Upgrade to mksh R56b.
R56b is a bugfix-only release everyone should upgrade to:
[tg] Reference the FAQ webpage
[panpo, Riviera] Fix documentation bug wrt. Esc+Ctrl-L
[tg, Larry Hynes] Fix “0” movement in vi mode
[tg] Replace broken libcs’ offsetof macro with MirBSD’s
R56 is a bugfix release with some experimental fixes:
[tg, Seb] Do not apply alias name restrictions to hash/tilde tracking
[tg] Restore ‘.’, ‘:’ and ‘[’ in alias names (“[[” is still forbidden)
[tg] Fix accidentally defanged $PATHSEP test
[tg] On ^C (INTR and QUIT edchars), shove edit line into history
[iSKUNK, tg] Begin porting to z/OS using EBCDIC encoding, incomplete
[tg] Redo fast character classes code, adding POSIX and other helpers
[tg] bind parses backslash-escaped ‘^’ (and ‘\’) as escaped
[tg] Building with -DMKSH_ASSUME_UTF8=0 no longer causes a known failure in the testsuite
[tg] New test.sh option -U to pass a UTF-8 locale to use in the tests
[tg] re_format(7) BSD: [[ $x = *[[:\<:]]foo[[:\>:]]* ]]
[tg, iSKUNK] Use Config in check.pl only if it exists
[tg] New matching code for bracket expressions, full POSIX (8bit)
[komh] Exclude FAT/HPFS/NTFS-unsafe tests on OS/2 (and Cygwin/MSYS)
[tg] Update to Unicode 10.0.0
[tg, selk] Make readonly idempotent
[tg, multiplexd] When truncating the persistent history, do not change the underlying file, do all operations on the locked one; do not stop using the history at all if it has been truncated
[tg, Jörg] Turn off UTF-8 mode upon turning on POSIX mode
[Martijn Dekker, Geoff Clare, many on the Austin list, tg] In POSIX mode, make the exec builtin force a $PATH search plus execve
[tg] Fix GCC 7, Coverity Scan warnings
[tg, Michal Hlavinka] Track background process PIDs even interactive
[tg] Always expose mksh’s hexdump shell function; speed it up by working on the input in chunks; use character classes to make it EBCDIC safe
[tg] Revamp dot.mkshrc default editor selection mechanism
Bug: N/A
Test: builds and boots
Change-Id: I35d624c4352c1b7c1ee499f9680712a30c64323b
diff --git a/Android.mk b/Android.mk
index 829fc7b..5a53d9f 100644
--- a/Android.mk
+++ b/Android.mk
@@ -32,48 +32,89 @@
MKSH_INCLUDES := $(LOCAL_PATH)/src
+# Compiler flags...
MKSH_CFLAGS += \
-Wno-deprecated-declarations \
-fno-asynchronous-unwind-tables \
-fno-strict-aliasing \
-fstack-protector -fwrapv \
-# ...and CPPFLAGS.
+# ...various options we choose...
MKSH_CFLAGS += \
- -DDEBUG_LEAKS -DMKSH_ASSUME_UTF8 \
+ -DDEBUG_LEAKS \
+ -DMKSH_ASSUME_UTF8 \
-DMKSH_DONT_EMIT_IDSTRING \
+ -DKSH_VERSIONNAME_VENDOR_EXT=\"\ Android\" \
+
+# ...and the defines from Build.sh.
+MKSH_CFLAGS += \
-DMKSH_BUILDSH \
- -D_GNU_SOURCE -DSETUID_CAN_FAIL_WITH_EAGAIN \
- -DHAVE_ATTRIBUTE_BOUNDED=0 -DHAVE_ATTRIBUTE_FORMAT=1 \
+ -D_GNU_SOURCE \
+ -DSETUID_CAN_FAIL_WITH_EAGAIN \
+ -DHAVE_STRING_POOLING=1 \
+ -DHAVE_ATTRIBUTE_BOUNDED=1 \
+ -DHAVE_ATTRIBUTE_FORMAT=1 \
-DHAVE_ATTRIBUTE_NORETURN=1 \
-DHAVE_ATTRIBUTE_PURE=1 \
-DHAVE_ATTRIBUTE_UNUSED=1 \
- -DHAVE_ATTRIBUTE_USED=1 -DHAVE_SYS_TIME_H=1 -DHAVE_TIME_H=1 \
- -DHAVE_BOTH_TIME_H=1 -DHAVE_SYS_BSDTYPES_H=0 \
- -DHAVE_SYS_FILE_H=1 -DHAVE_SYS_MKDEV_H=0 -DHAVE_SYS_MMAN_H=1 \
- -DHAVE_SYS_PARAM_H=1 -DHAVE_SYS_RESOURCE_H=1 \
- -DHAVE_SYS_SELECT_H=1 -DHAVE_SYS_SYSMACROS_H=1 \
- -DHAVE_BSTRING_H=0 -DHAVE_GRP_H=1 -DHAVE_IO_H=0 -DHAVE_LIBGEN_H=1 \
- -DHAVE_LIBUTIL_H=0 -DHAVE_PATHS_H=1 -DHAVE_STDINT_H=1 \
- -DHAVE_STRINGS_H=1 -DHAVE_TERMIOS_H=1 -DHAVE_ULIMIT_H=0 \
- -DHAVE_VALUES_H=0 -DHAVE_CAN_INTTYPES=1 -DHAVE_CAN_UCBINTS=1 \
- -DHAVE_CAN_INT8TYPE=1 -DHAVE_CAN_UCBINT8=1 -DHAVE_RLIM_T=1 \
+ -DHAVE_ATTRIBUTE_USED=1 \
+ -DHAVE_SYS_TIME_H=1 \
+ -DHAVE_TIME_H=1 \
+ -DHAVE_BOTH_TIME_H=1 \
+ -DHAVE_SYS_BSDTYPES_H=0 \
+ -DHAVE_SYS_FILE_H=1 \
+ -DHAVE_SYS_MKDEV_H=0 \
+ -DHAVE_SYS_MMAN_H=1 \
+ -DHAVE_SYS_PARAM_H=1 \
+ -DHAVE_SYS_RESOURCE_H=1 \
+ -DHAVE_SYS_SELECT_H=1 \
+ -DHAVE_SYS_SYSMACROS_H=1 \
+ -DHAVE_BSTRING_H=0 \
+ -DHAVE_GRP_H=1 \
+ -DHAVE_IO_H=0 \
+ -DHAVE_LIBGEN_H=1 \
+ -DHAVE_LIBUTIL_H=0 \
+ -DHAVE_PATHS_H=1 \
+ -DHAVE_STDINT_H=1 \
+ -DHAVE_STRINGS_H=1 \
+ -DHAVE_TERMIOS_H=1 \
+ -DHAVE_ULIMIT_H=0 \
+ -DHAVE_VALUES_H=0 \
+ -DHAVE_CAN_INTTYPES=1 \
+ -DHAVE_CAN_UCBINTS=1 \
+ -DHAVE_CAN_INT8TYPE=1 \
+ -DHAVE_CAN_UCBINT8=1 \
+ -DHAVE_RLIM_T=1 \
-DHAVE_SIG_T=1 \
- -DHAVE_STRING_POOLING=1 \
- -DHAVE_SYS_ERRLIST=0 -DHAVE_SYS_SIGNAME=1 \
- -DHAVE_SYS_SIGLIST=1 -DHAVE_FLOCK=1 -DHAVE_LOCK_FCNTL=1 \
+ -DHAVE_SYS_ERRLIST=0 \
+ -DHAVE_SYS_SIGNAME=1 \
+ -DHAVE_SYS_SIGLIST=1 \
+ -DHAVE_FLOCK=1 \
+ -DHAVE_LOCK_FCNTL=1 \
-DHAVE_GETRUSAGE=1 \
-DHAVE_GETSID=1 \
-DHAVE_GETTIMEOFDAY=1 \
- -DHAVE_ISSETUGID=0 \
-DHAVE_KILLPG=1 \
- -DHAVE_MEMMOVE=1 -DHAVE_MKNOD=0 -DHAVE_MMAP=1 -DHAVE_NICE=1 \
- -DHAVE_REVOKE=0 -DHAVE_SETLOCALE_CTYPE=0 \
- -DHAVE_LANGINFO_CODESET=0 -DHAVE_SELECT=1 -DHAVE_SETRESUGID=1 \
- -DHAVE_SETGROUPS=1 -DHAVE_STRERROR=1 -DHAVE_STRSIGNAL=0 \
- -DHAVE_STRLCPY=1 -DHAVE_FLOCK_DECL=1 -DHAVE_REVOKE_DECL=1 \
- -DHAVE_SYS_ERRLIST_DECL=0 -DHAVE_SYS_SIGLIST_DECL=1 \
- -DHAVE_PERSISTENT_HISTORY=0 -DMKSH_BUILD_R=551
+ -DHAVE_MEMMOVE=1 \
+ -DHAVE_MKNOD=0 \
+ -DHAVE_MMAP=1 \
+ -DHAVE_FTRUNCATE=1 \
+ -DHAVE_NICE=1 \
+ -DHAVE_REVOKE=0 \
+ -DHAVE_SETLOCALE_CTYPE=1 \
+ -DHAVE_LANGINFO_CODESET=1 \
+ -DHAVE_SELECT=1 \
+ -DHAVE_SETRESUGID=1 \
+ -DHAVE_SETGROUPS=1 \
+ -DHAVE_STRERROR=1 \
+ -DHAVE_STRSIGNAL=0 \
+ -DHAVE_STRLCPY=1 \
+ -DHAVE_FLOCK_DECL=1 \
+ -DHAVE_REVOKE_DECL=1 \
+ -DHAVE_SYS_ERRLIST_DECL=0 \
+ -DHAVE_SYS_SIGLIST_DECL=1 \
+ -DHAVE_PERSISTENT_HISTORY=0 \
+ -DMKSH_BUILD_R=562 \
LOCAL_SRC_FILES := $(MKSH_SRC_FILES)
diff --git a/Android.patch.txt b/Android.patch.txt
new file mode 100644
index 0000000..c3cf892
--- /dev/null
+++ b/Android.patch.txt
@@ -0,0 +1,48 @@
+--- mksh-R56b/funcs.c 2017-05-05 15:53:55.000000000 -0700
++++ src/funcs.c 2017-09-22 16:19:44.327000462 -0700
+@@ -103,7 +103,9 @@
+ {Tsgbreak, c_brkcont},
+ {T__builtin, c_builtin},
+ {Tbuiltin, c_builtin},
++#if !defined(__ANDROID__)
+ {Tbcat, c_cat},
++#endif
+ {Tcd, c_cd},
+ /* dash compatibility hack */
+ {"chdir", c_cd},
+@@ -126,7 +128,9 @@
+ {"pwd", c_pwd},
+ {Tread, c_read},
+ {Tdsgreadonly, c_typeset},
++#if !defined(__ANDROID__)
+ {"!realpath", c_realpath},
++#endif
+ {"~rename", c_rename},
+ {"*=return", c_exitreturn},
+ {Tsgset, c_set},
+@@ -160,8 +164,10 @@
+ {"~printf", c_printf},
+ #endif
+ #if HAVE_SELECT
++#if !defined(__ANDROID__)
+ {"sleep", c_sleep},
+ #endif
++#endif
+ #ifdef __MirBSD__
+ /* alias to "true" for historical reasons */
+ {"domainname", c_true},
+--- mksh-R56b/main.c 2017-04-28 04:14:14.000000000 -0700
++++ src/main.c 2017-09-22 15:58:14.134149037 -0700
+@@ -410,6 +410,12 @@
+ }
+ }
+
++ /* override default PATH regardless of environment */
++#ifdef MKSH_DEFPATH_OVERRIDE
++ vp = global(TPATH);
++ setstr(vp, MKSH_DEFPATH_OVERRIDE, KSH_RETURN_ERROR);
++#endif
++
+ /* for security */
+ typeset(TinitIFS, 0, 0, 0, 0);
+
diff --git a/src/Build.sh b/src/Build.sh
index ca88a06..78fe347 100644
--- a/src/Build.sh
+++ b/src/Build.sh
@@ -1,5 +1,5 @@
#!/bin/sh
-srcversion='$MirOS: src/bin/mksh/Build.sh,v 1.716 2017/04/12 18:33:22 tg Exp $'
+srcversion='$MirOS: src/bin/mksh/Build.sh,v 1.727 2017/08/29 13:38:28 tg Exp $'
#-
# Copyright (c) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010,
# 2011, 2012, 2013, 2014, 2015, 2016, 2017
@@ -53,6 +53,16 @@
alln=0123456789
alls=______________________________________________________________
+case `echo a | tr '\201' X` in
+X)
+ # EBCDIC build system
+ lfcr='\n\r'
+ ;;
+*)
+ lfcr='\012\015'
+ ;;
+esac
+
genopt_die() {
if test -n "$1"; then
echo >&2 "E: $*"
@@ -425,7 +435,7 @@
na=0
fi
hf=$1; shift
- hv=`echo "$hf" | tr -d '\012\015' | tr -c $alll$allu$alln $alls`
+ hv=`echo "$hf" | tr -d "$lfcr" | tr -c $alll$allu$alln $alls`
echo "/* NeXTstep bug workaround */" >x
for i
do
@@ -496,6 +506,7 @@
tfn=
legacy=0
textmode=0
+ebcdic=false
for i
do
@@ -519,6 +530,9 @@
:-c)
last=c
;;
+ :-E)
+ ebcdic=true
+ ;;
:-G)
echo "$me: Do not call me with '-G'!" >&2
exit 1
@@ -603,6 +617,10 @@
add_cppflags -DMKSH_LEGACY_MODE
fi
+if $ebcdic; then
+ add_cppflags -DMKSH_EBCDIC
+fi
+
if test $textmode = 0; then
check_categories="$check_categories shell:textmode-no shell:binmode-yes"
else
@@ -765,7 +783,9 @@
add_cppflags -DSETUID_CAN_FAIL_WITH_EAGAIN
;;
Haiku)
- add_cppflags -DMKSH_ASSUME_UTF8; HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ add_cppflags -DMKSH_ASSUME_UTF8
+ HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ HAVE_ISOFF_MKSH_ASSUME_UTF8=0
;;
Harvey)
add_cppflags -D_POSIX_SOURCE
@@ -773,11 +793,14 @@
add_cppflags -D_BSD_EXTENSION
add_cppflags -D_SUSV2_SOURCE
add_cppflags -D_GNU_SOURCE
- add_cppflags -DMKSH_ASSUME_UTF8; HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ add_cppflags -DMKSH_ASSUME_UTF8
+ HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ HAVE_ISOFF_MKSH_ASSUME_UTF8=0
add_cppflags -DMKSH_NO_CMDLINE_EDITING
add_cppflags -DMKSH__NO_SETEUGID
oswarn=' and will currently not work'
add_cppflags -DMKSH_UNEMPLOYED
+ add_cppflags -DMKSH_NOPROSPECTOFWORK
# these taken from Harvey-OS github and need re-checking
add_cppflags -D_setjmp=setjmp -D_longjmp=longjmp
: "${HAVE_CAN_NO_EH_FRAME=0}"
@@ -826,7 +849,9 @@
MirBSD)
;;
MSYS_*)
- add_cppflags -DMKSH_ASSUME_UTF8=0; HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ add_cppflags -DMKSH_ASSUME_UTF8=0
+ HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ HAVE_ISOFF_MKSH_ASSUME_UTF8=1
# almost same as CYGWIN* (from RT|Chatzilla)
: "${HAVE_SETLOCALE_CTYPE=0}"
# broken on this OE (from ir0nh34d)
@@ -860,7 +885,9 @@
: "${HAVE_SETLOCALE_CTYPE=0}"
;;
OS/2)
- add_cppflags -DMKSH_ASSUME_UTF8=0; HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ add_cppflags -DMKSH_ASSUME_UTF8=0
+ HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ HAVE_ISOFF_MKSH_ASSUME_UTF8=1
HAVE_TERMIOS_H=0
HAVE_MKNOD=0 # setmode() incompatible
oswarn="; it is being ported"
@@ -894,6 +921,16 @@
] incompatibilities with $y.
"
;;
+OS/390)
+ add_cppflags -DMKSH_ASSUME_UTF8=0
+ HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ HAVE_ISOFF_MKSH_ASSUME_UTF8=1
+ : "${CC=xlc}"
+ : "${SIZE=: size}"
+ add_cppflags -DMKSH_FOR_Z_OS
+ add_cppflags -D_ALL_SOURCE
+ oswarn='; EBCDIC support is incomplete'
+ ;;
OSF1)
HAVE_SIG_T=0 # incompatible
add_cppflags -D_OSF_SOURCE
@@ -907,7 +944,9 @@
add_cppflags -D_LIMITS_EXTENSION
add_cppflags -D_BSD_EXTENSION
add_cppflags -D_SUSV2_SOURCE
- add_cppflags -DMKSH_ASSUME_UTF8; HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ add_cppflags -DMKSH_ASSUME_UTF8
+ HAVE_ISSET_MKSH_ASSUME_UTF8=1
+ HAVE_ISOFF_MKSH_ASSUME_UTF8=0
add_cppflags -DMKSH_NO_CMDLINE_EDITING
add_cppflags -DMKSH__NO_SETEUGID
oswarn=' and will currently not work'
@@ -1047,7 +1086,7 @@
# - LLVM+clang defines __GNUC__ too
# - nwcc defines __GNUC__ too
CPP="$CC -E"
-$e ... which compiler seems to be used
+$e ... which compiler type seems to be used
cat >conftest.c <<'EOF'
const char *
#if defined(__ICC) || defined(__INTEL_COMPILER)
@@ -1297,7 +1336,7 @@
# huh?
;;
esac
-$e "$bi==> which compiler seems to be used...$ao $ui$ct$etd$ao"
+$e "$bi==> which compiler type seems to be used...$ao $ui$ct$etd$ao"
rmf conftest.c conftest.o conftest a.out* a.exe* conftest.exe* vv.out
#
@@ -1392,8 +1431,16 @@
DOWARN=-Wc,-we
;;
xlc)
- save_NOWARN=-qflag=i:e
- DOWARN=-qflag=i:i
+ case $TARGET_OS in
+ OS/390)
+ save_NOWARN=-qflag=e
+ DOWARN=-qflag=i
+ ;;
+ *)
+ save_NOWARN=-qflag=i:e
+ DOWARN=-qflag=i:i
+ ;;
+ esac
;;
*)
test x"$save_NOWARN" = x"" && save_NOWARN=-Wno-error
@@ -1563,10 +1610,24 @@
ac_flags 1 extansi -Xa
;;
xlc)
- ac_flags 1 rodata "-qro -qroconst -qroptr"
- ac_flags 1 rtcheck -qcheck=all
- #ac_flags 1 rtchkc -qextchk # reported broken
- ac_flags 1 wformat "-qformat=all -qformat=nozln"
+ case $TARGET_OS in
+ OS/390)
+ # On IBM z/OS, the following are warnings by default:
+ # CCN3296: #include file <foo.h> not found.
+ # CCN3944: Attribute "__foo__" is not supported and is ignored.
+ # CCN3963: The attribute "foo" is not a valid variable attribute and is ignored.
+ ac_flags 1 halton '-qhaltonmsg=CCN3296 -qhaltonmsg=CCN3944 -qhaltonmsg=CCN3963'
+ # CCN3290: Unknown macro name FOO on #undef directive.
+ # CCN4108: The use of keyword '__attribute__' is non-portable.
