Commit Graph

640 Commits

Author SHA1 Message Date
tg 65f0044ceb make it possible for the user to specify mksh_siglist and mksh_signame
in CPPFLAGS, thus overriding the system default
(think standard OS with different libc)
2006-08-01 13:19:42 +00:00
tg e65d9edc11 * use strsignal(3) not strerror(3), oops
* on GNU, that needs -D_GNU_SOURCE
* a little style(9) KNF while here
2006-08-01 12:57:07 +00:00
tg d121622e84 remove GNUish 'check' target, we always use 'regress' 2006-08-01 12:46:01 +00:00
tg db107a9b05 bring back the 'version' editing command at "ESC ^V" like AT&T ksh93+r 2006-08-01 12:44:17 +00:00
tg 711496f6fb move version definition into extra file 2006-08-01 12:22:26 +00:00
tg 3a9ac3916a * add new book reference from #ksh (freenode) topic
* add another ksh book, referenced from amazon
* fix reference mdoc code
* add page count for the books
* fix old 10-digit ISBNs
* add new 13-digit ISBNs
* add myself as current maintainer/developer
* add miros-discuss eMail address for bug reports
* note that no book documents mksh
2006-07-27 14:45:30 +00:00
tg 85a4745978 switch /bin/sh to mksh
this at least gets us rid of a bunch of segfaults on freewrt openssl builds
2006-07-23 19:10:59 +00:00
tg f1e8f970c4 * get dot.mkshrc sample from Debian mksh-27.4-2
* sync dot.mkshrc sample with FreeWRT r383
* add some stuff from src/etc/profile and contrib/samples/etc_profile
* sync the latter two with reality and shorten
* teach the user in the mksh installation instructions to install the sample

NB: the sample is also covered by the MirOS licence
2006-07-23 18:44:22 +00:00
tg 80c579a758 dates 2006-07-23 14:35:44 +00:00
tg afcbdad1c5 simplify, rid use of .Mx macro (portability) and direct users to miros-disgust@ 2006-07-23 14:35:10 +00:00
tg 2ffaa5ac46 while mksh isn't designed to run as sh(1) it _can_ do it,
and in fact, Han Boetes even uses it as /bin/sh on CRUX GNU/Linux
2006-07-21 21:47:43 +00:00
tg 64598758fb * TNF needs more to silence char subscript warnings (on 3.0-stable)
* bump vsn
2006-07-11 14:51:01 +00:00
tg f54064509e Certain Sourcemage GNU/Linux people like to build (cast?) as root...
adjust regression testsuite so that it doesn't fail w.r.t. $PS1
2006-07-07 09:52:20 +00:00
tg d854fb1a8d fix date/version 2006-07-03 12:19:43 +00:00
tg 8f2400d002 allow Test.sh to be passed arguments 2006-07-03 12:17:08 +00:00
tg bc93d1c90d fix the "char subscripts" issue, bump version 2006-07-03 12:16:31 +00:00
tg c9953f0ace adding -Wno-char-subscripts for Solaris is wrong
- I saw it on NetBSD®/alpha too
- these ought to be fixed

I've got to investigate that later.
2006-07-01 14:39:41 +00:00
tg fc87ecb1dd add a little portability 2006-06-23 15:05:39 +00:00
tg ab231f3595 new conditional: HAVE_ARC4RANDOM_PUSH 2006-06-21 19:35:38 +00:00
tg 569cf64ff1 * process ~/.mksrc only if FTALKING (i.e. interactive shell)
From: hondza <miscreant@tiscali.cz>
* document that in the manual page myself
* add regression test for that myself
2006-06-21 19:27:35 +00:00
tg 51d6772800 move -D_FILE_OFFSET_BITS=64 (not on interix tho, sadly) - XXX where else?
and -DNEED_COMPAT
2006-05-27 20:33:01 +00:00
tg 7bc318156c we only need to check for error, not for 'error code 1',
that's sufficient for make; Solaris 10 gives us an exit
code of 255 (exit status 65280) here according to Tonnerre
2006-05-27 11:36:50 +00:00
tg ea102ef66e fix build under Solaris 10
From: Tonnerre LOMBARD <tonnerre@bsdprojects.net>

doesn't break Solaris 8 on stinky.trash.net (tested)
2006-05-27 11:30:07 +00:00
tg daa40b4e45 this is mksh R27, tested on
* MirOS HEAD (i386, gcc 3.4)
* Debian GNU/Linux 3.1 (i386, gcc 3.3)
* Interix 3.5 (i386, gcc 3.3) - perl too old for regression tests
* GNU/Cygwin 2006-03-* (i386, gcc 3.4) - no perl installed
* Solaris 8 (sparc64, gcc 3.4)

no testing been done on
* Mac
* other BSDs
* gcc 4.1.1
* Solaris 10
because I asked for community feedback but...
2006-05-26 23:36:19 +00:00
tg 576e19b39e need confstr proto on Debian 2006-05-26 22:24:20 +00:00
tg 14a65d55e6 pull in last OpenBSD diff - I must admit I neither fully understand
the diff nor the commit message, but they're probably knowing what
they're doing

on an unrelated side note, mksh-beta.cpio.gz didn't get tested by
many people... so mksh R27 will be released without much more...

