
which has already been fixed by some patch, as verified in
Documentation to explain the behaviour of map().
Add an optimization for map-maps-a-list-element-to-more-list-elements
case, but add also notes explaining the relationship of this
patch and the earlier notes by Sarathy.
Subject: [ID 20000716.023] syslog test fails without sockets
Subject: Re: [PATCH] [ID 20000716.011] strangeness with split($_ =~ m/.../)
Test cases for #6431.
File::Spec::VMS fixup for tmpdir from Craig Berry.
Make the "uninit variable" warning to say "concat or string"
or "join or string" when in concat or join .
Get UTF16 BOMs working. Patch from
Subject: Re: [ID 20000719.001] Problem with bleadperl of 7/18/00
Subject: [PATCH] Make large file tests deal with SIGXFSZ
Subject: [ID 20000724.003] Documentation changes for perllocale.pod
Subject: [PATCH] av.c apidoc
Branch: maint-5.6/perl
!> (integrate 43 files)
____________________________________________________________________________
[ 7845] By: gsar on 2000/11/24 00:20:45
Log: integrate changes#6406..6414 from mainline
Merge perlhacktut into perlhack, update perlguts.
Fix AutoSplit to use File::Spec the right way in VMS,
from Peter Prymmer.
The bug report
[ID 19991110.002] minimal matching discrepancy found by pcre author
seems to have been fixed (though differently from what was suggested
in the report) in 5.6.0. Add tests to keep the bug from reappearing.
thinko fix in vms/descrip_mms.template, the win32.pod in lib,
not in pod, from Peter Prymmer
Subject: [docpatch] Re: [ID 19991002.011] perldoc -f shift
From: Hugo <hv@crypt.compulink.co.uk>
Date: Fri, 14 Jul 2000 23:05:20 +0100
Message-Id: <200007142205.XAA17882@crypt.compulink.co.uk>
Didn't anymore apply, but that point still could use another fix.
lib/b test fixes from Peter Prymmer.
More docs for sv functions.
perlvms.pod whitespace cleanup to keep pod utils happy.
another VMS build tweak from Peter Prymmer
Branch: maint-5.6/perl
!> embed.pl lib/AutoSplit.pm pod/perlapi.pod pod/perlfunc.pod
!> pod/perlguts.pod pod/perlhack.pod sv.c t/lib/b.t t/op/re_tests
!> vms/descrip_mms.template vms/perlvms.pod
____________________________________________________________________________
[ 7844] By: nick on 2000/11/23 23:42:45
Log: Win32/perlio Now just fails one io/argv.t test - lack
of default :crlf on standard streams.
Branch: perlio
! doio.c perlio.c
____________________________________________________________________________
[ 7843] By: nick on 2000/11/23 23:04:05
Log: Win32 passes all but t/lib/peek.t with perlio and home-grown crlf.
peek fail is showing a real problem (multiple crlf layers
are getting pushed.)
Branch: perlio
! doio.c perlio.c win32/makefile.mk
____________________________________________________________________________
[ 7842] By: nick on 2000/11/23 19:46:23
Log: Implement PerlIO_binmode()
Fix PerlIOCrlf_unread() (*--ptr rather than *ptr-- ...)
Test on UNIX with PERLIO="perlio crlf" to mimic Win32,
make binmode in t/lib/io_tell.t unconditional so that works.
Checkin just so Win32 machine can see these changes.
Branch: perlio
! doio.c perlio.c perlio.h pp_sys.c t/lib/io_tell.t
____________________________________________________________________________
[ 7841] By: jhi on 2000/11/23 19:30:00
Log: README.amiga changes from Jan-Erik Karlsson <trg@privat.utfors.se>.
Branch: perl
! README.amiga
____________________________________________________________________________
[ 7840] By: jhi on 2000/11/23 17:18:02
Log: SOCKS wrestling continues, patches from Jens Hamisch.
Branch: perl
! doio.c embed.h embed.pl perlsdio.h proto.h
____________________________________________________________________________
[ 7839] By: jhi on 2000/11/23 16:54:14
Log: Push Filter::Util::Call one level deeper.