+ ac_flags 1 supprss '-qsuppress=CCN3290 -qsuppress=CCN4108'
+ ;;
+ *)
+ ac_flags 1 rodata '-qro -qroconst -qroptr'
+ ac_flags 1 rtcheck -qcheck=all
+ #ac_flags 1 rtchkc -qextchk # reported broken
+ ac_flags 1 wformat '-qformat=all -qformat=nozln'
+ ;;
+ esac
#ac_flags 1 wp64 -qwarn64 # too verbose for now
;;
esac
@@ -1705,6 +1766,10 @@
check_categories="$check_categories arge nojsig"
ac_ifcpp 'ifdef MKSH_ASSUME_UTF8' isset_MKSH_ASSUME_UTF8 '' \
'if the default UTF-8 mode is specified' && : "${HAVE_SETLOCALE_CTYPE=0}"
+ac_ifcpp 'if !MKSH_ASSUME_UTF8' isoff_MKSH_ASSUME_UTF8 \
+ isset_MKSH_ASSUME_UTF8 0 \
+ 'if the default UTF-8 mode is disabled' && \
+ check_categories="$check_categories noutf8"
#ac_ifcpp 'ifdef MKSH_DISABLE_DEPRECATED' isset_MKSH_DISABLE_DEPRECATED '' \
# "if deprecated features are to be omitted" && \
# check_categories="$check_categories nodeprecated"
@@ -2025,6 +2090,11 @@
munmap(NULL, 0)); }
EOF
+ac_test ftruncate mmap 0 'for ftruncate' <<-'EOF'
+ #include <unistd.h>
+ int main(void) { return (ftruncate(0, 0)); }
+EOF
+
ac_test nice <<-'EOF'
#include <unistd.h>
int main(void) { return (nice(4)); }
@@ -2179,8 +2249,8 @@
# other checks
#
fd='if to use persistent history'
-ac_cache PERSISTENT_HISTORY || case $HAVE_MMAP$HAVE_FLOCK$HAVE_LOCK_FCNTL in
-11*|101) fv=1 ;;
+ac_cache PERSISTENT_HISTORY || case $HAVE_FTRUNCATE$HAVE_MMAP$HAVE_FLOCK$HAVE_LOCK_FCNTL in
+111*|1101) fv=1 ;;
esac
test 1 = $fv || check_categories="$check_categories no-histfile"
ac_testdone
@@ -2339,7 +2409,7 @@
addsrcs USE_PRINTF_BUILTIN printf.c
test 1 = "$USE_PRINTF_BUILTIN" && add_cppflags -DMKSH_PRINTF_BUILTIN
test 1 = "$HAVE_CAN_VERB" && CFLAGS="$CFLAGS -verbose"
-add_cppflags -DMKSH_BUILD_R=551
+add_cppflags -DMKSH_BUILD_R=562
$e $bi$me: Finished configuration testing, now producing output.$ao
@@ -2366,8 +2436,8 @@
set -A check_categories -- $check_categories
pflag='$curdir/$mkshexe'
sflag='$srcdir/check.t'
- usee=0 Pflag=0 Sflag=0 uset=0 vflag=1 xflag=0
- while getopts "C:e:fPp:QSs:t:v" ch; do case \$ch {
+ usee=0 useU=0 Pflag=0 Sflag=0 uset=0 vflag=1 xflag=0
+ while getopts "C:e:fPp:QSs:t:U:v" ch; do case \$ch {
(C) check_categories[\${#check_categories[*]}]=\$OPTARG ;;
(e) usee=1; eflag=\$OPTARG ;;
(f) check_categories[\${#check_categories[*]}]=fastbox ;;
@@ -2380,6 +2450,7 @@
(+S) Sflag=0 ;;
(s) sflag=\$OPTARG ;;
(t) uset=1; tflag=\$OPTARG ;;
+ (U) useU=1; Uflag=\$OPTARG ;;
(v) vflag=1 ;;
(+v) vflag=0 ;;
(*) xflag=1 ;;
@@ -2387,6 +2458,9 @@
done
shift \$((OPTIND - 1))
set -A args -- '$srcdir/check.pl' -p "\$pflag"
+ if $ebcdic; then
+ args[\${#args[*]}]=-E
+ fi
x=
for y in "\${check_categories[@]}"; do
x=\$x,\$y
@@ -2404,6 +2478,10 @@
args[\${#args[*]}]=-t
args[\${#args[*]}]=\$tflag
fi
+ if (( useU )); then
+ args[\${#args[*]}]=-U
+ args[\${#args[*]}]=\$Uflag
+ fi
(( vflag )) && args[\${#args[*]}]=-v
(( xflag )) && args[\${#args[*]}]=-x # force usage by synerr
if [[ -n \$TMPDIR && -d \$TMPDIR/. ]]; then
@@ -2647,7 +2725,7 @@
MKSH_ASSUME_UTF8 (0=disabled, 1=enabled; default: unset)
MKSH_BINSHPOSIX if */sh or */-sh, enable set -o posix
MKSH_BINSHREDUCED if */sh or */-sh, enable set -o sh
-MKSH_CLS_STRING "\033[;H\033[J"
+MKSH_CLS_STRING KSH_ESC_STRING "[;H" KSH_ESC_STRING "[J"
MKSH_DEFAULT_EXECSHELL "/bin/sh" (do not change)
MKSH_DEFAULT_PROFILEDIR "/etc" (do not change)
MKSH_DEFAULT_TMPDIR "/tmp" (do not change)
diff --git a/src/check.pl b/src/check.pl
index a80d4e1..e9c2437 100644
--- a/src/check.pl
+++ b/src/check.pl
@@ -1,8 +1,8 @@
-# $MirOS: src/bin/mksh/check.pl,v 1.42 2015/11/29 17:05:00 tg Exp $
+# $MirOS: src/bin/mksh/check.pl,v 1.49 2017/05/05 21:17:31 tg Exp $
# $OpenBSD: th,v 1.1 2013/12/02 20:39:44 millert Exp $
#-
# Copyright (c) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2011,
-# 2012, 2013, 2014, 2015
+# 2012, 2013, 2014, 2015, 2017
# mirabilos <m@mirbsd.org>
#
# Provided that these terms and disclaimer and all copyright notices
@@ -78,9 +78,9 @@
# the test harness).
# CYGWIN is set to nodosfilewarning.
# ENV is set to /nonexistant.
-# PATHSEP is set to either : or ;.
# __progname is set to the -p argument.
# __perlname is set to $^X (perlexe).
+# @utflocale@ is substituted from -U.
# file-setup mps Used to create files, directories
# and symlinks. First word is either
# file, dir or symlink; second word is
@@ -153,9 +153,15 @@
# p tag takes parameters (used with m).
# s tag can be used several times.
+# require Config only if it exists
# pull EINTR from POSIX.pm or Errno.pm if they exist
# otherwise just skip it
BEGIN {
+ eval {
+ require Config;
+ import Config;
+ 1;
+ };
$EINTR = 0;
eval {
require POSIX;
@@ -172,7 +178,6 @@
};
use Getopt::Std;
-use Config;
$os = defined $^O ? $^O : 'unknown';
@@ -180,7 +185,7 @@
$Usage = <<EOF ;
Usage: $prog [-Pv] [-C cat] [-e e=v] [-p prog] [-s fn] [-T dir] \
- [-t tmo] name ...
+ [-t tmo] [-U lcl] name ...
-C c Specify the comma separated list of categories the program
belongs to (see category field).
-e e=v Set the environment variable e to v for all tests
@@ -193,6 +198,7 @@
scaned for test files (which end in .t).
-T dir Use dir instead of /tmp to hold temporary files
-t t Use t as default time limit for tests (default is unlimited)
+ -U lcl Use lcl as UTF-8 locale (e.g. C.UTF-8) instead of the default
-v Verbose mode: print reason test failed.
name specifies the name of the test(s) to run; if none are
specified, all tests are run.
@@ -241,7 +247,7 @@
%known_tests = ();
-if (!getopts('C:e:Pp:s:T:t:v')) {
+if (!getopts('C:Ee:Pp:s:T:t:U:v')) {
print STDERR $Usage;
exit 1;
}
@@ -250,8 +256,10 @@
die "$prog: no test set specified (use -s)\n" if !defined $opt_s;
$test_prog = $opt_p;
$verbose = defined $opt_v && $opt_v;
+$is_ebcdic = defined $opt_E && $opt_E;
$test_set = $opt_s;
$temp_base = $opt_T || "/tmp";
+$utflocale = $opt_U || (($os eq "hpux") ? "en_US.utf8" : "en_US.UTF-8");
if (defined $opt_t) {
die "$prog: bad -t argument (should be number > 0): $opt_t\n"
if $opt_t !~ /^\d+$/ || $opt_t <= 0;
@@ -259,6 +267,14 @@
}
$program_kludge = defined $opt_P ? $opt_P : 0;
+if ($is_ebcdic) {
+ $categories{'shell:ebcdic-yes'} = 1;
+ $categories{'shell:ascii-no'} = 1;
+} else {
+ $categories{'shell:ebcdic-no'} = 1;
+ $categories{'shell:ascii-yes'} = 1;
+}
+
if (defined $opt_C) {
foreach $c (split(',', $opt_C)) {
$c =~ s/\s+//;
@@ -281,12 +297,24 @@
}
$new_env{'CYGWIN'} = 'nodosfilewarning';
$new_env{'ENV'} = '/nonexistant';
-$new_env{'PATHSEP'} = $os eq 'os2' ? ';' : ':';
+
if (($os eq 'VMS') || ($Config{perlpath} =~ m/$Config{_exe}$/i)) {
$new_env{'__perlname'} = $Config{perlpath};
} else {
$new_env{'__perlname'} = $Config{perlpath} . $Config{_exe};
}
+$new_env{'__perlname'} = $^X if ($new_env{'__perlname'} eq '') and -f $^X and -x $^X;
+if ($new_env{'__perlname'} eq '') {
+ foreach $pathelt (split /:/,$ENV{'PATH'}) {
+ chomp($pathelt = `pwd`) if $pathelt eq '';
+ my $x = $pathelt . '/' . $^X;
+ next unless -f $x and -x $x;
+ $new_env{'__perlname'} = $x;
+ last;
+ }
+}
+$new_env{'__perlname'} = $^X if ($new_env{'__perlname'} eq '');
+
if (defined $opt_e) {
# XXX need a way to allow many -e arguments...
if ($opt_e =~ /^([a-zA-Z_]\w*)(|=(.*))$/) {
@@ -866,38 +894,50 @@
$char = 1;
}
}
- return "first difference: line $lineno, char $char (wanted '"
- . &format_char($ce) . "', got '"
- . &format_char($cg) . "'";
+ return "first difference: line $lineno, char $char (wanted " .
+ &format_char($ce) . ", got " . &format_char($cg);
}
sub
format_char
{
- local($ch, $s);
+ local($ch, $s, $q);
$ch = ord($_[0]);
+ $q = "'";
+
+ if ($is_ebcdic) {
+ if ($ch == 0x15) {
+ return $q . '\n' . $q;
+ } elsif ($ch == 0x16) {
+ return $q . '\b' . $q;
+ } elsif ($ch == 0x05) {
+ return $q . '\t' . $q;
+ } elsif ($ch < 64 || $ch == 255) {
+ return sprintf("X'%02X'", $ch);
+ }
+ return sprintf("'%c' (X'%02X')", $ch, $ch);
+ }
+
+ $s = sprintf("0x%02X (", $ch);
if ($ch == 10) {
- return '\n';
+ return $s . $q . '\n' . $q . ')';
} elsif ($ch == 13) {
- return '\r';
+ return $s . $q . '\r' . $q . ')';
} elsif ($ch == 8) {
- return '\b';
+ return $s . $q . '\b' . $q . ')';
} elsif ($ch == 9) {
- return '\t';
+ return $s . $q . '\t' . $q . ')';
} elsif ($ch > 127) {
- $ch -= 127;
- $s = "M-";
- } else {
- $s = '';
+ $ch -= 128;
+ $s .= "M-";
}
if ($ch < 32) {
- $s .= '^';
- $ch += ord('@');
+ return sprintf("%s^%c)", $s, $ch + ord('@'));
} elsif ($ch == 127) {
- return $s . "^?";
+ return $s . "^?)";
}
- return $s . sprintf("%c", $ch);
+ return sprintf("%s'%c')", $s, $ch);
}
sub
@@ -1159,6 +1199,8 @@
print STDERR "$prog:$test{':long-name'}: env-setup field doesn't start and end with the same character\n";
return undef;
}
+
+ $test{'env-setup'} =~ s/\@utflocale\@/$utflocale/g;
}
if (defined $test{'expected-exit'}) {
local($val) = $test{'expected-exit'};
diff --git a/src/check.t b/src/check.t
index 93c614f..4473c08 100644
--- a/src/check.t
+++ b/src/check.t
Binary files differ
diff --git a/src/dot.mkshrc b/src/dot.mkshrc
index af55d7d..4a3dfea 100644
--- a/src/dot.mkshrc
+++ b/src/dot.mkshrc
@@ -1,5 +1,5 @@
# $Id$
-# $MirOS: src/bin/mksh/dot.mkshrc,v 1.114 2017/03/19 22:31:26 tg Exp $
+# $MirOS: src/bin/mksh/dot.mkshrc,v 1.121 2017/08/08 21:10:21 tg Exp $
#-
# Copyright (c) 2002, 2003, 2004, 2006, 2007, 2008, 2009, 2010,
# 2011, 2012, 2013, 2014, 2015, 2016, 2017
@@ -56,6 +56,13 @@
done
)
+# customise your favourite editor here; the first one found is used
+for EDITOR in "${EDITOR:-}" jupp jstar mcedit ed vi; do
+ EDITOR=$(\\builtin whence -p "$EDITOR") || EDITOR=
+ [[ -n $EDITOR && -x $EDITOR ]] && break
+ EDITOR=
+done
+
\\builtin alias ls=ls l='ls -F' la='l -a' ll='l -l' lo='l -alo'
\: "${HOSTNAME:=$(\\builtin ulimit -c 0; \\builtin print -r -- $(hostname \
2>/dev/null))}${EDITOR:=/bin/ed}${TERM:=vt100}${USER:=$(\\builtin ulimit \
@@ -73,6 +80,7 @@
\: "${MKSH:=$(\\builtin whence -p mksh)}${MKSH:=/bin/mksh}"
\\builtin export MKSH
+# prompts
PS4='[$EPOCHREALTIME] '; PS1='#'; (( USER_ID )) && PS1='$'; PS1=$'\001\r''${|
\\builtin typeset e=$?
@@ -86,6 +94,8 @@
\\builtin return $e
} '"$PS1 "
+
+# utilities
\\builtin alias doch='sudo mksh -c "$(\\builtin fc -ln -1)"'
\\builtin command -v rot13 >/dev/null || \\builtin alias rot13='tr \
abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ \
@@ -99,42 +109,47 @@
}
else
function hd {
- \\builtin typeset -Uui16 -Z11 pos=0
- \\builtin typeset -Uui16 -Z5 hv=2147483647
- \\builtin typeset dasc line i
- \\builtin set +U
-
- \\builtin cat "$@" | if \\builtin read -arN -1 line; then
- \\builtin typeset -i1 'line[*]'
- i=0
- while (( i < ${#line[*]} )); do
- hv=${line[i++]}
- if (( (pos & 15) == 0 )); then
- (( pos )) && \
- \\builtin print -r -- "$dasc|"
- \\builtin print -nr "${pos#16#} "
- dasc=' |'
- fi
- \\builtin print -nr "${hv#16#} "
- #XXX EBCDIC, but we need [[:print:]] to fix this
- if (( (hv < 32) || (hv > 126) )); then
- dasc+=.
- else
- dasc+=${line[i-1]#1#}
- fi
- (( (pos++ & 15) == 7 )) && \
- \\builtin print -nr -- '- '
- done
- while (( pos & 15 )); do
- \\builtin print -nr ' '
- (( (pos++ & 15) == 7 )) && \
- \\builtin print -nr -- '- '
- done
- (( hv == 2147483647 )) || \\builtin print -r -- "$dasc|"
- fi
+ \\builtin cat "$@" | hd_mksh "$@"
}
fi
+# NUL-safe and EBCDIC-safe hexdump (from stdin)
+function hd_mksh {
+ \\builtin typeset -Uui16 -Z11 pos=0
+ \\builtin typeset -Uui16 -Z5 hv=2147483647
+ \\builtin typeset dasc dn line i
+ \\builtin set +U
+
+ while \\builtin read -arn 512 line; do
+ \\builtin typeset -i1 'line[*]'
+ i=0
+ while (( i < ${#line[*]} )); do
+ dn=
+ (( (hv = line[i++]) != 0 )) && dn=${line[i-1]#1#}
+ if (( (pos & 15) == 0 )); then
+ (( pos )) && \
+ \\builtin print -r -- "$dasc|"
+ \\builtin print -nr "${pos#16#} "
+ dasc=' |'
+ fi
+ \\builtin print -nr "${hv#16#} "
+ if [[ $dn = [[:print:]] ]]; then
+ dasc+=$dn
+ else
+ dasc+=.
+ fi
+ (( (pos++ & 15) == 7 )) && \
+ \\builtin print -nr -- '- '
+ done
+ done
+ while (( pos & 15 )); do
+ \\builtin print -nr ' '
+ (( (pos++ & 15) == 7 )) && \
+ \\builtin print -nr -- '- '
+ done
+ (( hv == 2147483647 )) || \\builtin print -r -- "$dasc|"
+}
+
# Berkeley C shell compatible dirs, popd, and pushd functions
# Z shell compatible chpwd() hook, used to update DIRSTACK[0]
DIRSTACKBASE=$(\\builtin realpath ~/. 2>/dev/null || \
@@ -483,6 +498,7 @@
i_func[nfunc++]=setenv
i_func[nfunc++]=smores
i_func[nfunc++]=hd
+ i_func[nfunc++]=hd_mksh
i_func[nfunc++]=chpwd
i_func[nfunc++]=cd
i_func[nfunc++]=cd_csh
@@ -588,6 +604,11 @@
\: place customisations below this line
+# some defaults follow — you are supposed to adjust these to your
+# liking; by default we add ~/.etc/bin and ~/bin (whichever exist)
+# to $PATH, set $SHELL to mksh, set some defaults for man and less
+# and show a few more possible things for users to begin moving in
+
for p in ~/.etc/bin ~/bin; do
[[ -d $p/. ]] || \\builtin continue
[[ $PATHSEP$PATH$PATHSEP = *"$PATHSEP$p$PATHSEP"* ]] || \
diff --git a/src/edit.c b/src/edit.c
index 58eaf7f..8bccf13 100644
--- a/src/edit.c
+++ b/src/edit.c
@@ -28,16 +28,16 @@
#ifndef MKSH_NO_CMDLINE_EDITING
-__RCSID("$MirOS: src/bin/mksh/edit.c,v 1.321 2017/04/12 16:46:20 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/edit.c,v 1.340 2017/08/27 23:33:50 tg Exp $");
/*
* in later versions we might use libtermcap for this, but since external
* dependencies are problematic, this has not yet been decided on; another
- * good string is "\033c" except on hardware terminals like the DEC VT420
- * which do a full power cycle then...
+ * good string is KSH_ESC_STRING "c" except on hardware terminals like the
+ * DEC VT420 which do a full power cycle then...