Set the index of item[0], since it might not have been set before
(A=0; A[1]=1) and do not use a static buffer in str_val, since the
results might be used in a loop. Report from jared r r spiegel; help
from Matthias Kilian; ok beck@
2006-05-26 22:17:21 +00:00
tg f4ea190e65 why do I suddenly need -D_BSD_SOURCE on GNU where I didn't before? 2006-05-10 20:34:28 +00:00
tg 8297ca20c8 * this is almost mksh R27, bump
* add the O'Reilly books as references
2006-05-10 19:30:33 +00:00
tg 7672b9b346 apply some fixes from OpenBSD and don't apply some others
but sync RCS IDs for easier future adaption:
* Simplify savefd() by removing the "noclose" flag and make noclose
  behavior the default. Almost all uses of savefd() are followed
  by an implicit or explicit close.
* fix typos
* might as well make ksh_getopt() match real getopt(), ie. get rid of that
  stupid EOF concept that was never true. adobriyan@gmail
* use SEEK_* for lseek()
* fix lint comments, no functional changes
* remove excessive optimization; from adobriyan@gmail
* only santa checks things twice; from adobriyan@gmail
* Interpret zero-filled numbers as decimal; PR 4213; from Alexey Dobriyan
2006-05-10 18:54:13 +00:00
tg 301066165b auto-test check.t:mkshrc-1 as well (verified to work) 2006-05-08 12:18:49 +00:00
tg e28549e4ff illustrate creative use of $PS1
requested by Jari Aalto
2006-05-08 12:12:36 +00:00
tg b0eb8f5b5e * check.t: add two mkshrc-tests (turn on, turn off)
* check.pl, check.t: add mkshrc-test whether "real" $HOME/.mkshrc interfers
* mksh.1: sync with reality: $ENV unset isn't needed, $ENV='' is enough
  to turn on ~/.mkshrc processing
2006-05-08 11:59:41 +00:00
tg c4ef5b801a New feature: read ~/.mkshrc during startup, after the profiles
have been read, for non-priviledgued shells only. If $ENV is
set, ~/.mkshrc is ignored (even if the file pointed to by $ENV
does not exist in the first place).
Feature requested by Jari Aalto for portable mksh because some
operating systems' vendors do not allow touching the profile.

Initial diff and manual page addition by Jari Aalto <jari.aalto@cante.net>
Code changes and manual page correctness by me.
2006-05-08 11:42:36 +00:00
tg 41f310751e un-tabulate function declaration (style) 2006-05-08 11:07:38 +00:00
tg 5726a50ab9 * shuffle around _*_SOURCE definitions (sh.h -> Build.sh), per OS
* Tonnerre Lombard reports that Solaris 10 needs _BSD_SOURCE
2006-03-19 18:06:32 +00:00
tg 9e838e765c * convert most of the interna of edit.c to unsigned where it's
important for the signedness of the target, to enable using
  high-bit7 characters on the command line (eg. in UTF-8 locales)
* make emacs-usemeta and vi-show8 equivalent and document that
  in the manual page (i.e. beware if you're in a latin1 locale)
* do not reference SUSv3 (helps GNU groff), write \&" out instead
2006-02-16 11:48:32 +00:00
tg 9d6fc21153 better style 2006-01-30 12:39:14 +00:00
tg 3f2912d717 * update to mksh R26c
* there is no comma in front of "which" in English, I was taught
2006-01-30 12:37:24 +00:00
tg eb33e02f8e this can be released as mksh-R26c, works on Solaris, Interix (whose Perl
is too old for the testsuite), Debian and of course MirOS
2006-01-29 20:56:55 +00:00
tg e7f1c2c8c1 forgot one use_rand, found on Solaris 2006-01-29 20:16:51 +00:00
tg a09f40cacd optimise the non-arc4random case while not touching the default case 2006-01-29 20:10:16 +00:00
tg 94ee3b388a sync with OpenBSD: remove unused code and vars; name clash with libc
compiles under MirBSD
2006-01-29 20:04:54 +00:00
tg 7ca34b3b12 mirbsd 2006-01-27 10:59:30 +00:00
tg 18265d10f5 * add a few handy tricks (IFS, -r, -p and co-processes) to the description
of the 'read' builtin (it's not exactly where they belong, but hey, this
  man page is already better than the O'Reilly book)
* replace all | with \*(Ba
2006-01-27 10:58:36 +00:00
tg 8301a9bd51 flesh out regression test into an on-the-fly-generated ./Test.sh
so that the user doesn't have to copy and paste a super-long line

idea by Han Boetes and others
2006-01-27 00:50:12 +00:00
tg 0addfc1c92 * replace /*...*/ form of licence with easier # form (no practical change)
* bump to 2006
2006-01-27 00:44:16 +00:00
tg 67ce90703e sync licence 2005-12-06 18:14:27 +00:00
tg d347cdc4d8 sync with reality 2005-11-24 19:42:38 +00:00
tg 8226efa566 use \t not \n for separating RCS IDs
ident(1) finds both similarily, and as doesn't bitch
2005-11-24 19:39:10 +00:00
tg 95cfad6339 * only have one $MirOS RCS ID per file to shrink source size
(this is an exception from normal use)
* bump to R26
2005-11-22 18:40:44 +00:00