Branch: perl
+ t/lib/filter-util.t
+> ext/Filter/Util/Call/Call.pm ext/Filter/Util/Call/Call.xs
+> ext/Filter/Util/Call/Makefile.PL t/lib/filter-util.pl
- ext/Filter/Util/Call.pm ext/Filter/Util/Call.xs
- ext/Filter/Util/Makefile.PL t/lib/filt-util.pl
- t/lib/filt-util.t
! MANIFEST configure.com hints/uwin.sh hints/vmesa.sh
! win32/Makefile win32/makefile.mk
____________________________________________________________________________
[ 7838] By: jhi on 2000/11/23 16:52:40
Log: Reach back one higher up when searching for PERL_SRC.
Branch: perl
! lib/ExtUtils/MM_Unix.pm
____________________________________________________________________________
[ 7837] By: jhi on 2000/11/23 16:05:41
Log: Subject: Re: perl@7826
Date: Thu, 23 Nov 2000 02:20:25 -0600
From: "Craig A. Berry" <craigberry@mac.com>
Message-Id: <p04330102b6424b6cd4c9@[172.16.52.1]>
Make VMS to like Filter::Util::Call.
Branch: perl
! configure.com ext/Filter/Util/Makefile.PL
____________________________________________________________________________
[ 7836] By: nick on 2000/11/23 16:04:08
Log: Implement crlf layer - not ready for merge.
Branch: perlio
! perlio.c win32/makefile.mk
____________________________________________________________________________
[ 7835] By: jhi on 2000/11/23 15:56:52
Log: STDCHAR is different in Sparc v9 vs x86.
Branch: perl
! hints/solaris_2.sh
____________________________________________________________________________
[ 7834] By: jhi on 2000/11/23 15:50:47
Log: The type needs to be visible to protos.
Branch: perl
! perl.h
____________________________________________________________________________
[ 7833] By: jhi on 2000/11/23 15:40:46
Log: Finish unfinished SOCKS workaround changes.
Branch: perl
! doio.c
____________________________________________________________________________
[ 7832] By: jhi on 2000/11/23 15:33:30
Log: Integrate perlio.
Branch: perl
!> win32/config_H.bc win32/config_H.gc win32/config_H.vc
!> win32/makefile.mk
____________________________________________________________________________
[ 7831] By: nick on 2000/11/23 10:31:07
Log: Correct makefile.mk dependancies for ext/Filter/Util
Build ext/Encode
Run dmake regen_config_h
Branch: perlio
! win32/config_H.bc win32/config_H.gc win32/config_H.vc
! win32/makefile.mk
____________________________________________________________________________
[ 7830] By: nick on 2000/11/23 08:05:07
Log: Integrate mainline
Branch: perlio
+> ext/Filter/Util/Call.pm ext/Filter/Util/Call.xs
+> ext/Filter/Util/Makefile.PL lib/Filter/Simple.pm
+> t/lib/filt-util.pl t/lib/filt-util.t
!> (integrate 41 files)
____________________________________________________________________________
[ 7829] By: gsar on 2000/11/23 03:07:11
Log: build the Filter extension on windows
Branch: perl
! win32/Makefile win32/makefile.mk
____________________________________________________________________________
[ 7828] By: gsar on 2000/11/23 02:46:35
Log: regenerate win32/config* and fix some inconsistencies in the
makefiles; always export Perl_deb() because it is needed
by ext/re/* (whether perl itself is built with -DDEBUGGING
or not)
Branch: perl
! makedef.pl win32/Makefile win32/config_H.bc win32/config_H.gc
! win32/config_H.vc win32/makefile.mk
____________________________________________________________________________
[ 7827] By: gsar on 2000/11/23 02:18:38
Log: get sources building on windows+MSVC again (untested with other
compilers)
Branch: perl
! win32/include/sys/socket.h win32/perllib.c
____________________________________________________________________________
[ 7826] By: jhi on 2000/11/23 01:18:26
Log: Update Changes.
Branch: perl
! Changes patchlevel.h
____________________________________________________________________________
--
$jhi++; # http://www.iki.fi/jhi/
# There is this special biologist word we use for 'stable'.
# It is 'dead'. -- Jack Cohen