*/
#ifndef MKSH_CLS_STRING
-#define MKSH_CLS_STRING "\033[;H\033[J"
+#define MKSH_CLS_STRING KSH_ESC_STRING "[;H" KSH_ESC_STRING "[J"
#endif
/* tty driver characters we are interested in */
@@ -76,7 +76,7 @@
static char *holdbufp; /* place to hold last edit buffer */
/* 0=dumb 1=tmux (for now) */
-static bool x_term_mode;
+static uint8_t x_term_mode;
static void x_adjust(void);
static int x_getc(void);
@@ -97,6 +97,7 @@
#if !MKSH_S_NOVI
static int x_vi(char *);
#endif
+static void x_intr(int, int) MKSH_A_NORETURN;
#define x_flush() shf_flush(shl_out)
#if defined(MKSH_SMALL) && !defined(MKSH_SMALL_BUT_FAST)
@@ -309,14 +310,14 @@
* empirically made list of chars to escape
* for globbing as well as QCHAR itself
*/
- switch (ch) {
+ switch (ord(ch)) {
case QCHAR:
- case '$':
- case '*':
- case '?':
- case '[':
- case '\\':
- case '`':
+ case ord('$'):
+ case ord('*'):
+ case ord('?'):
+ case ord('['):
+ case ord('\\'):
+ case ord('`'):
*dp++ = QCHAR;
break;
}
@@ -467,7 +468,7 @@
const struct path_order_info *b = (const struct path_order_info *)bb;
int t;
- if ((t = strcmp(a->word + a->base, b->word + b->base)))
+ if ((t = ascstrcmp(a->word + a->base, b->word + b->base)))
return (t);
if (a->path_order > b->path_order)
return (1);
@@ -535,7 +536,7 @@
char **words = (char **)XPptrv(w);
size_t i, j;
- qsort(words, nwords, sizeof(void *), xstrcmp);
+ qsort(words, nwords, sizeof(void *), ascpstrcmp);
for (i = j = 0; i < nwords - 1; i++) {
if (strcmp(words[i], words[i + 1]))
words[j++] = words[i];
@@ -552,8 +553,7 @@
return (nwords);
}
-#define IS_WORDC(c) (!ctype(c, C_LEX1) && (c) != '\'' && (c) != '"' && \
- (c) != '`' && (c) != '=' && (c) != ':')
+#define IS_WORDC(c) (!ctype(c, C_EDNWC))
static int
x_locate_word(const char *buf, int buflen, int pos, int *startp,
@@ -588,9 +588,9 @@
int p = start - 1;
/* Figure out if this is a command */
- while (p >= 0 && ksh_isspace(buf[p]))
+ while (p >= 0 && ctype(buf[p], C_SPACE))
p--;
- iscmd = p < 0 || vstrchr(";|&()`", buf[p]);
+ iscmd = p < 0 || ctype(buf[p], C_EDCMD);
if (iscmd) {
/*
* If command has a /, path, etc. is not searched;
@@ -649,11 +649,12 @@
for (s = toglob; *s; s++) {
if (*s == '\\' && s[1])
s++;
- else if (*s == '?' || *s == '*' || *s == '[' ||
- *s == '$' ||
+ else if (ctype(*s, C_QUEST | C_DOLAR) ||
+ ord(*s) == ord('*') || ord(*s) == ord('[') ||
/* ?() *() +() @() !() but two already checked */
- (s[1] == '(' /*)*/ &&
- (*s == '+' || *s == '@' || *s == '!'))) {
+ (ord(s[1]) == ord('(' /*)*/) &&
+ (ord(*s) == ord('+') || ord(*s) == ord('@') ||
+ ord(*s) == ord('!')))) {
/*
* just expand based on the extglob
* or parameter
@@ -714,8 +715,8 @@
break;
}
/* false for nwords==1 as 0 = words[0][prefix_len] then */
- if (UTFMODE && prefix_len && (words[0][prefix_len] & 0xC0) == 0x80)
- while (prefix_len && (words[0][prefix_len] & 0xC0) != 0xC0)
+ if (UTFMODE && prefix_len && (rtt2asc(words[0][prefix_len]) & 0xC0) == 0x80)
+ while (prefix_len && (rtt2asc(words[0][prefix_len]) & 0xC0) != 0xC0)
--prefix_len;
return (prefix_len);
}
@@ -747,7 +748,7 @@
const char *p;
if (se == NULL)
- se = s + strlen(s);
+ se = strnul(s);
if (s == se)
return (0);
@@ -799,7 +800,7 @@
while (sp) {
xp = Xstring(xs, xp);
if (!(p = cstrchr(sp, MKSH_PATHSEPC)))
- p = sp + strlen(sp);
+ p = strnul(sp);
pathlen = p - sp;
if (pathlen) {
/*
@@ -858,8 +859,7 @@
int rval = 0;
while (wlen - add > 0)
- if (vstrchr("\"#$&'()*:;<=>?[\\`{|}", s[add]) ||
- ctype(s[add], C_IFS)) {
+ if (ctype(s[add], C_IFS | C_EDQ)) {
if (putbuf_func(s, add) != 0) {
rval = -1;
break;
@@ -908,11 +908,6 @@
#define XF_NOBIND 2 /* not allowed to bind to function */
#define XF_PREFIX 4 /* function sets prefix */
-/* Separator for completion */
-#define is_cfs(c) ((c) == ' ' || (c) == '\t' || (c) == '"' || (c) == '\'')
-/* Separator for motion */
-#define is_mfs(c) (!(ksh_isalnux(c) || (c) == '$' || ((c) & 0x80)))
-
#define X_NTABS 4 /* normal, meta1, meta2, pc */
#define X_TABSZ 256 /* size of keydef tables etc */
@@ -991,6 +986,7 @@
static int x_size2(char *, char **);
static void x_zots(char *);
static void x_zotc3(char **);
+static void x_vi_zotc(int);
static void x_load_hist(char **);
static int x_search(char *, int, int);
#ifndef MKSH_SMALL
@@ -1036,56 +1032,56 @@
};
static struct x_defbindings const x_defbindings[] = {
- { XFUNC_del_back, 0, CTRL('?') },
- { XFUNC_del_bword, 1, CTRL('?') },
- { XFUNC_eot_del, 0, CTRL('D') },
- { XFUNC_del_back, 0, CTRL('H') },
- { XFUNC_del_bword, 1, CTRL('H') },
+ { XFUNC_del_back, 0, CTRL_QM },
+ { XFUNC_del_bword, 1, CTRL_QM },
+ { XFUNC_eot_del, 0, CTRL_D },
+ { XFUNC_del_back, 0, CTRL_H },
+ { XFUNC_del_bword, 1, CTRL_H },
{ XFUNC_del_bword, 1, 'h' },
{ XFUNC_mv_bword, 1, 'b' },
{ XFUNC_mv_fword, 1, 'f' },
{ XFUNC_del_fword, 1, 'd' },
- { XFUNC_mv_back, 0, CTRL('B') },
- { XFUNC_mv_forw, 0, CTRL('F') },
- { XFUNC_search_char_forw, 0, CTRL(']') },
- { XFUNC_search_char_back, 1, CTRL(']') },
- { XFUNC_newline, 0, CTRL('M') },
- { XFUNC_newline, 0, CTRL('J') },
- { XFUNC_end_of_text, 0, CTRL('_') },
- { XFUNC_abort, 0, CTRL('G') },
- { XFUNC_prev_com, 0, CTRL('P') },
- { XFUNC_next_com, 0, CTRL('N') },
- { XFUNC_nl_next_com, 0, CTRL('O') },
- { XFUNC_search_hist, 0, CTRL('R') },
+ { XFUNC_mv_back, 0, CTRL_B },
+ { XFUNC_mv_forw, 0, CTRL_F },
+ { XFUNC_search_char_forw, 0, CTRL_BC },
+ { XFUNC_search_char_back, 1, CTRL_BC },
+ { XFUNC_newline, 0, CTRL_M },
+ { XFUNC_newline, 0, CTRL_J },
+ { XFUNC_end_of_text, 0, CTRL_US },
+ { XFUNC_abort, 0, CTRL_G },
+ { XFUNC_prev_com, 0, CTRL_P },
+ { XFUNC_next_com, 0, CTRL_N },
+ { XFUNC_nl_next_com, 0, CTRL_O },
+ { XFUNC_search_hist, 0, CTRL_R },
{ XFUNC_beg_hist, 1, '<' },
{ XFUNC_end_hist, 1, '>' },
{ XFUNC_goto_hist, 1, 'g' },
- { XFUNC_mv_end, 0, CTRL('E') },
- { XFUNC_mv_beg, 0, CTRL('A') },
- { XFUNC_draw_line, 0, CTRL('L') },
- { XFUNC_cls, 1, CTRL('L') },
- { XFUNC_meta1, 0, CTRL('[') },
- { XFUNC_meta2, 0, CTRL('X') },
- { XFUNC_kill, 0, CTRL('K') },
- { XFUNC_yank, 0, CTRL('Y') },
+ { XFUNC_mv_end, 0, CTRL_E },
+ { XFUNC_mv_beg, 0, CTRL_A },
+ { XFUNC_draw_line, 0, CTRL_L },
+ { XFUNC_cls, 1, CTRL_L },
+ { XFUNC_meta1, 0, CTRL_BO },
+ { XFUNC_meta2, 0, CTRL_X },
+ { XFUNC_kill, 0, CTRL_K },
+ { XFUNC_yank, 0, CTRL_Y },
{ XFUNC_meta_yank, 1, 'y' },
- { XFUNC_literal, 0, CTRL('^') },
+ { XFUNC_literal, 0, CTRL_CA },
{ XFUNC_comment, 1, '#' },
- { XFUNC_transpose, 0, CTRL('T') },
- { XFUNC_complete, 1, CTRL('[') },
- { XFUNC_comp_list, 0, CTRL('I') },
+ { XFUNC_transpose, 0, CTRL_T },
+ { XFUNC_complete, 1, CTRL_BO },
+ { XFUNC_comp_list, 0, CTRL_I },
{ XFUNC_comp_list, 1, '=' },
{ XFUNC_enumerate, 1, '?' },
{ XFUNC_expand, 1, '*' },
- { XFUNC_comp_file, 1, CTRL('X') },
- { XFUNC_comp_comm, 2, CTRL('[') },
+ { XFUNC_comp_file, 1, CTRL_X },
+ { XFUNC_comp_comm, 2, CTRL_BO },
{ XFUNC_list_comm, 2, '?' },
- { XFUNC_list_file, 2, CTRL('Y') },
+ { XFUNC_list_file, 2, CTRL_Y },
{ XFUNC_set_mark, 1, ' ' },
- { XFUNC_kill_region, 0, CTRL('W') },
- { XFUNC_xchg_point_mark, 2, CTRL('X') },
- { XFUNC_literal, 0, CTRL('V') },
- { XFUNC_version, 1, CTRL('V') },
+ { XFUNC_kill_region, 0, CTRL_W },
+ { XFUNC_xchg_point_mark, 2, CTRL_X },
+ { XFUNC_literal, 0, CTRL_V },
+ { XFUNC_version, 1, CTRL_V },
{ XFUNC_prev_histword, 1, '.' },
{ XFUNC_prev_histword, 1, '_' },
{ XFUNC_set_arg, 1, '0' },
@@ -1148,7 +1144,7 @@
#endif
#ifndef MKSH_SMALL
/* more non-standard ones */
- { XFUNC_eval_region, 1, CTRL('E') },
+ { XFUNC_eval_region, 1, CTRL_E },
{ XFUNC_edit_line, 2, 'e' }
#endif
};
@@ -1191,17 +1187,19 @@
if (c == -1)
return (-1);
if (UTFMODE) {
- if ((buf[0] >= 0xC2) && (buf[0] < 0xF0)) {
+ if ((rtt2asc(buf[0]) >= (unsigned char)0xC2) &&
+ (rtt2asc(buf[0]) < (unsigned char)0xF0)) {
c = x_e_getc();
if (c == -1)
return (-1);
- if ((c & 0xC0) != 0x80) {
+ if ((rtt2asc(c) & 0xC0) != 0x80) {
x_e_ungetc(c);
return (1);
}
buf[pos++] = c;
}
- if ((buf[0] >= 0xE0) && (buf[0] < 0xF0)) {
+ if ((rtt2asc(buf[0]) >= (unsigned char)0xE0) &&
+ (rtt2asc(buf[0]) < (unsigned char)0xF0)) {
/* XXX x_e_ungetc is one-octet only */
buf[pos++] = c = x_e_getc();
if (c == -1)
@@ -1299,9 +1297,7 @@
return (i);
case KINTR:
/* special case for interrupt */
- trapsig(SIGINT);
- x_mode(false);
- unwind(LSHELL);
+ x_intr(SIGINT, c);
}
/* ad-hoc hack for fixing the cursor position */
x_goto(xcp);
@@ -1320,11 +1316,11 @@
if (c == 0) {
invmbs:
left = 0;
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
if (UTFMODE) {
- if (((c & 0xC0) == 0x80) && left) {
+ if (((rtt2asc(c) & 0xC0) == 0x80) && left) {
str[pos++] = c;
if (!--left) {
str[pos] = '\0';
@@ -1382,7 +1378,7 @@
x_do_ins(const char *cp, size_t len)
{
if (xep + len >= xend) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (-1);
}
memmove(xcp + len, xcp, xep - xcp + 1);
@@ -1422,7 +1418,7 @@
ssize_t i = 0;
if (xcp == xbuf) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
do {
@@ -1448,7 +1444,7 @@
}
if (!i) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
x_delete(i, false);
@@ -1558,15 +1554,15 @@
char *cp = xcp;
if (cp == xbuf) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (0);
}
while (x_arg--) {
- while (cp != xbuf && is_mfs(cp[-1])) {
+ while (cp != xbuf && ctype(cp[-1], C_MFS)) {
cp--;
nb++;
}
- while (cp != xbuf && !is_mfs(cp[-1])) {
+ while (cp != xbuf && !ctype(cp[-1], C_MFS)) {
cp--;
nb++;
}
@@ -1582,13 +1578,13 @@
char *cp = xcp;
if (cp == xep) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (0);
}
while (x_arg--) {
- while (cp != xep && is_mfs(*cp))
+ while (cp != xep && ctype(*cp, C_MFS))
cp++;
- while (cp != xep && !is_mfs(*cp))
+ while (cp != xep && !ctype(*cp, C_MFS))
cp++;
}
nc = x_nb2nc(cp - xcp);
@@ -1621,7 +1617,7 @@
{
if (UTFMODE)
while ((!lower_bound || (cp > lower_bound)) &&
- ((*(unsigned char *)cp & 0xC0) == 0x80))
+ ((rtt2asc(*cp) & 0xC0) == 0x80))
--cp;
return (cp);
}
@@ -1642,14 +1638,14 @@
{
uint8_t c = *(unsigned char *)cp;
- if (UTFMODE && (c > 0x7F))
+ if (UTFMODE && (rtt2asc(c) > 0x7F))
return (utf_widthadj(cp, (const char **)dcp));
if (dcp)
*dcp = cp + 1;
if (c == '\t')
/* Kludge, tabs are always four spaces. */
return (4);
- if (ISCTRL(c) && /* but not C1 */ c < 0x80)
+ if (ksh_isctrl(c))
/* control unsigned char */
return (2);
return (1);
@@ -1674,9 +1670,9 @@
/* Kludge, tabs are always four spaces. */
x_e_puts(T4spaces);
(*cp)++;
- } else if (ISCTRL(c) && /* but not C1 */ c < 0x80) {
+ } else if (ksh_isctrl(c)) {
x_e_putc2('^');
- x_e_putc2(UNCTRL(c));
+ x_e_putc2(ksh_unctrl(c));
(*cp)++;
} else
x_e_putc3((const char **)cp);
@@ -1686,7 +1682,7 @@
x_mv_back(int c MKSH_A_UNUSED)
{
if (xcp == xbuf) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
while (x_arg--) {
@@ -1703,7 +1699,7 @@
char *cp = xcp, *cp2;
if (xcp == xep) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
while (x_arg--) {
@@ -1724,13 +1720,13 @@
*xep = '\0';
if (x_e_getmbc(tmp) < 0) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
while (x_arg--) {
if ((cp = (cp == xep) ? NULL : strstr(cp + 1, tmp)) == NULL &&
(cp = strstr(xbuf, tmp)) == NULL) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
}
@@ -1745,7 +1741,7 @@
bool b;
if (x_e_getmbc(tmp) < 0) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
for (; x_arg--; cp = p)
@@ -1753,7 +1749,7 @@
if (p-- == xbuf)
p = xep;
if (p == cp) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
if ((tmp[1] && ((p+1) > xep)) ||
@@ -1789,7 +1785,7 @@
unsigned char tmp[1], *cp = tmp;
*tmp = isedchar(edchars.eof) ? (unsigned char)edchars.eof :
- (unsigned char)CTRL('D');
+ (unsigned char)CTRL_D;
x_zotc3((char **)&cp);
x_putc('\r');
x_putc('\n');
@@ -1849,7 +1845,7 @@
sp = holdbufp;
modified = 0;
} else if (hp < history || hp > histptr) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return;
}
if (sp == NULL)
@@ -1859,7 +1855,7 @@
strlcpy(holdbufp, xbuf, LINE);
strlcpy(xbuf, sp, xend - xbuf);
xbp = xbuf;
- xep = xcp = xbuf + strlen(xbuf);
+ xep = xcp = strnul(xbuf);
x_adjust();
modified = 0;
}
@@ -1904,13 +1900,13 @@
if ((c = x_e_getc()) < 0)
return (KSTD);
f = x_tab[0][c];
- if (c == CTRL('[')) {
+ if (c == CTRL_BO) {
if ((f & 0x7F) == XFUNC_meta1) {
if ((c = x_e_getc()) < 0)
return (KSTD);
f = x_tab[1][c] & 0x7F;
if (f == XFUNC_meta1 || f == XFUNC_meta2)
- x_meta1(CTRL('['));
+ x_meta1(CTRL_BO);
x_e_ungetc(c);
}
break;
@@ -1942,7 +1938,7 @@
/* add char to pattern */
/* overflow check... */
if ((size_t)(p - pat) >= sizeof(pat) - 1) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
continue;
}
*p++ = c, *p = '\0';
@@ -1988,7 +1984,7 @@
return (i);
}
}
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
x_histp = histptr;
return (-1);
}
@@ -2094,7 +2090,7 @@
int col;
if (lastch == ' ' && !line_was_cleared && x_term_mode == 1) {
- shf_puts("\033[K", shl_out);
+ shf_puts(KSH_ESC_STRING "[K", shl_out);
line_was_cleared = true;
}
if (lastch == ' ' && line_was_cleared)
@@ -2168,11 +2164,11 @@
* to the one they want.
*/
if (xcp == xbuf) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
} else if (xcp == xep || Flag(FGMACS)) {
if (xcp - xbuf == 1) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
/*
@@ -2181,12 +2177,12 @@
*/
x_bs3(&xcp);
if (utf_mbtowc(&tmpa, xcp) == (size_t)-1) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
x_bs3(&xcp);
if (utf_mbtowc(&tmpb, xcp) == (size_t)-1) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
utf_wctomb(xcp, tmpa);
@@ -2199,12 +2195,12 @@
* cursor, move cursor position along one.
*/
if (utf_mbtowc(&tmpa, xcp) == (size_t)-1) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
x_bs3(&xcp);
if (utf_mbtowc(&tmpb, xcp) == (size_t)-1) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
utf_wctomb(xcp, tmpa);
@@ -2313,21 +2309,35 @@
return (KSTD);
}
-static int
-x_abort(int c MKSH_A_UNUSED)
+/* fake receiving an interrupt */
+static void
+x_intr(int signo, int c)
{
- /* x_zotc(c); */
+ x_vi_zotc(c);
+ *xep = '\0';
+ strip_nuls(xbuf, xep - xbuf);
+ if (*xbuf)
+ histsave(&source->line, xbuf, HIST_STORE, true);
xlp = xep = xcp = xbp = xbuf;
xlp_valid = true;
*xcp = 0;
x_modified();
+ x_flush();
+ trapsig(signo);
+ x_mode(false);
+ unwind(LSHELL);
+}
+
+static int
+x_abort(int c MKSH_A_UNUSED)
+{
return (KINTR);
}
static int
x_error(int c MKSH_A_UNUSED)
{
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
@@ -2387,19 +2397,18 @@
strdupx(news, cp, ap);
op = news;
while (*cp) {
- /* XXX -- should handle \^ escape? */
- if (*cp == '^') {
+ switch (*cp) {
+ case '^':
cp++;
- /*XXX or ^^ escape? this is ugly. */
- if (*cp >= '?')
- /* includes '?'; ASCII */
- *op++ = CTRL(*cp);
- else {
- *op++ = '^';
- cp--;
- }
- } else
+ *op++ = ksh_toctrl(*cp);
+ break;
+ case '\\':
+ if (cp[1] == '\\' || cp[1] == '^')
+ ++cp;
+ /* FALLTHROUGH */
+ default:
*op++ = *cp;
+ }
cp++;
}
*op = '\0';
@@ -2412,9 +2421,9 @@
{
char *p = *buf;
- if (ISCTRL(c)) {
+ if (ksh_isctrl(c)) {
*p++ = '^';
- *p++ = UNCTRL(c);
+ *p++ = ksh_unctrl(c);
} else
*p++ = c;
*p = 0;
@@ -2437,9 +2446,9 @@
int f = x_tab[prefix][key];
if (prefix)
- /* prefix == 1 || prefix == 2 */
- shf_puts(x_mapout(prefix == 1 ? CTRL('[') :
- prefix == 2 ? CTRL('X') : 0), shl_stdout);
+ /* prefix == 1 || prefix == 2 || prefix == 3 */
+ shf_puts(x_mapout(prefix == 1 ? CTRL_BO :
+ prefix == 2 ? CTRL_X : 0), shl_stdout);
#ifdef MKSH_SMALL
shprintf("%s = ", x_mapout(key));
#else
@@ -2603,7 +2612,7 @@
char *xr;
if (xmp == NULL) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
if (xmp > xcp) {
@@ -2625,7 +2634,7 @@
char *tmp;
if (xmp == NULL) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
tmp = xmp;
@@ -2703,7 +2712,7 @@
&start, &end, &words);
if (nwords == 0) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
x_goto(xbuf + start);
@@ -2713,7 +2722,7 @@
while (i < nwords) {
if (x_escape(words[i], strlen(words[i]), x_do_ins) < 0 ||
(++i < nwords && x_ins(T1space) < 0)) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
}
@@ -2737,7 +2746,7 @@
&start, &end, &words);
/* no match */
if (nwords == 0) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return;
}
if (type == CT_LIST) {
@@ -2894,9 +2903,10 @@
{
int width = 1;
- if (c == '\r' || c == '\n')
+ if (ctype(c, C_CR | C_LF))
x_col = 0;
if (x_col < xx_cols) {
+#ifndef MKSH_EBCDIC
if (UTFMODE && (c > 0x7F)) {
char utf_tmp[3];
size_t x;
@@ -2911,9 +2921,10 @@
x_putc(utf_tmp[2]);
width = utf_wcwidth(c);
} else
+#endif
x_putc(c);
switch (c) {
- case 7:
+ case KSH_BEL:
break;
case '\r':
case '\n':
@@ -2935,7 +2946,7 @@
{
int width = 1, c = **(const unsigned char **)cp;
- if (c == '\r' || c == '\n')
+ if (ctype(c, C_CR | C_LF))
x_col = 0;
if (x_col < xx_cols) {
if (UTFMODE && (c > 0x7F)) {
@@ -2944,7 +2955,13 @@
width = utf_widthadj(*cp, (const char **)&cp2);
if (cp2 == *cp + 1) {
(*cp)++;
+#ifdef MKSH_EBCDIC
+ x_putc(asc2rtt(0xEF));
+ x_putc(asc2rtt(0xBF));
+ x_putc(asc2rtt(0xBD));
+#else
shf_puts("\xEF\xBF\xBD", shl_out);
+#endif
} else
while (*cp < cp2)
x_putcf(*(*cp)++);
@@ -2953,7 +2970,7 @@
x_putc(c);
}
switch (c) {
- case 7:
+ case KSH_BEL:
break;
case '\r':
case '\n':
@@ -2997,7 +3014,7 @@
/* strip command prefix */
c &= 255;
- while (c >= 0 && ksh_isdigit(c)) {
+ while (c >= 0 && ctype(c, C_DIGIT)) {
n = n * 10 + ksh_numdig(c);
if (n > LINE)
/* upper bound for repeat */
@@ -3007,7 +3024,7 @@
}
if (c < 0 || first) {
x_set_arg_too_big:
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
x_arg = 1;
x_arg_defaulted = true;
} else {
@@ -3026,7 +3043,7 @@
int ret = x_do_comment(xbuf, xend - xbuf, &len);
if (ret < 0)
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
else {
x_modified();
xep = xbuf + len;
@@ -3049,7 +3066,7 @@
strdupx(v, KSH_VERSION, ATEMP);
xbuf = xbp = xcp = v;
- xend = xep = v + strlen(v);
+ xend = xep = strnul(v);
x_redraw('\r');
x_flush();
@@ -3077,7 +3094,7 @@
{
if (x_arg_defaulted) {
if (xep == xbuf) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
if (modified) {
@@ -3092,7 +3109,7 @@
"fc -e ${VISUAL:-${EDITOR:-vi}} --", x_arg);
else
strlcpy(xbuf, "fc -e ${VISUAL:-${EDITOR:-vi}} --", xend - xbuf);
- xep = xbuf + strlen(xbuf);
+ xep = strnul(xbuf);
return (x_newline('\n'));
}
#endif
@@ -3132,7 +3149,7 @@
last_arg = x_arg_defaulted ? -1 : x_arg;
xhp = histptr - (m - 1);
if ((xhp < history) || !(cp = *xhp)) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
x_modified();
return (KSTD);
}
@@ -3144,11 +3161,11 @@
/*
* ignore white-space after the last word
*/
- while (rcp > cp && is_cfs(*rcp))
+ while (rcp > cp && ctype(*rcp, C_CFS))
rcp--;
- while (rcp > cp && !is_cfs(*rcp))
+ while (rcp > cp && !ctype(*rcp, C_CFS))
rcp--;
- if (is_cfs(*rcp))
+ if (ctype(*rcp, C_CFS))
rcp++;
x_ins(rcp);
} else {
@@ -3159,16 +3176,16 @@
/*
* ignore white-space at start of line
*/
- while (*rcp && is_cfs(*rcp))
+ while (*rcp && ctype(*rcp, C_CFS))
rcp++;
while (x_arg-- > 0) {
- while (*rcp && !is_cfs(*rcp))
+ while (*rcp && !ctype(*rcp, C_CFS))
rcp++;
- while (*rcp && is_cfs(*rcp))
+ while (*rcp && ctype(*rcp, C_CFS))
rcp++;
}
cp = rcp;
- while (*rcp && !is_cfs(*rcp))
+ while (*rcp && !ctype(*rcp, C_CFS))
rcp++;
ch = *rcp;
*rcp = '\0';
@@ -3220,14 +3237,14 @@
char *cp = xcp;
if (cp == xep) {
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
return (KSTD);
}
while (x_arg--) {
/*
* first skip over any white-space
*/
- while (cp != xep && is_mfs(*cp))
+ while (cp != xep && ctype(*cp, C_MFS))
cp++;
/*
* do the first char on its own since it may be
@@ -3245,7 +3262,7 @@
/*
* now for the rest of the word
*/
- while (cp != xep && !is_mfs(*cp)) {
+ while (cp != xep && !ctype(*cp, C_MFS)) {
if (c == 'U')
/* uppercase */
*cp = ksh_toupper(*cp);
@@ -3324,17 +3341,17 @@
#endif
if (!edchars.erase)
- edchars.erase = CTRL('H');
+ edchars.erase = CTRL_H;
if (!edchars.kill)
- edchars.kill = CTRL('U');
+ edchars.kill = CTRL_U;
if (!edchars.intr)
- edchars.intr = CTRL('C');
+ edchars.intr = CTRL_C;
if (!edchars.quit)
- edchars.quit = CTRL('\\');
+ edchars.quit = CTRL_BK;
if (!edchars.eof)
- edchars.eof = CTRL('D');
+ edchars.eof = CTRL_D;
if (!edchars.werase)
- edchars.werase = CTRL('W');
+ edchars.werase = CTRL_W;
if (isedchar(edchars.erase)) {
bind_if_not_bound(0, edchars.erase, XFUNC_del_back);
@@ -3368,6 +3385,7 @@
static int vi_insert(int);
static int vi_cmd(int, const char *);
static int domove(int, const char *, int);
+static int domovebeg(void);
static int redo_insert(int);
static void yank_range(int, int);
static int bracktype(int);
@@ -3394,12 +3412,10 @@
static int expand_word(int);
static int complete_word(int, int);
static int print_expansions(struct edstate *, int);
-#define char_len(c) ((ISCTRL((unsigned char)c) && \
- /* but not C1 */ (unsigned char)c < 0x80) ? 2 : 1)
-static void x_vi_zotc(int);
static void vi_error(void);
static void vi_macro_reset(void);
static int x_vi_putbuf(const char *, size_t);
+#define char_len(c) (ksh_isctrl(c) ? 2 : 1)
#define vC 0x01 /* a valid command that isn't a vM, vE, vU */
#define vM 0x02 /* movement command (h, l, etc.) */
@@ -3410,14 +3426,14 @@
#define vZ 0x40 /* repeat count defaults to 0 (not 1) */
#define vS 0x80 /* search (/, ?) */
-#define is_bad(c) (classify[(c)&0x7f]&vB)
-#define is_cmd(c) (classify[(c)&0x7f]&(vM|vE|vC|vU))
-#define is_move(c) (classify[(c)&0x7f]&vM)
-#define is_extend(c) (classify[(c)&0x7f]&vE)
-#define is_long(c) (classify[(c)&0x7f]&vX)
-#define is_undoable(c) (!(classify[(c)&0x7f]&vU))
-#define is_srch(c) (classify[(c)&0x7f]&vS)
-#define is_zerocount(c) (classify[(c)&0x7f]&vZ)
+#define is_bad(c) (classify[rtt2asc(c) & 0x7F] & vB)
+#define is_cmd(c) (classify[rtt2asc(c) & 0x7F] & (vM | vE | vC | vU))
+#define is_move(c) (classify[rtt2asc(c) & 0x7F] & vM)
+#define is_extend(c) (classify[rtt2asc(c) & 0x7F] & vE)
+#define is_long(c) (classify[rtt2asc(c) & 0x7F] & vX)
+#define is_undoable(c) (!(classify[rtt2asc(c) & 0x7F] & vU))
+#define is_srch(c) (classify[rtt2asc(c) & 0x7F] & vS)
+#define is_zerocount(c) (classify[rtt2asc(c) & 0x7F] & vZ)
static const unsigned char classify[128] = {
/* 0 1 2 3 4 5 6 7 */
@@ -3587,13 +3603,14 @@
if (state != VLIT) {
if (isched(c, edchars.intr) ||
isched(c, edchars.quit)) {
+ /* shove input buffer away */
+ xbuf = ebuf.cbuf;
+ xep = xbuf;
+ if (ebuf.linelen > 0)
+ xep += ebuf.linelen;
/* pretend we got an interrupt */
- x_vi_zotc(c);
- x_flush();
- trapsig(isched(c, edchars.intr) ?
- SIGINT : SIGQUIT);
- x_mode(false);
- unwind(LSHELL);
+ x_intr(isched(c, edchars.intr) ?
+ SIGINT : SIGQUIT, c);
} else if (isched(c, edchars.eof) &&
state != VVERSION) {
if (vs->linelen == 0) {
@@ -3646,7 +3663,7 @@
default: ch = 0; goto vi_insert_failed;
}
if (insert != 0) {
- if (ch == CTRL('v')) {
+ if (ch == CTRL_V) {
state = VLIT;
ch = '^';
}
@@ -3667,11 +3684,11 @@
return (1);
}
} else {
- if (ch == '\r' || ch == '\n')
+ if (ctype(ch, C_CR | C_LF))
return (1);
cmdlen = 0;
argc1 = 0;
- if (ch >= ord('1') && ch <= ord('9')) {
+ if (ctype(ch, C_DIGIT) && ord(ch) != ord('0')) {
argc1 = ksh_numdig(ch);
state = VARG1;
} else {
@@ -3716,7 +3733,7 @@
break;
case VARG1:
- if (ksh_isdigit(ch))
+ if (ctype(ch, C_DIGIT))
argc1 = argc1 * 10 + ksh_numdig(ch);
else {
curcmd[cmdlen++] = ch;
@@ -3726,7 +3743,7 @@
case VEXTCMD:
argc2 = 0;
- if (ch >= ord('1') && ch <= ord('9')) {
+ if (ctype(ch, C_DIGIT) && ord(ch) != ord('0')) {
argc2 = ksh_numdig(ch);
state = VARG2;
return (0);
@@ -3742,7 +3759,7 @@
break;
case VARG2:
- if (ksh_isdigit(ch))
+ if (ctype(ch, C_DIGIT))
argc2 = argc2 * 10 + ksh_numdig(ch);
else {
if (argc1 == 0)
@@ -3760,7 +3777,7 @@
break;
case VXCH:
- if (ch == CTRL('['))
+ if (ch == CTRL_BO)
state = VNORMAL;
else {
curcmd[cmdlen++] = ch;
@@ -3769,7 +3786,7 @@
break;
case VSEARCH:
- if (ch == '\r' || ch == '\n' /*|| ch == CTRL('[')*/ ) {
+ if (ctype(ch, C_CR | C_LF) /* || ch == CTRL_BO */ ) {
restore_cbuf();
/* Repeat last search? */
if (srchlen == 0) {
@@ -3784,7 +3801,7 @@
memcpy(srchpat, locpat, srchlen + 1);
}
state = VCMD;
- } else if (isched(ch, edchars.erase) || ch == CTRL('h')) {
+ } else if (isched(ch, edchars.erase) || ch == CTRL_H) {
if (srchlen != 0) {
srchlen--;
vs->linelen -= char_len(locpat[srchlen]);
@@ -3825,12 +3842,12 @@
vi_error();
else {
locpat[srchlen++] = ch;
- if (ISCTRL(ch) && /* but not C1 */ ch < 0x80) {
+ if (ksh_isctrl(ch)) {
if ((size_t)vs->linelen + 2 >
(size_t)vs->cbufsize)
vi_error();
vs->cbuf[vs->linelen++] = '^';
- vs->cbuf[vs->linelen++] = UNCTRL(ch);
+ vs->cbuf[vs->linelen++] = ksh_unctrl(ch);
} else {
if (vs->linelen >= vs->cbufsize)
vi_error();
@@ -3903,8 +3920,8 @@
break;
case 0:
if (insert != 0) {
- if (lastcmd[0] == 's' || lastcmd[0] == 'c' ||
- lastcmd[0] == 'C') {
+ if (lastcmd[0] == 's' ||
+ ksh_eq(lastcmd[0], 'C', 'c')) {
if (redo_insert(1) != 0)
vi_error();
} else {
@@ -3942,7 +3959,7 @@
return (VXCH);
else if (ch == '.')
return (VREDO);
- else if (ch == CTRL('v'))
+ else if (ch == CTRL_V)
return (VVERSION);
else if (is_cmd(ch))
return (VCMD);
@@ -3955,7 +3972,7 @@
{
int tcursor;
- if (isched(ch, edchars.erase) || ch == CTRL('h')) {
+ if (isched(ch, edchars.erase) || ch == CTRL_H) {
if (insert == REPLACE) {
if (vs->cursor == undo->cursor) {
vi_error();
@@ -4012,7 +4029,7 @@
* buffer (if user inserts & deletes char, ibuf gets trashed and
* we don't want to use it)
*/
- if (first_insert && ch != CTRL('['))
+ if (first_insert && ch != CTRL_BO)
saved_inslen = 0;
switch (ch) {
case '\0':
@@ -4022,7 +4039,7 @@
case '\n':
return (1);
- case CTRL('['):
+ case CTRL_BO:
expanded = NONE;
if (first_insert) {
first_insert = false;
@@ -4033,26 +4050,25 @@
lastcmd[0] = 'a';
lastac = 1;
}
- if (lastcmd[0] == 's' || lastcmd[0] == 'c' ||
- lastcmd[0] == 'C')
+ if (lastcmd[0] == 's' || ksh_eq(lastcmd[0], 'C', 'c'))
return (redo_insert(0));
else
return (redo_insert(lastac - 1));
/* { start nonstandard vi commands */
- case CTRL('x'):
+ case CTRL_X:
expand_word(0);
break;
- case CTRL('f'):
+ case CTRL_F:
complete_word(0, 0);
break;
- case CTRL('e'):
+ case CTRL_E:
print_expansions(vs, 0);
break;
- case CTRL('i'):
+ case CTRL_I:
if (Flag(FVITABCOMPLETE)) {
complete_word(0, 0);
break;
@@ -4105,14 +4121,14 @@
lastac = argcnt;
memmove(lastcmd, cmd, MAXVICMD);
}
- switch (*cmd) {
+ switch (ord(*cmd)) {
- case CTRL('l'):
- case CTRL('r'):
+ case CTRL_L:
+ case CTRL_R:
redraw_line(true);
break;
- case '@':
+ case ord('@'):
{
static char alias[] = "_\0";
struct tbl *ap;
@@ -4153,7 +4169,7 @@
}
break;
- case 'a':
+ case ord('a'):
modified = 1;
hnum = hlast;
if (vs->linelen != 0)
@@ -4161,7 +4177,7 @@
insert = INSERT;
break;
- case 'A':
+ case ord('A'):
modified = 1;
hnum = hlast;
del_range(0, 0);
@@ -4169,36 +4185,35 @@
insert = INSERT;
break;
- case 'S':
- vs->cursor = domove(1, "^", 1);
+ case ord('S'):
+ vs->cursor = domovebeg();
del_range(vs->cursor, vs->linelen);
modified = 1;
hnum = hlast;
insert = INSERT;
break;
- case 'Y':
+ case ord('Y'):
cmd = "y$";
/* ahhhhhh... */
/* FALLTHROUGH */
- case 'c':
- case 'd':
- case 'y':
+ case ord('c'):
+ case ord('d'):
+ case ord('y'):
if (*cmd == cmd[1]) {
- c1 = *cmd == 'c' ? domove(1, "^", 1) : 0;
+ c1 = *cmd == 'c' ? domovebeg() : 0;
c2 = vs->linelen;
} else if (!is_move(cmd[1]))
return (-1);
else {
if ((ncursor = domove(argcnt, &cmd[1], 1)) < 0)
return (-1);
- if (*cmd == 'c' &&
- (cmd[1] == 'w' || cmd[1] == 'W') &&
- !ksh_isspace(vs->cbuf[vs->cursor])) {
+ if (*cmd == 'c' && ksh_eq(cmd[1], 'W', 'w') &&
+ !ctype(vs->cbuf[vs->cursor], C_SPACE)) {
do {
--ncursor;
- } while (ksh_isspace(vs->cbuf[ncursor]));
+ } while (ctype(vs->cbuf[ncursor], C_SPACE));
ncursor++;
}
if (ncursor > vs->cursor) {
@@ -4224,7 +4239,7 @@
}
break;
- case 'p':
+ case ord('p'):
modified = 1;
hnum = hlast;
if (vs->linelen != 0)
@@ -4238,7 +4253,7 @@
return (-1);
break;
- case 'P':
+ case ord('P'):
modified = 1;
hnum = hlast;
any = 0;
@@ -4251,25 +4266,25 @@
return (-1);
break;
- case 'C':
+ case ord('C'):
modified = 1;
hnum = hlast;
del_range(vs->cursor, vs->linelen);
insert = INSERT;
break;
- case 'D':
+ case ord('D'):
yank_range(vs->cursor, vs->linelen);
del_range(vs->cursor, vs->linelen);
if (vs->cursor != 0)
vs->cursor--;
break;
- case 'g':
+ case ord('g'):
if (!argcnt)
argcnt = hlast;
/* FALLTHROUGH */
- case 'G':
+ case ord('G'):
if (!argcnt)
argcnt = 1;
else
@@ -4282,22 +4297,22 @@
}
break;
- case 'i':
+ case ord('i'):
modified = 1;
hnum = hlast;
insert = INSERT;
break;
- case 'I':
+ case ord('I'):
modified = 1;
hnum = hlast;
- vs->cursor = domove(1, "^", 1);
+ vs->cursor = domovebeg();
insert = INSERT;
break;
- case 'j':
- case '+':
- case CTRL('n'):
+ case ord('j'):
+ case ord('+'):
+ case CTRL_N:
if (grabhist(modified, hnum + argcnt) < 0)
return (-1);
else {
@@ -4306,9 +4321,9 @@
}
break;
- case 'k':
- case '-':
- case CTRL('p'):
+ case ord('k'):
+ case ord('-'):
+ case CTRL_P:
if (grabhist(modified, hnum - argcnt) < 0)
return (-1);
else {
@@ -4317,7 +4332,7 @@
}
break;
- case 'r':
+ case ord('r'):
if (vs->linelen == 0)
return (-1);
modified = 1;
@@ -4335,13 +4350,13 @@
}
break;
- case 'R':
+ case ord('R'):
modified = 1;
hnum = hlast;
insert = REPLACE;
break;
- case 's':
+ case ord('s'):
if (vs->linelen == 0)
return (-1);
modified = 1;
@@ -4352,7 +4367,7 @@
insert = INSERT;
break;
- case 'v':
+ case ord('v'):
if (!argcnt) {
if (vs->linelen == 0)
return (-1);
@@ -4375,7 +4390,7 @@
vs->linelen = strlen(vs->cbuf);
return (2);
- case 'x':
+ case ord('x'):
if (vs->linelen == 0)
return (-1);
modified = 1;
@@ -4386,7 +4401,7 @@
del_range(vs->cursor, vs->cursor + argcnt);
break;
- case 'X':
+ case ord('X'):
if (vs->cursor > 0) {
modified = 1;
hnum = hlast;
@@ -4399,13 +4414,13 @@
return (-1);
break;
- case 'u':
+ case ord('u'):
t = vs;
vs = undo;
undo = t;
break;
- case 'U':
+ case ord('U'):
if (!modified)
return (-1);
if (grabhist(modified, ohnum) < 0)
@@ -4414,19 +4429,19 @@
hnum = ohnum;
break;
- case '?':
+ case ord('?'):
if (hnum == hlast)
hnum = -1;
/* ahhh */
/* FALLTHROUGH */
- case '/':
+ case ord('/'):
c3 = 1;
srchlen = 0;
lastsearch = *cmd;
/* FALLTHROUGH */
- case 'n':
- case 'N':
+ case ord('n'):
+ case ord('N'):
if (lastsearch == ' ')
return (-1);
if (lastsearch == '?')
@@ -4453,7 +4468,7 @@
return (0);
}
break;
- case '_':
+ case ord('_'):
{
bool inspace;
char *p, *sp;
@@ -4461,14 +4476,13 @@
if (histnum(-1) < 0)
return (-1);
p = *histpos();
-#define issp(c) (ksh_isspace(c) || (c) == '\n')
if (argcnt) {
- while (*p && issp(*p))
+ while (ctype(*p, C_SPACE))
p++;
while (*p && --argcnt) {
- while (*p && !issp(*p))
+ while (*p && !ctype(*p, C_SPACE))
p++;
- while (*p && issp(*p))
+ while (ctype(*p, C_SPACE))
p++;
}
if (!*p)
@@ -4478,7 +4492,7 @@
sp = p;
inspace = false;
while (*p) {
- if (issp(*p))
+ if (ctype(*p, C_SPACE))
inspace = true;
else if (inspace) {
inspace = false;
@@ -4492,7 +4506,7 @@
hnum = hlast;
if (vs->cursor != vs->linelen)
vs->cursor++;
- while (*p && !issp(*p)) {
+ while (*p && !ctype(*p, C_SPACE)) {
argcnt++;
p++;
}
@@ -4506,7 +4520,7 @@
}
break;
- case '~':
+ case ord('~'):
{
char *p;
int i;
@@ -4515,11 +4529,11 @@
return (-1);
for (i = 0; i < argcnt; i++) {
p = &vs->cbuf[vs->cursor];
- if (ksh_islower(*p)) {
+ if (ctype(*p, C_LOWER)) {
modified = 1;
hnum = hlast;
*p = ksh_toupper(*p);
- } else if (ksh_isupper(*p)) {
+ } else if (ctype(*p, C_UPPER)) {
modified = 1;
hnum = hlast;
*p = ksh_tolower(*p);
@@ -4530,7 +4544,7 @@
break;
}
- case '#':
+ case ord('#'):
{
int ret = x_do_comment(vs->cbuf, vs->cbufsize,
&vs->linelen);
@@ -4540,44 +4554,44 @@
}
/* AT&T ksh */
- case '=':
+ case ord('='):
/* Nonstandard vi/ksh */
- case CTRL('e'):
+ case CTRL_E:
print_expansions(vs, 1);
break;
/* Nonstandard vi/ksh */
- case CTRL('i'):
+ case CTRL_I:
if (!Flag(FVITABCOMPLETE))
return (-1);
complete_word(1, argcnt);
break;
/* some annoying AT&T kshs */
- case CTRL('['):
+ case CTRL_BO:
if (!Flag(FVIESCCOMPLETE))
return (-1);
/* FALLTHROUGH */
/* AT&T ksh */
- case '\\':
+ case ord('\\'):
/* Nonstandard vi/ksh */
- case CTRL('f'):
+ case CTRL_F:
complete_word(1, argcnt);
break;
/* AT&T ksh */
- case '*':
+ case ord('*'):
/* Nonstandard vi/ksh */
- case CTRL('x'):
+ case CTRL_X:
expand_word(1);
break;
/* mksh: cursor movement */
- case '[':
- case 'O':
+ case ord('['):
+ case ord('O'):
state = VPREFIX2;
if (vs->linelen != 0)
vs->cursor++;
@@ -4596,20 +4610,20 @@
int ncursor = 0, i = 0, t;
unsigned int bcount;
- switch (*cmd) {
- case 'b':
+ switch (ord(*cmd)) {
+ case ord('b'):
if (!sub && vs->cursor == 0)
return (-1);
ncursor = backword(argcnt);
break;
- case 'B':
+ case ord('B'):
if (!sub && vs->cursor == 0)
return (-1);
ncursor = Backword(argcnt);
break;
- case 'e':
+ case ord('e'):
if (!sub && vs->cursor + 1 >= vs->linelen)
return (-1);
ncursor = endword(argcnt);
@@ -4617,7 +4631,7 @@
ncursor++;
break;
- case 'E':
+ case ord('E'):
if (!sub && vs->cursor + 1 >= vs->linelen)
return (-1);
ncursor = Endword(argcnt);
@@ -4625,18 +4639,18 @@
ncursor++;
break;
- case 'f':
- case 'F':
- case 't':
- case 'T':
+ case ord('f'):
+ case ord('F'):
+ case ord('t'):
+ case ord('T'):
fsavecmd = *cmd;
fsavech = cmd[1];
/* FALLTHROUGH */
- case ',':
- case ';':
+ case ord(','):
+ case ord(';'):
if (fsavecmd == ' ')
return (-1);
- i = fsavecmd == 'f' || fsavecmd == 'F';
+ i = ksh_eq(fsavecmd, 'F', 'f');
t = fsavecmd > 'a';
if (*cmd == ',')
t = !t;
@@ -4647,8 +4661,8 @@
ncursor++;
break;
- case 'h':
- case CTRL('h'):
+ case ord('h'):
+ case CTRL_H:
if (!sub && vs->cursor == 0)
return (-1);
ncursor = vs->cursor - argcnt;
@@ -4656,8 +4670,8 @@
ncursor = 0;
break;
- case ' ':
- case 'l':
+ case ord(' '):
+ case ord('l'):
if (!sub && vs->cursor + 1 >= vs->linelen)
return (-1);
if (vs->linelen != 0) {
@@ -4667,30 +4681,27 @@
}
break;
- case 'w':
+ case ord('w'):
if (!sub && vs->cursor + 1 >= vs->linelen)
return (-1);
ncursor = forwword(argcnt);
break;
- case 'W':
+ case ord('W'):
if (!sub && vs->cursor + 1 >= vs->linelen)
return (-1);
ncursor = Forwword(argcnt);
break;
- case '0':
+ case ord('0'):
ncursor = 0;
break;
- case '^':
- ncursor = 0;
- while (ncursor < vs->linelen - 1 &&
- ksh_isspace(vs->cbuf[ncursor]))
- ncursor++;
+ case ord('^'):
+ ncursor = domovebeg();
break;
- case '|':
+ case ord('|'):
ncursor = argcnt;
if (ncursor > vs->linelen)
ncursor = vs->linelen;
@@ -4698,14 +4709,14 @@
ncursor--;
break;
- case '$':
+ case ord('$'):
if (vs->linelen != 0)
ncursor = vs->linelen;
else
ncursor = 0;
break;
- case '%':
+ case ord('%'):
ncursor = vs->cursor;
while (ncursor < vs->linelen &&
(i = bracktype(vs->cbuf[ncursor])) == 0)
@@ -4738,6 +4749,17 @@
}
static int
+domovebeg(void)
+{
+ int ncursor = 0;
+
+ while (ncursor < vs->linelen - 1 &&
+ ctype(vs->cbuf[ncursor], C_SPACE))
+ ncursor++;
+ return (ncursor);
+}
+
+static int
redo_insert(int count)
{
while (count-- > 0)
@@ -4760,24 +4782,24 @@
static int
bracktype(int ch)
{
- switch (ch) {
+ switch (ord(ch)) {
- case '(':
+ case ord('('):
return (1);
- case '[':
+ case ord('['):
return (2);
- case '{':
+ case ord('{'):
return (3);
- case ')':
+ case ord(')'):
return (-1);
- case ']':
+ case ord(']'):
return (-2);
- case '}':
+ case ord('}'):
return (-3);
default:
@@ -4912,17 +4934,16 @@
ncursor = vs->cursor;
while (ncursor < vs->linelen && argcnt--) {
- if (ksh_isalnux(vs->cbuf[ncursor]))
+ if (ctype(vs->cbuf[ncursor], C_ALNUX))
while (ncursor < vs->linelen &&
- ksh_isalnux(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_ALNUX))
ncursor++;
- else if (!ksh_isspace(vs->cbuf[ncursor]))
+ else if (!ctype(vs->cbuf[ncursor], C_SPACE))
while (ncursor < vs->linelen &&
- !ksh_isalnux(vs->cbuf[ncursor]) &&
- !ksh_isspace(vs->cbuf[ncursor]))
+ !ctype(vs->cbuf[ncursor], C_ALNUX | C_SPACE))
ncursor++;
while (ncursor < vs->linelen &&
- ksh_isspace(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_SPACE))
ncursor++;
}
return (ncursor);
@@ -4935,17 +4956,16 @@
ncursor = vs->cursor;
while (ncursor > 0 && argcnt--) {
- while (--ncursor > 0 && ksh_isspace(vs->cbuf[ncursor]))
+ while (--ncursor > 0 && ctype(vs->cbuf[ncursor], C_SPACE))
;
if (ncursor > 0) {
- if (ksh_isalnux(vs->cbuf[ncursor]))
+ if (ctype(vs->cbuf[ncursor], C_ALNUX))
while (--ncursor >= 0 &&
- ksh_isalnux(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_ALNUX))
;
else
while (--ncursor >= 0 &&
- !ksh_isalnux(vs->cbuf[ncursor]) &&
- !ksh_isspace(vs->cbuf[ncursor]))
+ !ctype(vs->cbuf[ncursor], C_ALNUX | C_SPACE))
;
ncursor++;
}
@@ -4961,17 +4981,16 @@
ncursor = vs->cursor;
while (ncursor < vs->linelen && argcnt--) {
while (++ncursor < vs->linelen - 1 &&
- ksh_isspace(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_SPACE))
;
if (ncursor < vs->linelen - 1) {
- if (ksh_isalnux(vs->cbuf[ncursor]))
+ if (ctype(vs->cbuf[ncursor], C_ALNUX))
while (++ncursor < vs->linelen &&
- ksh_isalnux(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_ALNUX))
;
else
while (++ncursor < vs->linelen &&
- !ksh_isalnux(vs->cbuf[ncursor]) &&
- !ksh_isspace(vs->cbuf[ncursor]))
+ !ctype(vs->cbuf[ncursor], C_ALNUX | C_SPACE))
;
ncursor--;
}
@@ -4987,10 +5006,10 @@
ncursor = vs->cursor;
while (ncursor < vs->linelen && argcnt--) {
while (ncursor < vs->linelen &&
- !ksh_isspace(vs->cbuf[ncursor]))
+ !ctype(vs->cbuf[ncursor], C_SPACE))
ncursor++;
while (ncursor < vs->linelen &&
- ksh_isspace(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_SPACE))
ncursor++;
}
return (ncursor);
@@ -5003,9 +5022,9 @@
ncursor = vs->cursor;
while (ncursor > 0 && argcnt--) {
- while (--ncursor >= 0 && ksh_isspace(vs->cbuf[ncursor]))
+ while (--ncursor >= 0 && ctype(vs->cbuf[ncursor], C_SPACE))
;
- while (ncursor >= 0 && !ksh_isspace(vs->cbuf[ncursor]))
+ while (ncursor >= 0 && !ctype(vs->cbuf[ncursor], C_SPACE))
ncursor--;
ncursor++;
}
@@ -5020,11 +5039,11 @@
ncursor = vs->cursor;
while (ncursor < vs->linelen - 1 && argcnt--) {
while (++ncursor < vs->linelen - 1 &&
- ksh_isspace(vs->cbuf[ncursor]))
+ ctype(vs->cbuf[ncursor], C_SPACE))
;
if (ncursor < vs->linelen - 1) {
while (++ncursor < vs->linelen &&
- !ksh_isspace(vs->cbuf[ncursor]))
+ !ctype(vs->cbuf[ncursor], C_SPACE))
;
ncursor--;
}
@@ -5187,10 +5206,10 @@
*twb1++ = ' ';
} while (++col < winwidth && (col & 7) != 0);
else if (col < winwidth) {
- if (ISCTRL(ch) && /* but not C1 */ ch < 0x80) {
+ if (ksh_isctrl(ch)) {
*twb1++ = '^';
if (++col < winwidth) {
- *twb1++ = UNCTRL(ch);
+ *twb1++ = ksh_unctrl(ch);
col++;
}
} else {
@@ -5460,24 +5479,26 @@
redraw_line(false);
return (0);
}
+#endif /* !MKSH_S_NOVI */
/* Similar to x_zotc(emacs.c), but no tab weirdness */
static void
x_vi_zotc(int c)
{
- if (ISCTRL(c)) {
+ if (ksh_isctrl(c)) {
x_putc('^');
- c = UNCTRL(c);
+ c = ksh_unctrl(c);
}
x_putc(c);
}
+#if !MKSH_S_NOVI
static void
vi_error(void)
{
/* Beem out of any macros as soon as an error occurs */
vi_macro_reset();
- x_putc(7);
+ x_putc(KSH_BEL);
x_flush();
}
@@ -5602,7 +5623,7 @@
if (cp == NULL) {
/* command cannot be parsed */
x_eval_region_err:
- x_e_putc2(7);
+ x_e_putc2(KSH_BEL);
x_redraw('\r');
return (KSTD);
}
diff --git a/src/eval.c b/src/eval.c
index 23894d6..7a892c0 100644
--- a/src/eval.c
+++ b/src/eval.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/eval.c,v 1.201 2017/04/06 01:59:54 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/eval.c,v 1.215 2017/08/28 23:27:51 tg Exp $");
/*
* string expansion
@@ -65,6 +65,12 @@
#define IFS_IWS 3 /* beginning of word, ignore IFS WS */
#define IFS_QUOTE 4 /* beg.w/quote, become IFS_WORD unless "$@" */
+#define STYPE_CHAR 0xFF
+#define STYPE_DBL 0x100
+#define STYPE_AT 0x200
+#define STYPE_SINGLE 0x2FF
+#define STYPE_MASK 0x300
+
static int varsub(Expand *, const char *, const char *, int *, int *);
static int comsub(Expand *, const char *, int);
static char *valsub(struct op *, Area *);
@@ -277,18 +283,18 @@
switch (type) {
case XBASE:
/* original prefixed string */
- c = *sp++;
+ c = ord(*sp++);
switch (c) {
case EOS:
c = 0;
break;
case CHAR:
- c = *sp++;
+ c = ord(*sp++);
break;
case QCHAR:
/* temporary quote */
quote |= 2;
- c = *sp++;
+ c = ord(*sp++);
break;
case OQUOTE:
if (word != IFS_WORD)
@@ -314,21 +320,21 @@
case COMASUB:
case COMSUB:
*dp++ = '(';
- c = ')';
+ c = ord(')');
break;
case FUNASUB:
case FUNSUB:
case VALSUB:
*dp++ = '{';
*dp++ = c == VALSUB ? '|' : ' ';
- c = '}';
+ c = ord('}');
break;
}
while (*sp != '\0') {
Xcheck(ds, dp);
*dp++ = *sp++;
}
- if (c == '}')
+ if (c == ord('}'))
*dp++ = ';';
*dp++ = c;
} else {
@@ -429,12 +435,12 @@
/* skip qualifier(s) */
if (stype)
sp += slen;
- switch (stype & 0x17F) {
- case 0x100 | '#':
+ switch (stype & STYPE_SINGLE) {
+ case ord('#') | STYPE_AT:
x.str = shf_smprintf("%08X",
(unsigned int)hash(str_val(st->var)));
break;
- case 0x100 | 'Q': {
+ case ord('Q') | STYPE_AT: {
struct shf shf;
shf_sopen(NULL, 0, SHF_WR|SHF_DYNAMIC, &shf);
@@ -442,7 +448,7 @@
x.str = shf_sclose(&shf);
break;
}
- case '0': {
+ case ord('0'): {
char *beg, *mid, *end, *stg;
mksh_ari_t from = 0, num = -1, flen, finc = 0;
@@ -450,13 +456,13 @@
mid = beg + (wdscan(sp, ADELIM) - sp);
stg = beg + (wdscan(sp, CSUBST) - sp);
mid[-2] = EOS;
- if (mid[-1] == /*{*/'}') {
+ if (ord(mid[-1]) == ord(/*{*/ '}')) {
sp += mid - beg - 1;
end = NULL;
} else {
end = mid +
(wdscan(mid, ADELIM) - mid);
- if (end[-1] != /*{*/ '}')
+ if (ord(end[-1]) != ord(/*{*/ '}'))
/* more than max delimiters */
goto unwind_substsyn;
end[-2] = EOS;
@@ -489,8 +495,8 @@
strndupx(x.str, beg, num, ATEMP);
goto do_CSUBST;
}
- case 0x100 | '/':
- case '/': {
+ case ord('/') | STYPE_AT:
+ case ord('/'): {
char *s, *p, *d, *sbeg, *end;
char *pat = NULL, *rrep = null;
char fpat = 0, *tpat1, *tpat2;
@@ -500,18 +506,18 @@
p = s + (wdscan(sp, ADELIM) - sp);
d = s + (wdscan(sp, CSUBST) - sp);
p[-2] = EOS;
- if (p[-1] == /*{*/'}')
+ if (ord(p[-1]) == ord(/*{*/ '}'))
d = NULL;
else
d[-2] = EOS;
sp += (d ? d : p) - s - 1;
- if (!(stype & 0x180) &&
+ if (!(stype & STYPE_MASK) &&
s[0] == CHAR &&
- (s[1] == '#' || s[1] == '%'))
+ ctype(s[1], C_SUB2))
fpat = s[1];
wpat = s + (fpat ? 2 : 0);
wrep = d ? p : NULL;
- if (!(stype & 0x100)) {
+ if (!(stype & STYPE_AT)) {
rrep = wrep ? evalstr(wrep,
DOTILDE | DOSCALAR) :
null;
@@ -531,21 +537,21 @@
*/
goto no_repl;
}
- if ((stype & 0x180) &&
+ if ((stype & STYPE_MASK) &&
gmatchx(null, pat, false)) {
/*
* pattern matches empty
* string => don't loop
*/
- stype &= ~0x180;
+ stype &= ~STYPE_MASK;
}
/* first see if we have any match at all */
- if (fpat == '#') {
+ if (ord(fpat) == ord('#')) {
/* anchor at the beginning */
tpat1 = shf_smprintf("%s%c*", pat, MAGIC);
tpat2 = tpat1;
- } else if (fpat == '%') {
+ } else if (ord(fpat) == ord('%')) {
/* anchor at the end */
tpat1 = shf_smprintf("%c*%s", MAGIC, pat);
tpat2 = pat;
@@ -563,7 +569,7 @@
goto end_repl;
end = strnul(s);
/* now anchor the beginning of the match */
- if (fpat != '#')
+ if (ord(fpat) != ord('#'))
while (sbeg <= end) {
if (gmatchx(sbeg, tpat2, false))
break;
@@ -572,7 +578,7 @@
}
/* now anchor the end of the match */
p = end;
- if (fpat != '%')
+ if (ord(fpat) != ord('%'))
while (p >= sbeg) {
bool gotmatch;
@@ -587,7 +593,7 @@
strndupx(end, sbeg, p - sbeg, ATEMP);
record_match(end);
afree(end, ATEMP);
- if (stype & 0x100) {
+ if (stype & STYPE_AT) {
if (rrep != null)
afree(rrep, ATEMP);
rrep = wrep ? evalstr(wrep,
@@ -600,11 +606,11 @@
sbeg = d + (sbeg - s) + strlen(rrep);
afree(s, ATEMP);
s = d;
- if (stype & 0x100) {
+ if (stype & STYPE_AT) {
afree(tpat1, ATEMP);
afree(pat, ATEMP);
goto again_search;
- } else if (stype & 0x80)
+ } else if (stype & STYPE_DBL)
goto again_repl;
end_repl:
afree(tpat1, ATEMP);
@@ -616,8 +622,8 @@
afree(ws, ATEMP);
goto do_CSUBST;
}
- case '#':
- case '%':
+ case ord('#'):
+ case ord('%'):
/* ! DOBLANK,DOBRACE */
f = (f & DONTRUNCOMMAND) |
DOPAT | DOTILDE |
@@ -631,10 +637,10 @@
*/
if (!Flag(FSH)) {
*dp++ = MAGIC;
- *dp++ = 0x80 | '@';
+ *dp++ = ord(0x80 | '@');
}
break;
- case '=':
+ case ord('='):
/*
* Tilde expansion for string
* variables in POSIX mode is
@@ -658,7 +664,7 @@
f &= ~(DOBLANK|DOGLOB|DOBRACE);
tilde_ok = 1;
break;
- case '?':
+ case ord('?'):
if (*sp == CSUBST)
errorf("%s: parameter null or not set",
st->var->name);
@@ -692,9 +698,9 @@
f = st->f;
if (f & DOBLANK)
doblank--;
- switch (st->stype & 0x17F) {
- case '#':
- case '%':
+ switch (st->stype & STYPE_SINGLE) {
+ case ord('#'):
+ case ord('%'):
if (!Flag(FSH)) {
/* Append end-pattern */
*dp++ = MAGIC;
@@ -724,7 +730,7 @@
doblank++;
st = st->prev;
continue;
- case '=':
+ case ord('='):
/*
* Restore our position and substitute
* the value of st->var (may not be
@@ -757,17 +763,17 @@
st = st->prev;
word = quote || (!*x.str && (f & DOSCALAR)) ? IFS_WORD : IFS_IWS;
continue;
- case '?':
+ case ord('?'):
dp = Xrestpos(ds, dp, st->base);
errorf(Tf_sD_s, st->var->name,
debunk(dp, dp, strlen(dp) + 1));
break;
- case '0':
- case 0x100 | '/':
- case '/':
- case 0x100 | '#':
- case 0x100 | 'Q':
+ case ord('0'):
+ case ord('/') | STYPE_AT:
+ case ord('/'):
+ case ord('#') | STYPE_AT:
+ case ord('Q') | STYPE_AT:
dp = Xrestpos(ds, dp, st->base);
type = XSUB;
word = quote || (!*x.str && (f & DOSCALAR)) ? IFS_WORD : IFS_IWS;
@@ -845,7 +851,7 @@
doblank--;
continue;
}
- c = ifs0;
+ c = ord(ifs0);
if ((f & DOHEREDOC)) {
/* pseudo-field-split reliably */
if (c == 0)
@@ -891,10 +897,7 @@
--newlines;
} else {
while ((c = shf_getc(x.u.shf)) == 0 ||
-#ifdef MKSH_WITH_TEXTMODE
- c == '\r' ||
-#endif
- c == '\n') {
+ ctype(c, C_NL)) {
#ifdef MKSH_WITH_TEXTMODE
if (c == '\r') {
c = shf_getc(x.u.shf);
@@ -999,11 +1002,11 @@
tilde_ok <<= 1;
/* mark any special second pass chars */
if (!quote)
- switch (c) {
- case '[':
- case '!':
- case '-':
- case ']':
+ switch (ord(c)) {
+ case ord('['):
+ case ord('!'):
+ case ord('-'):
+ case ord(']'):
/*
* For character classes - doesn't hurt
* to have magic !,-,]s outside of
@@ -1011,28 +1014,29 @@
*/
if (f & (DOPAT | DOGLOB)) {
fdo |= DOMAGIC;
- if (c == '[')
+ if (c == ord('['))
fdo |= f & DOGLOB;
*dp++ = MAGIC;
}
break;
- case '*':
- case '?':
+ case ord('*'):
+ case ord('?'):
if (f & (DOPAT | DOGLOB)) {
fdo |= DOMAGIC | (f & DOGLOB);
*dp++ = MAGIC;
}
break;
- case '{':
- case '}':
- case ',':
- if ((f & DOBRACE) && (c == '{' /*}*/ ||
+ case ord('{'):
+ case ord('}'):
+ case ord(','):
+ if ((f & DOBRACE) &&
+ (ord(c) == ord('{' /*}*/) ||
(fdo & DOBRACE))) {
fdo |= DOBRACE|DOMAGIC;
*dp++ = MAGIC;
}
break;
- case '=':
+ case ord('='):
/* Note first unquoted = for ~ */
if (!(f & DOTEMP) && (!Flag(FPOSIX) ||
(f & DOASNTILDE)) && !saw_eq) {
@@ -1040,13 +1044,13 @@
tilde_ok = 1;
}
break;
- case ':':
+ case ord(':'):
/* : */
/* Note unquoted : for ~ */
if (!(f & DOTEMP) && (f & DOASNTILDE))
tilde_ok = 1;
break;
- case '~':
+ case ord('~'):
/*
* tilde_ok is reset whenever
* any of ' " $( $(( ${ } are seen.
@@ -1118,7 +1122,7 @@
struct tbl *vp;
bool zero_ok = false;
- if ((stype = sp[0]) == '\0')
+ if ((stype = ord(sp[0])) == '\0')
/* Bad variable name */
return (-1);
@@ -1128,20 +1132,20 @@
* ${#var}, string length (-U: characters, +U: octets) or array size
* ${%var}, string width (-U: screen columns, +U: octets)
*/
- c = sp[1];
- if (stype == '%' && c == '\0')
+ c = ord(sp[1]);
+ if (stype == ord('%') && c == '\0')
return (-1);
- if ((stype == '#' || stype == '%') && c != '\0') {
+ if (ctype(stype, C_SUB2) && c != '\0') {
/* Can't have any modifiers for ${#...} or ${%...} */
if (*word != CSUBST)
return (-1);
sp++;
/* Check for size of array */
- if ((p = cstrchr(sp, '[')) && (p[1] == '*' || p[1] == '@') &&
- p[2] == ']') {
+ if ((p = cstrchr(sp, '[')) && (ord(p[1]) == ord('*') ||
+ ord(p[1]) == ord('@')) && ord(p[2]) == ord(']')) {
int n = 0;
- if (stype != '#')
+ if (stype != ord('#'))
return (-1);
vp = global(arrayname(sp));
if (vp->flag & (ISSET|ARRAY))
@@ -1150,14 +1154,14 @@
if (vp->flag & ISSET)
n++;
c = n;
- } else if (c == '*' || c == '@') {
- if (stype != '#')
+ } else if (c == ord('*') || c == ord('@')) {
+ if (stype != ord('#'))
return (-1);
c = e->loc->argc;
} else {
p = str_val(global(sp));
zero_ok = p != null;
- if (stype == '#')
+ if (stype == ord('#'))
c = utflen(p);
else {
/* partial utf_mbswidth reimplementation */
@@ -1171,7 +1175,7 @@
if (!UTFMODE || (len = utf_mbtowc(&wc,
s)) == (size_t)-1)
/* not UTFMODE or not UTF-8 */
- wc = (unsigned char)(*s++);
+ wc = rtt2asc(*s++);
else
/* UTFMODE and UTF-8 */
s += len;
@@ -1192,11 +1196,11 @@
xp->str = shf_smprintf(Tf_d, c);
return (XSUB);
}
- if (stype == '!' && c != '\0' && *word == CSUBST) {
+ if (stype == ord('!') && c != '\0' && *word == CSUBST) {
sp++;
- if ((p = cstrchr(sp, '[')) && (p[1] == '*' || p[1] == '@') &&
- p[2] == ']') {
- c = '!';
+ if ((p = cstrchr(sp, '[')) && (ord(p[1]) == ord('*') ||
+ ord(p[1]) == ord('@')) && ord(p[2]) == ord(']')) {
+ c = ord('!');
stype = 0;
goto arraynames;
}
@@ -1209,43 +1213,46 @@
/* Check for qualifiers in word part */
stype = 0;
- c = word[slen + 0] == CHAR ? word[slen + 1] : 0;
- if (c == ':') {
+ c = word[slen + 0] == CHAR ? ord(word[slen + 1]) : 0;
+ if (c == ord(':')) {
slen += 2;
- stype = 0x80;
- c = word[slen + 0] == CHAR ? word[slen + 1] : 0;
+ stype = STYPE_DBL;
+ c = word[slen + 0] == CHAR ? ord(word[slen + 1]) : 0;
}
- if (!stype && c == '/') {
+ if (!stype && c == ord('/')) {
slen += 2;
stype = c;
- if (word[slen] == ADELIM && word[slen + 1] == c) {
+ if (word[slen] == ADELIM &&
+ ord(word[slen + 1]) == (unsigned int)c) {
slen += 2;
- stype |= 0x80;
+ stype |= STYPE_DBL;
}
- } else if (stype == 0x80 && (c == ' ' || c == '0')) {
- stype |= '0';
- } else if (ctype(c, C_SUBOP1)) {
+ } else if (stype == STYPE_DBL && (c == ord(' ') || c == ord('0'))) {
+ stype |= ord('0');
+ } else if (ctype(c, C_SUB1)) {
slen += 2;
stype |= c;
- } else if (ksh_issubop2(c)) {
+ } else if (ctype(c, C_SUB2)) {
/* Note: ksh88 allows :%, :%%, etc */
slen += 2;
stype = c;
- if (word[slen + 0] == CHAR && c == word[slen + 1]) {
- stype |= 0x80;
+ if (word[slen + 0] == CHAR &&
+ ord(word[slen + 1]) == (unsigned int)c) {
+ stype |= STYPE_DBL;
slen += 2;
}
- } else if (c == '@') {
+ } else if (c == ord('@')) {
/* @x where x is command char */
- switch (c = word[slen + 2] == CHAR ? word[slen + 3] : 0) {
- case '#':
- case '/':
- case 'Q':
+ switch (c = ord(word[slen + 2]) == CHAR ?
+ ord(word[slen + 3]) : 0) {
+ case ord('#'):
+ case ord('/'):
+ case ord('Q'):
break;
default:
return (-1);
}
- stype |= 0x100 | c;
+ stype |= STYPE_AT | c;
slen += 4;
} else if (stype)
/* : is not ok */
@@ -1253,51 +1260,51 @@
if (!stype && *word != CSUBST)
return (-1);
- c = sp[0];
- if (c == '*' || c == '@') {
- switch (stype & 0x17F) {
+ c = ord(sp[0]);
+ if (c == ord('*') || c == ord('@')) {
+ switch (stype & STYPE_SINGLE) {
/* can't assign to a vector */
- case '=':
+ case ord('='):
/* can't trim a vector (yet) */
- case '%':
- case '#':
- case '?':
- case '0':
- case 0x100 | '/':
- case '/':
- case 0x100 | '#':
- case 0x100 | 'Q':
+ case ord('%'):
+ case ord('#'):
+ case ord('?'):
+ case ord('0'):
+ case ord('/') | STYPE_AT:
+ case ord('/'):
+ case ord('#') | STYPE_AT:
+ case ord('Q') | STYPE_AT:
return (-1);
}
if (e->loc->argc == 0) {
xp->str = null;
xp->var = global(sp);
- state = c == '@' ? XNULLSUB : XSUB;
+ state = c == ord('@') ? XNULLSUB : XSUB;
} else {
xp->u.strv = (const char **)e->loc->argv + 1;
xp->str = *xp->u.strv++;
/* $@ */
- xp->split = tobool(c == '@');
+ xp->split = tobool(c == ord('@'));
state = XARG;
}
/* POSIX 2009? */
zero_ok = true;
- } else if ((p = cstrchr(sp, '[')) && (p[1] == '*' || p[1] == '@') &&
- p[2] == ']') {
+ } else if ((p = cstrchr(sp, '[')) && (ord(p[1]) == ord('*') ||
+ ord(p[1]) == ord('@')) && ord(p[2]) == ord(']')) {
XPtrV wv;
- switch (stype & 0x17F) {
+ switch (stype & STYPE_SINGLE) {
/* can't assign to a vector */
- case '=':
+ case ord('='):
/* can't trim a vector (yet) */
- case '%':
- case '#':
- case '?':
- case '0':
- case 0x100 | '/':
- case '/':
- case 0x100 | '#':
- case 0x100 | 'Q':
+ case ord('%'):
+ case ord('#'):
+ case ord('?'):
+ case ord('0'):
+ case ord('/') | STYPE_AT:
+ case ord('/'):
+ case ord('#') | STYPE_AT:
+ case ord('Q') | STYPE_AT:
return (-1);
}
c = 0;
@@ -1307,45 +1314,45 @@
for (; vp; vp = vp->u.array) {
if (!(vp->flag&ISSET))
continue;
- XPput(wv, c == '!' ? shf_smprintf(Tf_lu,
+ XPput(wv, c == ord('!') ? shf_smprintf(Tf_lu,
arrayindex(vp)) :
str_val(vp));
}
if (XPsize(wv) == 0) {
xp->str = null;
- state = p[1] == '@' ? XNULLSUB : XSUB;
+ state = ord(p[1]) == ord('@') ? XNULLSUB : XSUB;
XPfree(wv);
} else {
XPput(wv, 0);
xp->u.strv = (const char **)XPptrv(wv);
xp->str = *xp->u.strv++;
/* ${foo[@]} */
- xp->split = tobool(p[1] == '@');
+ xp->split = tobool(ord(p[1]) == ord('@'));
state = XARG;
}
} else {
xp->var = global(sp);
xp->str = str_val(xp->var);
/* can't assign things like $! or $1 */
- if ((stype & 0x17F) == '=' && !*xp->str &&
+ if ((stype & STYPE_SINGLE) == ord('=') && !*xp->str &&
ctype(*sp, C_VAR1 | C_DIGIT))
return (-1);
state = XSUB;
}
- c = stype & 0x7F;
+ c = stype & STYPE_CHAR;
/* test the compiler's code generator */
- if (((stype < 0x100) && (ksh_issubop2(c) ||
- (((stype & 0x80) ? *xp->str == '\0' : xp->str == null) &&
+ if ((!(stype & STYPE_AT) && (ctype(c, C_SUB2) ||
+ (((stype & STYPE_DBL) ? *xp->str == '\0' : xp->str == null) &&
(state != XARG || (ifs0 || xp->split ?
(xp->u.strv[0] == NULL) : !hasnonempty(xp->u.strv))) ?
- c == '=' || c == '-' || c == '?' : c == '+'))) ||
- stype == (0x80 | '0') || stype == (0x100 | '#') ||
- stype == (0x100 | 'Q') || (stype & 0x7F) == '/')
+ ctype(c, C_EQUAL | C_MINUS | C_QUEST) : c == ord('+')))) ||
+ stype == (ord('0') | STYPE_DBL) || stype == (ord('#') | STYPE_AT) ||
+ stype == (ord('Q') | STYPE_AT) || (stype & STYPE_CHAR) == ord('/'))
/* expand word instead of variable value */
state = XBASE;
if (Flag(FNOUNSET) && xp->str == null && !zero_ok &&
- (ksh_issubop2(c) || (state != XBASE && c != '+')))
+ (ctype(c, C_SUB2) || (state != XBASE && c != ord('+'))))
errorf(Tf_parm, sp);
*stypep = stype;
*slenp = slen;
@@ -1408,7 +1415,7 @@
if (!herein(io, &name)) {
xp->str = name;
/* as $(…) requires, trim trailing newlines */
- name += strlen(name);
+ name = strnul(name);
while (name > xp->str && name[-1] == '\n')
--name;
*name = '\0';
@@ -1483,8 +1490,8 @@
char *end = strnul(str);
char *p, c;
- switch (how & 0xFF) {
- case '#':
+ switch (how & (STYPE_CHAR | STYPE_DBL)) {
+ case ord('#'):
/* shortest match at beginning */
for (p = str; p <= end; p += utf_ptradj(p)) {
c = *p; *p = '\0';
@@ -1496,7 +1503,7 @@
*p = c;
}
break;
- case '#'|0x80:
+ case ord('#') | STYPE_DBL:
/* longest match at beginning */
for (p = end; p >= str; p--) {
c = *p; *p = '\0';
@@ -1508,7 +1515,7 @@
*p = c;
}
break;
- case '%':
+ case ord('%'):
/* shortest match at end */
p = end;
while (p >= str) {
@@ -1516,7 +1523,7 @@
goto trimsub_match;
if (UTFMODE) {
char *op = p;
- while ((p-- > str) && ((*p & 0xC0) == 0x80))
+ while ((p-- > str) && ((rtt2asc(*p) & 0xC0) == 0x80))
;
if ((p < str) || (p + utf_ptradj(p) != op))
p = op - 1;
@@ -1524,7 +1531,7 @@
--p;
}
break;
- case '%'|0x80:
+ case ord('%') | STYPE_DBL:
/* longest match at end */
for (p = str; p <= end; p++)
if (gmatchx(p, pat, false)) {
@@ -1555,7 +1562,7 @@
XPput(*wp, debunk(cp, cp, strlen(cp) + 1));
else
qsort(XPptrv(*wp) + oldsize, XPsize(*wp) - oldsize,
- sizeof(void *), xstrcmp);
+ sizeof(void *), ascpstrcmp);
}
#define GF_NONE 0
@@ -1658,7 +1665,7 @@
*np++ = '\0';
} else {
odirsep = '\0'; /* keep gcc quiet */
- se = sp + strlen(sp);
+ se = strnul(sp);
}
@@ -1669,10 +1676,10 @@
* directory isn't readable - if no globbing is needed, only execute
* permission should be required (as per POSIX)).
*/
- if (!has_globbing(sp, se)) {
+ if (!has_globbing(sp)) {
XcheckN(*xs, xp, se - sp + 1);
debunk(xp, sp, Xnleft(*xs, xp));
- xp += strlen(xp);
+ xp = strnul(xp);
*xpp = xp;
globit(xs, xpp, np, wp, check);
} else {
@@ -1701,9 +1708,8 @@
XcheckN(*xs, xp, len);
memcpy(xp, name, len);
*xpp = xp + len - 1;
- globit(xs, xpp, np, wp,
- (check & GF_MARKDIR) | GF_GLOBBED
- | (np ? GF_EXCHECK : GF_NONE));
+ globit(xs, xpp, np, wp, (check & GF_MARKDIR) |
+ GF_GLOBBED | (np ? GF_EXCHECK : GF_NONE));
xp = Xstring(*xs, xp) + prefix_len;
}
closedir(dirp);
@@ -1728,7 +1734,7 @@
memmove(dp, sp, s - sp);
for (d = dp + (s - sp); *s && (d - dp < (ssize_t)dlen); s++)
if (!ISMAGIC(*s) || !(*++s & 0x80) ||
- !vstrchr("*+?@! ", *s & 0x7f))
+ !ctype(*s & 0x7F, C_PATMO | C_SPC))
*d++ = *s;
else {
/* extended pattern operators: *+?@! */
@@ -1857,7 +1863,7 @@
char *p = exp_start;
/* search for open brace */
- while ((p = strchr(p, MAGIC)) && p[1] != '{' /*}*/)
+ while ((p = strchr(p, MAGIC)) && ord(p[1]) != ord('{' /*}*/))
p += 2;
brace_start = p;
@@ -1868,9 +1874,9 @@
p += 2;
while (*p && count) {
if (ISMAGIC(*p++)) {
- if (*p == '{' /*}*/)
+ if (ord(*p) == ord('{' /*}*/))
++count;
- else if (*p == /*{*/ '}')
+ else if (ord(*p) == ord(/*{*/ '}'))
--count;
else if (*p == ',' && count == 1)
comma = p;
@@ -1902,9 +1908,9 @@
count = 1;
for (p = brace_start + 2; p != brace_end; p++) {
if (ISMAGIC(*p)) {
- if (*++p == '{' /*}*/)
+ if (ord(*++p) == ord('{' /*}*/))
++count;
- else if ((*p == /*{*/ '}' && --count == 0) ||
+ else if ((ord(*p) == ord(/*{*/ '}') && --count == 0) ||
(*p == ',' && count == 1)) {
char *news;
int l1, l2, l3;
diff --git a/src/exec.c b/src/exec.c
index 6307bce..56a42f6 100644
--- a/src/exec.c
+++ b/src/exec.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/exec.c,v 1.196 2017/04/12 16:46:21 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/exec.c,v 1.199 2017/08/07 21:16:31 tg Exp $");
#ifndef MKSH_DEFAULT_EXECSHELL
#define MKSH_DEFAULT_EXECSHELL MKSH_UNIXROOT "/bin/sh"
@@ -554,6 +554,9 @@
}
ap += builtin_opt.optind;
flags |= XEXEC;
+ /* POSuX demands ksh88-like behaviour here */
+ if (Flag(FPOSIX))
+ fcflags = FC_PATH;
} else if (tp->val.f == c_command) {
bool saw_p = false;
@@ -885,7 +888,9 @@
#ifndef MKSH_SMALL
if ((fd = binopen2(tp->str, O_RDONLY)) >= 0) {
unsigned char *cp;
+#ifndef MKSH_EBCDIC
unsigned short m;
+#endif
ssize_t n;
#if defined(__OS2__) && defined(MKSH_WITH_TEXTMODE)
@@ -905,7 +910,7 @@
(buf[2] == 0xBF)) ? 3 : 0);
/* scan for newline or NUL (end of buffer) */
- while (*cp && *cp != '\n')
+ while (!ctype(*cp, C_NL | C_NUL))
++cp;
/* if the shebang line is longer than MAXINTERP, bail out */
if (!*cp)
@@ -920,13 +925,13 @@
cp += 2;
#ifdef __OS2__
else if (!strncmp(cp, Textproc, 7) &&
- (cp[7] == ' ' || cp[7] == '\t'))
+ ctype(cp[7], C_BLANK))
cp += 8;
#endif
else
goto noshebang;
/* skip whitespace before shell name */
- while (*cp == ' ' || *cp == '\t')
+ while (ctype(*cp, C_BLANK))
++cp;
/* just whitespace on the line? */
if (*cp == '\0')
@@ -934,13 +939,13 @@
/* no, we actually found an interpreter name */
sh = (char *)cp;
/* look for end of shell/interpreter name */
- while (*cp != ' ' && *cp != '\t' && *cp != '\0')
+ while (!ctype(*cp, C_BLANK | C_NUL))
++cp;
/* any arguments? */
if (*cp) {
*cp++ = '\0';
/* skip spaces before arguments */
- while (*cp == ' ' || *cp == '\t')
+ while (ctype(*cp, C_BLANK))
++cp;
/* pass it all in ONE argument (historic reasons) */
if (*cp)
@@ -959,6 +964,7 @@
#endif
goto nomagic;
noshebang:
+#ifndef MKSH_EBCDIC
m = buf[0] << 8 | buf[1];
if (m == 0x7F45 && buf[2] == 'L' && buf[3] == 'F')
errorf("%s: not executable: %d-bit ELF file", tp->str,
@@ -977,6 +983,7 @@
buf[4] == 'Z') || (m == /* 7zip */ 0x377A) ||
(m == /* gzip */ 0x1F8B) || (m == /* .Z */ 0x1F9D))
errorf("%s: not executable: magic %04X", tp->str, m);
+#endif
#ifdef __OS2__
cp = _getext(tp->str);
if (cp && (!stricmp(cp, ".cmd") || !stricmp(cp, ".bat"))) {
@@ -1337,7 +1344,7 @@
while (sp != NULL) {
xp = Xstring(xs, xp);
if (!(p = cstrchr(sp, MKSH_PATHSEPC)))
- p = sp + strlen(sp);
+ p = strnul(sp);
if (p != sp) {
XcheckN(xs, xp, p - sp);
memcpy(xp, sp, p - sp);
diff --git a/src/expr.c b/src/expr.c
index 124dc17..12989d4 100644
--- a/src/expr.c
+++ b/src/expr.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/expr.c,v 1.93 2017/04/02 16:47:41 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/expr.c,v 1.100 2017/08/07 21:38:55 tg Exp $");
#define EXPRTOK_DEFNS
#include "exprtok.h"
@@ -558,9 +558,9 @@
/* skip whitespace */
skip_spaces:
- while ((c = *cp), ksh_isspace(c))
+ while (ctype(ord((c = *cp)), C_SPACE))
++cp;
- if (es->tokp == es->expression && c == '#') {
+ if (es->tokp == es->expression && c == ord('#')) {
/* expression begins with # */
/* switch to unsigned */
es->natural = true;
@@ -571,11 +571,11 @@
if (c == '\0')
es->tok = END;
- else if (ksh_isalphx(c)) {
+ else if (ctype(c, C_ALPHX)) {
do {
- c = *++cp;
- } while (ksh_isalnux(c));
- if (c == '[') {
+ c = ord(*++cp);
+ } while (ctype(c, C_ALNUX));
+ if (c == ord('[')) {
size_t len;
len = array_ref_len(cp);
@@ -617,9 +617,9 @@
tvar[c] = '\0';
goto process_tvar;
#endif
- } else if (ksh_isdigit(c)) {
- while (c != '_' && (ksh_isalnux(c) || c == '#'))
- c = *cp++;
+ } else if (ctype(c, C_DIGIT)) {
+ while (ctype(c, C_ALNUM | C_HASH))
+ c = ord(*cp++);
strndupx(tvar, es->tokp, --cp - es->tokp, ATEMP);
process_tvar:
es->val = tempvar("");
@@ -633,7 +633,7 @@
} else {
int i, n0;
- for (i = 0; (n0 = opname[i][0]); i++)
+ for (i = 0; (n0 = ord(opname[i][0])); i++)
if (c == n0 && strncmp(cp, opname[i],
(size_t)oplen[i]) == 0) {
es->tok = (enum token)i;
@@ -772,8 +772,7 @@
{
register size_t n;
- if (!UTFMODE ||
- *(const unsigned char *)(src) < 0xC2 ||
+ if (!UTFMODE || rtt2asc(*src) < 0xC2 ||
(n = utf_mbtowc(NULL, src)) == (size_t)-1)
n = 1;
return (n);
@@ -791,7 +790,7 @@
const unsigned char *s = (const unsigned char *)src;
unsigned int c, wc;
- if ((wc = *s++) < 0x80) {
+ if ((wc = ord(rtt2asc(*s++))) < 0x80) {
out:
if (dst != NULL)
*dst = wc;
@@ -805,7 +804,7 @@
if (wc < 0xE0) {
wc = (wc & 0x1F) << 6;
- if (((c = *s++) & 0xC0) != 0x80)
+ if (((c = ord(rtt2asc(*s++))) & 0xC0) != 0x80)
goto ilseq;
wc |= c & 0x3F;
goto out;
@@ -813,11 +812,11 @@
wc = (wc & 0x0F) << 12;
- if (((c = *s++) & 0xC0) != 0x80)
+ if (((c = ord(rtt2asc(*s++))) & 0xC0) != 0x80)
goto ilseq;
wc |= (c & 0x3F) << 6;
- if (((c = *s++) & 0xC0) != 0x80)
+ if (((c = ord(rtt2asc(*s++))) & 0xC0) != 0x80)
goto ilseq;
wc |= c & 0x3F;
@@ -834,18 +833,18 @@
unsigned char *d;
if (wc < 0x80) {
- *dst = wc;
+ *dst = asc2rtt(wc);
return (1);
}
d = (unsigned char *)dst;
if (wc < 0x0800)
- *d++ = (wc >> 6) | 0xC0;
+ *d++ = asc2rtt((wc >> 6) | 0xC0);
else {
- *d++ = ((wc = wc > 0xFFFD ? 0xFFFD : wc) >> 12) | 0xE0;
- *d++ = ((wc >> 6) & 0x3F) | 0x80;
+ *d++ = asc2rtt(((wc = wc > 0xFFFD ? 0xFFFD : wc) >> 12) | 0xE0);
+ *d++ = asc2rtt(((wc >> 6) & 0x3F) | 0x80);
}
- *d++ = (wc & 0x3F) | 0x80;
+ *d++ = asc2rtt((wc & 0x3F) | 0x80);
return ((char *)d - dst);
}
@@ -873,7 +872,7 @@
}
#ifndef MIRBSD_BOOTFLOPPY
-/* From: X11/xc/programs/xterm/wcwidth.c,v 1.9 */
+/* From: X11/xc/programs/xterm/wcwidth.c,v 1.10 */
struct mb_ucsrange {
unsigned short beg;
@@ -884,8 +883,8 @@
unsigned int val) MKSH_A_PURE;
/*
- * Generated from the Unicode Character Database, Version 9.0.0, by
- * MirOS: contrib/code/Snippets/eawparse,v 1.3 2014/11/16 12:16:24 tg Exp $
+ * Generated from the Unicode Character Database, Version 10.0.0, by
+ * MirOS: contrib/code/Snippets/eawparse,v 1.10 2017/07/12 22:47:26 tg Exp $
*/
static const struct mb_ucsrange mb_ucs_combining[] = {
@@ -941,6 +940,7 @@
{ 0x0AC7, 0x0AC8 },
{ 0x0ACD, 0x0ACD },
{ 0x0AE2, 0x0AE3 },
+ { 0x0AFA, 0x0AFF },
{ 0x0B01, 0x0B01 },
{ 0x0B3C, 0x0B3C },
{ 0x0B3F, 0x0B3F },
@@ -963,7 +963,8 @@
{ 0x0CC6, 0x0CC6 },
{ 0x0CCC, 0x0CCD },
{ 0x0CE2, 0x0CE3 },
- { 0x0D01, 0x0D01 },
+ { 0x0D00, 0x0D01 },
+ { 0x0D3B, 0x0D3C },
{ 0x0D41, 0x0D44 },
{ 0x0D4D, 0x0D4D },
{ 0x0D62, 0x0D63 },
@@ -1048,7 +1049,7 @@
{ 0x1CED, 0x1CED },
{ 0x1CF4, 0x1CF4 },
{ 0x1CF8, 0x1CF9 },
- { 0x1DC0, 0x1DF5 },
+ { 0x1DC0, 0x1DF9 },
{ 0x1DFB, 0x1DFF },
{ 0x200B, 0x200F },
{ 0x202A, 0x202E },
@@ -1136,14 +1137,16 @@
{ 0x2B1B, 0x2B1C },
{ 0x2B50, 0x2B50 },
{ 0x2B55, 0x2B55 },
- { 0x2E80, 0x303E },
- { 0x3040, 0xA4CF },
+ { 0x2E80, 0x3029 },
+ { 0x302E, 0x303E },
+ { 0x3040, 0x3098 },
+ { 0x309B, 0xA4CF },
{ 0xA960, 0xA97F },
{ 0xAC00, 0xD7A3 },
{ 0xF900, 0xFAFF },
{ 0xFE10, 0xFE19 },
{ 0xFE30, 0xFE6F },
- { 0xFF00, 0xFF60 },
+ { 0xFF01, 0xFF60 },
{ 0xFFE0, 0xFFE6 }
};
diff --git a/src/funcs.c b/src/funcs.c
index 930462d..38e66f8 100644
--- a/src/funcs.c
+++ b/src/funcs.c
@@ -38,7 +38,7 @@
#endif
#endif
-__RCSID("$MirOS: src/bin/mksh/funcs.c,v 1.340 2017/04/12 17:46:29 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/funcs.c,v 1.350 2017/05/05 22:53:28 tg Exp $");
#if HAVE_KILLPG
/*
@@ -751,11 +751,15 @@
bool
valid_alias_name(const char *cp)
{
+ if (ord(*cp) == ord('-'))
+ return (false);
+ if (ord(cp[0]) == ord('[') && ord(cp[1]) == ord('[') && !cp[2])
+ return (false);
while (*cp)
- if (!ksh_isalias(*cp))
- return (false);
- else
+ if (ctype(*cp, C_ALIAS))
++cp;
+ else
+ return (false);
return (true);
}
@@ -764,7 +768,7 @@
{
struct table *t = &aliases;
int rv = 0, prefix = 0;
- bool rflag = false, tflag, Uflag = false, pflag = false;
+ bool rflag = false, tflag, Uflag = false, pflag = false, chkalias;
uint32_t xflag = 0;
int optc;
@@ -809,12 +813,13 @@
wp += builtin_opt.optind;
if (!(builtin_opt.info & GI_MINUSMINUS) && *wp &&
- (wp[0][0] == '-' || wp[0][0] == '+') && wp[0][1] == '\0') {
+ ctype(wp[0][0], C_MINUS | C_PLUS) && wp[0][1] == '\0') {
prefix = wp[0][0];
wp++;
}
tflag = t == &taliases;
+ chkalias = t == &aliases;
/* "hash -r" means reset all the tracked aliases.. */
if (rflag) {
@@ -857,7 +862,7 @@
strndupx(xalias, alias, val++ - alias, ATEMP);
alias = xalias;
}
- if (!valid_alias_name(alias) || *alias == '-') {
+ if (chkalias && !valid_alias_name(alias)) {
bi_errorf(Tinvname, alias, Talias);
afree(xalias, ATEMP);
return (1);
@@ -1072,8 +1077,7 @@
int i, n, rv, sig;
/* assume old style options if -digits or -UPPERCASE */
- if ((p = wp[1]) && *p == '-' && (ksh_isdigit(p[1]) ||
- ksh_isupper(p[1]))) {
+ if ((p = wp[1]) && *p == '-' && ctype(p[1], C_DIGIT | C_UPPER)) {
if (!(t = gettrap(p + 1, false, false))) {
bi_errorf(Tbad_sig_s, p + 1);
return (1);
@@ -1422,9 +1426,9 @@
} else {
mode_t new_umask;
- if (ksh_isdigit(*cp)) {
+ if (ctype(*cp, C_DIGIT)) {
new_umask = 0;
- while (*cp >= ord('0') && *cp <= ord('7')) {
+ while (ctype(*cp, C_OCTAL)) {
new_umask = new_umask * 8 + ksh_numdig(*cp);
++cp;
}
@@ -1462,7 +1466,7 @@
if (!positions)
/* default is a */
positions = 0111;
- if (!vstrchr("=+-", op = *cp))
+ if (!ctype((op = *cp), C_EQUAL | C_MINUS | C_PLUS))
break;
cp++;
new_val = 0;
@@ -1503,7 +1507,7 @@
if (*cp == ',') {
positions = 0;
cp++;
- } else if (!vstrchr("=+-", *cp))
+ } else if (!ctype(*cp, C_EQUAL | C_MINUS | C_PLUS))
break;
}
if (*cp) {
@@ -1585,7 +1589,7 @@
return (rv);
}
-static char REPLY[] = "REPLY";
+static const char REPLY[] = "REPLY";
int
c_read(const char **wp)
{
@@ -2300,8 +2304,9 @@
size_t n;
n = strlen(id);
- if (n > 3 && id[n-3] == '[' && id[n-2] == '*' &&
- id[n-1] == ']') {
+ if (n > 3 && ord(id[n - 3]) == ord('[') &&
+ ord(id[n - 2]) == ord('*') &&
+ ord(id[n - 1]) == ord(']')) {
strndupx(cp, id, n - 3, ATEMP);
id = cp;
optc = 3;
@@ -3350,7 +3355,7 @@
* If this causes problems, will have to add parameter to
* evaluate() to control if unset params are 0 or an error.
*/
- if (!rval && !ksh_isdigit(v[0])) {
+ if (!rval && !ctype(v[0], C_DIGIT)) {
bi_errorf("invalid %s limit: %s", l->name, v);
return (1);
}
diff --git a/src/histrap.c b/src/histrap.c
index 26dd521..6b9396e 100644
--- a/src/histrap.c
+++ b/src/histrap.c
@@ -3,7 +3,7 @@
/*-
* Copyright (c) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010,
- * 2011, 2012, 2014, 2015, 2016
+ * 2011, 2012, 2014, 2015, 2016, 2017
* mirabilos <m@mirbsd.org>
*
* Provided that these terms and disclaimer and all copyright notices
@@ -27,7 +27,7 @@
#include <sys/file.h>
#endif
-__RCSID("$MirOS: src/bin/mksh/histrap.c,v 1.160 2017/04/08 01:07:16 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/histrap.c,v 1.166 2017/08/07 23:25:09 tg Exp $");
Trap sigtraps[ksh_NSIG + 1];
static struct sigaction Sigact_ign;
@@ -629,7 +629,7 @@
if (svmode == HIST_FLUSH)
return;
- ccp = cmd + strlen(cmd);
+ ccp = strnul(cmd);
while (ccp > cmd && ccp[-1] == '\n')
--ccp;
strndupx(c, cmd, ccp - cmd, APERM);
@@ -714,26 +714,66 @@
#if HAVE_PERSISTENT_HISTORY
static const unsigned char sprinkle[2] = { HMAGIC1, HMAGIC2 };
-#endif
-void
-hist_init(Source *s)
+static int
+hist_persist_back(int srcfd)
{
-#if HAVE_PERSISTENT_HISTORY
+ off_t tot, mis;
+ ssize_t n, w;
+ char *buf, *cp;
+ int rv = 0;
+#define MKSH_HS_BUFSIZ 4096
+
+ if ((tot = lseek(srcfd, (off_t)0, SEEK_END)) < 0 ||
+ lseek(srcfd, (off_t)0, SEEK_SET) < 0 ||
+ lseek(histfd, (off_t)0, SEEK_SET) < 0)
+ return (1);
+
+ if ((buf = malloc_osfunc(MKSH_HS_BUFSIZ)) == NULL)
+ return (1);
+
+ mis = tot;
+ while (mis > 0) {
+ if ((n = blocking_read(srcfd, (cp = buf),
+ MKSH_HS_BUFSIZ)) == -1) {
+ if (errno == EINTR) {
+ intrcheck();
+ continue;
+ }
+ goto copy_error;
+ }
+ mis -= n;
+ while (n) {
+ if (intrsig)
+ goto has_intrsig;
+ if ((w = write(histfd, cp, n)) != -1) {
+ n -= w;
+ cp += w;
+ continue;
+ }
+ if (errno == EINTR) {
+ has_intrsig:
+ intrcheck();
+ continue;
+ }
+ goto copy_error;
+ }
+ }
+ if (ftruncate(histfd, tot)) {
+ copy_error:
+ rv = 1;
+ }
+ free_osfunc(buf);
+ return (rv);
+}
+
+static void
+hist_persist_init(void)
+{
unsigned char *base;
int lines, fd;
- enum { hist_init_first, hist_init_retry, hist_init_restore } hs;
-#endif
+ enum { hist_init_first, hist_init_retry, hist_use_it } hs;
- histsave(NULL, NULL, HIST_DISCARD, true);
-
- if (Flag(FTALKING) == 0)
- return;
-
- hstarted = true;
- hist_source = s;
-
-#if HAVE_PERSISTENT_HISTORY
if (((hname = str_val(global("HISTFILE"))) == NULL) || !*hname) {
hname = NULL;
return;
@@ -745,17 +785,16 @@
/* we have a file and are interactive */
if ((fd = binopen3(hname, O_RDWR | O_CREAT | O_APPEND, 0600)) < 0)
return;
-
- histfd = savefd(fd);
+ if ((histfd = savefd(fd)) < 0)
+ return;
if (histfd != fd)
close(fd);
mksh_lockfd(histfd);
histfsize = lseek(histfd, (off_t)0, SEEK_END);
- if (histfsize > MKSH_MAXHISTFSIZE || hs == hist_init_restore) {
+ if (histfsize > MKSH_MAXHISTFSIZE) {
/* we ignore too large files but still append to them */
- /* we also don't need to re-read after truncation */
goto hist_init_tail;
} else if (histfsize > 2) {
/* we have some data, check its validity */
@@ -781,6 +820,7 @@
if ((fd = binopen3(nhname, O_RDWR | O_CREAT | O_TRUNC |
O_EXCL, 0600)) < 0) {
/* just don't truncate then, meh. */
+ hs = hist_use_it;
goto hist_trunc_dont;
}
if (fstat(histfd, &sb) >= 0 &&
@@ -795,28 +835,26 @@
hp = history;
while (hp < histptr) {
if (!writehistline(fd,
- s->line - (histptr - hp), *hp))
+ hist_source->line - (histptr - hp), *hp))
goto hist_trunc_abort;
++hp;
}
- /* now unlock, close both, rename, rinse, repeat */
+ /* now transfer back */
+ if (!hist_persist_back(fd)) {
+ /* success! */
+ hs = hist_use_it;
+ }
+ hist_trunc_abort:
+ /* remove temporary file */
close(fd);
fd = -1;
- hist_finish();
- if (rename(nhname, hname) < 0) {
- hist_trunc_abort:
- if (fd != -1)
- close(fd);
- unlink(nhname);
- if (fd != -1)
- goto hist_trunc_dont;
- /* darn! restore histfd and pray */
- }
- hs = hist_init_restore;
+ unlink(nhname);
+ /* use whatever is in the file now */
hist_trunc_dont:
afree(nhname, ATEMP);
- if (hs == hist_init_restore)
- goto retry;
+ if (hs == hist_use_it)
+ goto hist_trunc_done;
+ goto hist_init_fail;
}
} else if (histfsize != 0) {
/* negative or too small... */
@@ -840,9 +878,26 @@
return;
}
}
+ hist_trunc_done:
histfsize = lseek(histfd, (off_t)0, SEEK_END);
hist_init_tail:
mksh_unlkfd(histfd);
+}
+#endif
+
+void
+hist_init(Source *s)
+{
+ histsave(NULL, NULL, HIST_DISCARD, true);
+
+ if (Flag(FTALKING) == 0)
+ return;
+
+ hstarted = true;
+ hist_source = s;
+
+#if HAVE_PERSISTENT_HISTORY
+ hist_persist_init();
#endif
}
@@ -909,10 +964,11 @@
mksh_lockfd(histfd);
sizenow = lseek(histfd, (off_t)0, SEEK_END);
if (sizenow < histfsize) {
- /* the file has shrunk; give up */
- goto bad;
- }
- if (
+ /* the file has shrunk; trust it just appending the new data */
+ /* well, for now, anyway… since mksh strdups all into memory */
+ /* we can use a nicer approach some time later… */
+ ;
+ } else if (
/* ignore changes when the file is too large */
sizenow <= MKSH_MAXHISTFSIZE
&&
@@ -1114,7 +1170,7 @@
/* signal number (1..ksh_NSIG) or 0? */
- if (ksh_isdigit(*cs))
+ if (ctype(*cs, C_DIGIT))
return ((getn(cs, &i) && 0 <= i && i < ksh_NSIG) ?
(&sigtraps[i]) : NULL);
diff --git a/src/jobs.c b/src/jobs.c
index 0366004..4df98b7 100644
--- a/src/jobs.c
+++ b/src/jobs.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/jobs.c,v 1.121 2016/07/25 00:04:44 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/jobs.c,v 1.124 2017/08/08 14:30:10 tg Exp $");
#if HAVE_KILLPG
#define mksh_killpg killpg
@@ -39,14 +39,27 @@
#define PSTOPPED 3
typedef struct proc Proc;
-struct proc {
- Proc *next; /* next process in pipeline (if any) */
- pid_t pid; /* process id */
+/* to take alignment into consideration */
+struct proc_dummy {
+ Proc *next;
+ pid_t pid;
int state;
- int status; /* wait status */
+ int status;
+ char command[128];
+};
+/* real structure */
+struct proc {
+ /* next process in pipeline (if any) */
+ Proc *next;
+ /* process id of this Unix process in the job */
+ pid_t pid;
+ /* one of the four P… above */
+ int state;
+ /* wait status */
+ int status;
/* process command string from vistree */
- char command[256 - (ALLOC_OVERHEAD + sizeof(Proc *) +
- sizeof(pid_t) + 2 * sizeof(int))];
+ char command[256 - (ALLOC_OVERHEAD +
+ offsetof(struct proc_dummy, command[0]))];
};
/* Notify/print flag - j_print() argument */
@@ -1009,8 +1022,14 @@
}
for (j = job_list; j; j = tmp) {
tmp = j->next;
- if (j->flags & JF_REMOVE)
- remove_job(j, "notify");
+ if (j->flags & JF_REMOVE) {
+ if (j == async_job || (j->flags & JF_KNOWN)) {
+ j->flags = (j->flags & ~JF_REMOVE) | JF_ZOMBIE;
+ j->job = -1;
+ nzombie++;
+ } else
+ remove_job(j, "notify");
+ }
}
shf_flush(shl_out);
#ifndef MKSH_NOPROSPECTOFWORK
@@ -1651,7 +1670,7 @@
size_t len;
int job = 0;
- if (ksh_isdigit(*cp) && getn(cp, &job)) {
+ if (ctype(*cp, C_DIGIT) && getn(cp, &job)) {
/* Look for last_proc->pid (what $! returns) first... */
for (j = job_list; j != NULL; j = j->next)
if (j->last_proc && j->last_proc->pid == job)
diff --git a/src/lex.c b/src/lex.c
index 78c2ee7..f450221 100644
--- a/src/lex.c
+++ b/src/lex.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/lex.c,v 1.234 2017/04/06 01:59:55 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/lex.c,v 1.239 2017/05/05 22:53:29 tg Exp $");
/*
* states while lexing word
@@ -131,7 +131,7 @@
}
#if defined(MKSH_SMALL) && !defined(MKSH_SMALL_BUT_FAST)
-#define getsc getsc_i
+#define getsc() ord(getsc_i())
#else
static int getsc_r(int);
@@ -141,7 +141,7 @@
o_getsc_r(c);
}
-#define getsc() getsc_r(o_getsc())
+#define getsc() ord(getsc_r(o_getsc()))
#endif
#define STATE_BSIZE 8
@@ -220,11 +220,11 @@
} else {
/* normal lexing */
state = (cf & HEREDELIM) ? SHEREDELIM : SBASE;
- while ((c = getsc()) == ' ' || c == '\t')
+ while (ctype((c = getsc()), C_BLANK))
;
if (c == '#') {
ignore_backslash_newline++;
- while ((c = getsc()) != '\0' && c != '\n')
+ while (!ctype((c = getsc()), C_NUL | C_LF))
;
ignore_backslash_newline--;
}
@@ -245,30 +245,30 @@
while (!((c = getsc()) == 0 ||
((state == SBASE || state == SHEREDELIM) && ctype(c, C_LEX1)))) {
if (state == SBASE &&
- subshell_nesting_type == /*{*/ '}' &&
- c == /*{*/ '}')
+ subshell_nesting_type == ord(/*{*/ '}') &&
+ c == ord(/*{*/ '}'))
/* possibly end ${ :;} */
break;
Xcheck(ws, wp);
switch (state) {
case SADELIM:
- if (c == '(')
+ if (c == ord('('))
statep->nparen++;
- else if (c == ')')
+ else if (c == ord(')'))
statep->nparen--;
- else if (statep->nparen == 0 && (c == /*{*/ '}' ||
+ else if (statep->nparen == 0 && (c == ord(/*{*/ '}') ||
c == (int)statep->ls_adelim.delimiter)) {
*wp++ = ADELIM;
*wp++ = c;
- if (c == /*{*/ '}' || --statep->ls_adelim.num == 0)
+ if (c == ord(/*{*/ '}') || --statep->ls_adelim.num == 0)
POP_STATE();
- if (c == /*{*/ '}')
+ if (c == ord(/*{*/ '}'))
POP_STATE();
break;
}
/* FALLTHROUGH */
case SBASE:
- if (c == '[' && (cf & CMDASN)) {
+ if (c == ord('[') && (cf & CMDASN)) {
/* temporary */
*wp = EOS;
if (is_wdvarname(Xstring(ws, wp), false)) {
@@ -301,10 +301,9 @@
}
/* FALLTHROUGH */
Sbase1: /* includes *(...|...) pattern (*+?@!) */
- if (c == '*' || c == '@' || c == '+' || c == '?' ||
- c == '!') {
+ if (ctype(c, C_PATMO)) {
c2 = getsc();
- if (c2 == '(' /*)*/ ) {
+ if (c2 == ord('(' /*)*/)) {
*wp++ = OPAT;
*wp++ = c;
PUSH_STATE(SPATTERN);
@@ -315,7 +314,7 @@
/* FALLTHROUGH */
Sbase2: /* doesn't include *(...|...) pattern (*+?@!) */
switch (c) {
- case '\\':
+ case ord('\\'):
getsc_qchar:
if ((c = getsc())) {
/* trailing \ is lost */
@@ -323,7 +322,7 @@
*wp++ = c;
}
break;
- case '\'':
+ case ord('\''):
open_ssquote_unless_heredoc:
if ((cf & HEREDOC))
goto store_char;
@@ -331,12 +330,12 @@
ignore_backslash_newline++;
PUSH_STATE(SSQUOTE);
break;
- case '"':
+ case ord('"'):
open_sdquote:
*wp++ = OQUOTE;
PUSH_STATE(SDQUOTE);
break;
- case '$':
+ case ord('$'):
/*
* processing of dollar sign belongs into
* Subst, except for those which can open
@@ -345,9 +344,9 @@
subst_dollar_ex:
c = getsc();
switch (c) {
- case '"':
+ case ord('"'):
goto open_sdquote;
- case '\'':
+ case ord('\''):
goto open_sequote;
default:
goto SubstS;
@@ -359,15 +358,16 @@
Subst:
switch (c) {
- case '\\':
+ case ord('\\'):
c = getsc();
switch (c) {
- case '"':
+ case ord('"'):
if ((cf & HEREDOC))
goto heredocquote;
/* FALLTHROUGH */
- case '\\':
- case '$': case '`':
+ case ord('\\'):
+ case ord('$'):
+ case ord('`'):
store_qchar:
*wp++ = QCHAR;
*wp++ = c;
@@ -385,12 +385,12 @@
break;
}
break;
- case '$':
+ case ord('$'):
c = getsc();
SubstS:
- if (c == '(') /*)*/ {
+ if (c == ord('(' /*)*/)) {
c = getsc();
- if (c == '(') /*)*/ {
+ if (c == ord('(' /*)*/)) {
*wp++ = EXPRSUB;
PUSH_SRETRACE(SASPAREN);
statep->nparen = 2;
@@ -407,8 +407,8 @@
memcpy(wp, sp, cz);
wp += cz;
}
- } else if (c == '{') /*}*/ {
- if ((c = getsc()) == '|') {
+ } else if (c == ord('{' /*}*/)) {
+ if ((c = getsc()) == ord('|')) {
/*
* non-subenvironment
* value substitution
@@ -425,15 +425,15 @@
}
ungetsc(c);
*wp++ = OSUBST;
- *wp++ = '{'; /*}*/
+ *wp++ = '{' /*}*/;
wp = get_brace_var(&ws, wp);
c = getsc();
/* allow :# and :% (ksh88 compat) */
- if (c == ':') {
+ if (c == ord(':')) {
*wp++ = CHAR;
*wp++ = c;
c = getsc();
- if (c == ':') {
+ if (c == ord(':')) {
*wp++ = CHAR;
*wp++ = '0';
*wp++ = ADELIM;
@@ -444,10 +444,9 @@
statep->ls_adelim.num = 1;
statep->nparen = 0;
break;
- } else if (ksh_isdigit(c) ||
- c == '('/*)*/ || c == ' ' ||
+ } else if (ctype(c, C_DIGIT | C_DOLAR | C_SPC) ||
/*XXX what else? */
- c == '$') {
+ c == '(' /*)*/) {
/* substring subst. */
if (c != ' ') {
*wp++ = CHAR;
@@ -466,7 +465,7 @@
parse_adelim_slash:
*wp++ = CHAR;
*wp++ = c;
- if ((c = getsc()) == '/') {
+ if ((c = getsc()) == ord('/')) {
*wp++ = c2;
*wp++ = c;
} else
@@ -480,7 +479,7 @@
} else if (c == '@') {
c2 = getsc();
ungetsc(c2);
- if (c2 == '/') {
+ if (c2 == ord('/')) {
c2 = CHAR;
goto parse_adelim_slash;
}
@@ -489,7 +488,7 @@
* If this is a trim operation,
* treat (,|,) specially in STBRACE.
*/
- if (ksh_issubop2(c)) {
+ if (ctype(c, C_SUB2)) {
ungetsc(c);
if (Flag(FSH))
PUSH_STATE(STBRACEBOURNE);
@@ -503,14 +502,14 @@
else
PUSH_STATE(SBRACE);
}
- } else if (ksh_isalphx(c)) {
+ } else if (ctype(c, C_ALPHX)) {
*wp++ = OSUBST;
*wp++ = 'X';
do {
Xcheck(ws, wp);
*wp++ = c;
c = getsc();
- } while (ksh_isalnux(c));
+ } while (ctype(c, C_ALNUX));
*wp++ = '\0';
*wp++ = CSUBST;
*wp++ = 'X';
@@ -529,7 +528,7 @@
ungetsc(c);
}
break;
- case '`':
+ case ord('`'):
subst_gravis:
PUSH_STATE(SBQUOTE);
*wp++ = COMASUB;
@@ -573,11 +572,11 @@
break;
case SEQUOTE:
- if (c == '\'') {
+ if (c == ord('\'')) {
POP_STATE();
*wp++ = CQUOTE;
ignore_backslash_newline--;
- } else if (c == '\\') {
+ } else if (c == ord('\\')) {
if ((c2 = unbksl(true, getsc_i, ungetsc)) == -1)
c2 = getsc();
if (c2 == 0)
@@ -605,7 +604,7 @@
break;
case SSQUOTE:
- if (c == '\'') {
+ if (c == ord('\'')) {
POP_STATE();
if ((cf & HEREDOC) || state == SQBRACE)
goto store_char;
@@ -618,7 +617,7 @@
break;
case SDQUOTE:
- if (c == '"') {
+ if (c == ord('"')) {
POP_STATE();
*wp++ = CQUOTE;
} else
@@ -627,15 +626,15 @@
/* $(( ... )) */
case SASPAREN:
- if (c == '(')
+ if (c == ord('('))
statep->nparen++;
- else if (c == ')') {
+ else if (c == ord(')')) {
statep->nparen--;
if (statep->nparen == 1) {
/* end of EXPRSUB */
POP_SRETRACE();
- if ((c2 = getsc()) == /*(*/ ')') {
+ if ((c2 = getsc()) == ord(/*(*/ ')')) {
cz = strlen(sp) - 2;
XcheckN(ws, wp, cz);
memcpy(wp, sp + 1, cz);
@@ -667,7 +666,7 @@
goto Sbase2;
case SQBRACE:
- if (c == '\\') {
+ if (c == ord('\\')) {
/*
* perform POSIX "quote removal" if the back-
* slash is "special", i.e. same cases as the
@@ -676,26 +675,26 @@
* write QCHAR+c, otherwise CHAR+\+CHAR+c are
* emitted (in heredocquote:)
*/
- if ((c = getsc()) == '"' || c == '\\' ||
- c == '$' || c == '`' || c == /*{*/'}')
+ if ((c = getsc()) == ord('"') || c == ord('\\') ||
+ ctype(c, C_DOLAR | C_GRAVE) || c == ord(/*{*/ '}'))
goto store_qchar;
goto heredocquote;
}
goto common_SQBRACE;
case SBRACE:
- if (c == '\'')
+ if (c == ord('\''))
goto open_ssquote_unless_heredoc;
- else if (c == '\\')
+ else if (c == ord('\\'))
goto getsc_qchar;
common_SQBRACE:
- if (c == '"')
+ if (c == ord('"'))
goto open_sdquote;
- else if (c == '$')
+ else if (c == ord('$'))
goto subst_dollar_ex;
- else if (c == '`')
+ else if (c == ord('`'))
goto subst_gravis;
- else if (c != /*{*/ '}')
+ else if (c != ord(/*{*/ '}'))
goto store_char;
POP_STATE();
*wp++ = CSUBST;
@@ -704,16 +703,16 @@
/* Same as SBASE, except (,|,) treated specially */
case STBRACEKORN:
- if (c == '|')
+ if (c == ord('|'))
*wp++ = SPAT;
- else if (c == '(') {
+ else if (c == ord('(')) {
*wp++ = OPAT;
/* simile for @ */
*wp++ = ' ';
PUSH_STATE(SPATTERN);
} else /* FALLTHROUGH */
case STBRACEBOURNE:
- if (c == /*{*/ '}') {
+ if (c == ord(/*{*/ '}')) {
POP_STATE();
*wp++ = CSUBST;
*wp++ = /*{*/ '}';
@@ -722,20 +721,20 @@
break;
case SBQUOTE:
- if (c == '`') {
+ if (c == ord('`')) {
*wp++ = 0;
POP_STATE();
- } else if (c == '\\') {
+ } else if (c == ord('\\')) {
switch (c = getsc()) {
case 0:
/* trailing \ is lost */
break;
- case '$':
- case '`':
- case '\\':
+ case ord('$'):
+ case ord('`'):
+ case ord('\\'):
*wp++ = c;
break;
- case '"':
+ case ord('"'):
if (statep->ls_bool) {
*wp++ = c;
break;
@@ -756,10 +755,10 @@
/* LETEXPR: (( ... )) */
case SLETPAREN:
- if (c == /*(*/ ')') {
+ if (c == ord(/*(*/ ')')) {
if (statep->nparen > 0)
--statep->nparen;
- else if ((c2 = getsc()) == /*(*/ ')') {
+ else if ((c2 = getsc()) == ord(/*(*/ ')')) {
c = 0;
*wp++ = CQUOTE;
goto Done;
@@ -780,10 +779,10 @@
s->start = s->str = s->u.freeme = dp;
s->next = source;
source = s;
- ungetsc('('/*)*/);
- return ('('/*)*/);
+ ungetsc('(' /*)*/);
+ return (ord('(' /*)*/));
}
- } else if (c == '(')
+ } else if (c == ord('('))
/*
* parentheses inside quotes and
* backslashes are lost, but AT&T ksh
@@ -799,26 +798,26 @@
* $ and `...` are not to be treated specially
*/
switch (c) {
- case '\\':
+ case ord('\\'):
if ((c = getsc())) {
/* trailing \ is lost */
*wp++ = QCHAR;
*wp++ = c;
}
break;
- case '\'':
+ case ord('\''):
goto open_ssquote_unless_heredoc;
- case '$':
- if ((c2 = getsc()) == '\'') {
+ case ord('$'):
+ if ((c2 = getsc()) == ord('\'')) {
open_sequote:
*wp++ = OQUOTE;
ignore_backslash_newline++;
PUSH_STATE(SEQUOTE);
statep->ls_bool = false;
break;
- } else if (c2 == '"') {
+ } else if (c2 == ord('"')) {
/* FALLTHROUGH */
- case '"':
+ case ord('"'):
PUSH_SRETRACE(SHEREDQUOTE);
break;
}
@@ -832,7 +831,7 @@
/* " in << or <<- delimiter */
case SHEREDQUOTE:
- if (c != '"')
+ if (c != ord('"'))
goto Subst;
POP_SRETRACE();
dp = strnul(sp) - 1;
@@ -845,10 +844,10 @@
while ((c = *dp++)) {
if (c == '\\') {
switch ((c = *dp++)) {
- case '\\':
- case '"':
- case '$':
- case '`':
+ case ord('\\'):
+ case ord('"'):
+ case ord('$'):
+ case ord('`'):
break;
default:
*wp++ = CHAR;
@@ -866,12 +865,12 @@
/* in *(...|...) pattern (*+?@!) */
case SPATTERN:
- if (c == /*(*/ ')') {
+ if (c == ord(/*(*/ ')')) {
*wp++ = CPAT;
POP_STATE();
- } else if (c == '|') {
+ } else if (c == ord('|')) {
*wp++ = SPAT;
- } else if (c == '(') {
+ } else if (c == ord('(')) {
*wp++ = OPAT;
/* simile for @ */
*wp++ = ' ';
@@ -894,14 +893,14 @@
dp = Xstring(ws, wp);
if (state == SBASE && (
(c == '&' && !Flag(FSH) && !Flag(FPOSIX)) ||
- c == '<' || c == '>') && ((c2 = Xlength(ws, wp)) == 0 ||
- (c2 == 2 && dp[0] == CHAR && ksh_isdigit(dp[1])))) {
+ ctype(c, C_ANGLE)) && ((c2 = Xlength(ws, wp)) == 0 ||
+ (c2 == 2 && dp[0] == CHAR && ctype(dp[1], C_DIGIT)))) {
struct ioword *iop = alloc(sizeof(struct ioword), ATEMP);
iop->unit = c2 == 2 ? ksh_numdig(dp[1]) : c == '<' ? 0 : 1;
if (c == '&') {
- if ((c2 = getsc()) != '>') {
+ if ((c2 = getsc()) != ord('>')) {
ungetsc(c2);
goto no_iop;
}
@@ -912,22 +911,22 @@
c2 = getsc();
/* <<, >>, <> are ok, >< is not */
- if (c == c2 || (c == '<' && c2 == '>')) {
+ if (c == c2 || (c == ord('<') && c2 == ord('>'))) {
iop->ioflag |= c == c2 ?
- (c == '>' ? IOCAT : IOHERE) : IORDWR;
+ (c == ord('>') ? IOCAT : IOHERE) : IORDWR;
if (iop->ioflag == IOHERE) {
- if ((c2 = getsc()) == '-')
+ if ((c2 = getsc()) == ord('-'))
iop->ioflag |= IOSKIP;
- else if (c2 == '<')
+ else if (c2 == ord('<'))
iop->ioflag |= IOHERESTR;
else
ungetsc(c2);
}
- } else if (c2 == '&')
- iop->ioflag |= IODUP | (c == '<' ? IORDUP : 0);
+ } else if (c2 == ord('&'))
+ iop->ioflag |= IODUP | (c == ord('<') ? IORDUP : 0);
else {
- iop->ioflag |= c == '>' ? IOWRITE : IOREAD;
- if (c == '>' && c2 == '|')
+ iop->ioflag |= c == ord('>') ? IOWRITE : IOREAD;
+ if (c == ord('>') && c2 == ord('|'))
iop->ioflag |= IOCLOB;
else
ungetsc(c2);
@@ -948,29 +947,30 @@
/* free word */
Xfree(ws, wp);
/* no word, process LEX1 character */
- if ((c == '|') || (c == '&') || (c == ';') || (c == '('/*)*/)) {
+ if ((c == ord('|')) || (c == ord('&')) || (c == ord(';')) ||
+ (c == ord('(' /*)*/))) {
if ((c2 = getsc()) == c)
- c = (c == ';') ? BREAK :
- (c == '|') ? LOGOR :
- (c == '&') ? LOGAND :
- /* c == '(' ) */ MDPAREN;
- else if (c == '|' && c2 == '&')
+ c = (c == ord(';')) ? BREAK :
+ (c == ord('|')) ? LOGOR :
+ (c == ord('&')) ? LOGAND :
+ /* c == ord('(' )) */ MDPAREN;
+ else if (c == ord('|') && c2 == ord('&'))
c = COPROC;
- else if (c == ';' && c2 == '|')
+ else if (c == ord(';') && c2 == ord('|'))
c = BRKEV;
- else if (c == ';' && c2 == '&')
+ else if (c == ord(';') && c2 == ord('&'))
c = BRKFT;
else
ungetsc(c2);
#ifndef MKSH_SMALL
if (c == BREAK) {
- if ((c2 = getsc()) == '&')
+ if ((c2 = getsc()) == ord('&'))
c = BRKEV;
else
ungetsc(c2);
}
#endif
- } else if (c == '\n') {
+ } else if (c == ord('\n')) {
if (cf & HEREDELIM)
ungetsc(c);
else {
@@ -1025,7 +1025,7 @@
if ((cf & KEYWORD) && (p = ktsearch(&keywords, ident, h)) &&
(!(cf & ESACONLY) || p->val.i == ESAC ||
- p->val.i == /*{*/ '}')) {
+ p->val.i == ord(/*{*/ '}'))) {
afree(yylval.cp, ATEMP);
return (p->val.i);
}
@@ -1038,7 +1038,7 @@
const char *cp = source->str;
/* prefer POSIX but not Korn functions over aliases */
- while (*cp == ' ' || *cp == '\t')
+ while (ctype(*cp, C_BLANK))
/*
* this is like getsc() without skipping
* over Source boundaries (including not
@@ -1136,7 +1136,7 @@
if (!*eofp) {
/* end of here document marker, what to do? */
switch (c) {
- case /*(*/ ')':
+ case ord(/*(*/ ')'):
if (!subshell_nesting_type)
/*-
* not allowed outside $(...) or (...)
@@ -1151,7 +1151,7 @@
* Allow EOF here to commands without trailing
* newlines (mksh -c '...') will work as well.
*/
- case '\n':
+ case ord('\n'):
/* Newline terminates here document marker */
goto heredoc_found_terminator;
}
@@ -1233,7 +1233,7 @@
Source *s = source;
int c;
- while ((c = *s->str++) == 0) {
+ while ((c = ord(*s->str++)) == 0) {
/* return 0 for EOF by default */
s->str = NULL;
switch (s->type) {
@@ -1275,7 +1275,7 @@
source->flags |= s->flags & SF_ALIAS;
s = source;
} else if (*s->u.tblp->val.s &&
- (c = strnul(s->u.tblp->val.s)[-1], ksh_isspace(c))) {
+ ctype((c = strnul(s->u.tblp->val.s)[-1]), C_SPACE)) {
/* pop source stack */
source = s = s->next;
/*
@@ -1435,7 +1435,7 @@
} else if (interactive && cur_prompt == PS1) {
check_for_sole_return:
cp = Xstring(s->xs, xp);
- while (*cp && ctype(*cp, C_IFSWS))
+ while (ctype(*cp, C_IFSWS))
++cp;
if (!*cp) {
histsave(&s->line, NULL, HIST_FLUSH, true);
@@ -1528,7 +1528,7 @@
for (; *cp; cp++) {
if (indelimit && *cp != delimiter)
;
- else if (*cp == '\n' || *cp == '\r') {
+ else if (ctype(*cp, C_CR | C_LF)) {
lines += columns / x_cols + ((*cp == '\n') ? 1 : 0);
columns = 0;
} else if (*cp == '\t') {
@@ -1538,7 +1538,7 @@
columns--;
} else if (*cp == delimiter)
indelimit = !indelimit;
- else if (UTFMODE && ((unsigned char)*cp > 0x7F)) {
+ else if (UTFMODE && (rtt2asc(*cp) > 0x7F)) {
const char *cp2;
columns += utf_widthadj(cp, &cp2);
if (doprint && (indelimit ||
@@ -1580,39 +1580,39 @@
c2 = getsc();
ungetsc(c2);
- if (c2 != /*{*/ '}') {
+ if (ord(c2) != ord(/*{*/ '}')) {
ungetsc(c);
goto out;
}
}
goto ps_common;
case PS_SAW_BANG:
- switch (c) {
- case '@':
- case '#':
- case '-':
- case '?':
+ switch (ord(c)) {
+ case ord('@'):
+ case ord('#'):
+ case ord('-'):
+ case ord('?'):
goto out;
}
goto ps_common;
case PS_INITIAL:
- switch (c) {
- case '%':
+ switch (ord(c)) {
+ case ord('%'):
state = PS_SAW_PERCENT;
goto next;
- case '#':
+ case ord('#'):
state = PS_SAW_HASH;
goto next;
- case '!':
+ case ord('!'):
state = PS_SAW_BANG;
goto next;
}
/* FALLTHROUGH */
case PS_SAW_PERCENT:
ps_common:
- if (ksh_isalphx(c))
+ if (ctype(c, C_ALPHX))
state = PS_IDENT;
- else if (ksh_isdigit(c))
+ else if (ctype(c, C_DIGIT))
state = PS_NUMBER;
else if (ctype(c, C_VAR1))
state = PS_VAR1;
@@ -1620,14 +1620,15 @@
goto out;
break;
case PS_IDENT:
- if (!ksh_isalnux(c)) {
- if (c == '[') {
+ if (!ctype(c, C_ALNUX)) {
+ if (ord(c) == ord('[')) {
char *tmp, *p;
if (!arraysub(&tmp))
yyerror("missing ]");
*wp++ = c;
- for (p = tmp; *p; ) {
+ p = tmp;
+ while (*p) {
Xcheck(*wsp, wp);
*wp++ = *p++;
}
@@ -1640,7 +1641,7 @@
next:
break;
case PS_NUMBER:
- if (!ksh_isdigit(c))
+ if (!ctype(c, C_DIGIT))
goto out;
break;
case PS_VAR1:
@@ -1675,9 +1676,9 @@
c = getsc();
Xcheck(ws, wp);
*wp++ = c;
- if (c == '[')
+ if (ord(c) == ord('['))
depth++;
- else if (c == ']')
+ else if (ord(c) == ord(']'))
depth--;
} while (depth > 0 && c && c != '\n');
@@ -1756,19 +1757,19 @@
{
int c;
- if ((unsigned char)(c = o_getsc_u()) != 0xEF) {
+ if (rtt2asc((c = o_getsc_u())) != 0xEF) {
ungetsc_i(c);
return;
}
- if ((unsigned char)(c = o_getsc_u()) != 0xBB) {
+ if (rtt2asc((c = o_getsc_u())) != 0xBB) {
ungetsc_i(c);
- ungetsc_i(0xEF);
+ ungetsc_i(asc2rtt(0xEF));
return;
}
- if ((unsigned char)(c = o_getsc_u()) != 0xBF) {
+ if (rtt2asc((c = o_getsc_u())) != 0xBF) {
ungetsc_i(c);
- ungetsc_i(0xBB);
- ungetsc_i(0xEF);
+ ungetsc_i(asc2rtt(0xBB));
+ ungetsc_i(asc2rtt(0xEF));
return;
}
UTFMODE |= 8;
diff --git a/src/main.c b/src/main.c
index 1286b07..b4d7244 100644
--- a/src/main.c
+++ b/src/main.c
@@ -34,7 +34,7 @@
#include <locale.h>
#endif
-__RCSID("$MirOS: src/bin/mksh/main.c,v 1.332 2017/04/12 16:01:45 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/main.c,v 1.342 2017/04/28 11:13:47 tg Exp $");
extern char **environ;
@@ -236,6 +236,11 @@
ssize_t k;
#endif
+#if defined(MKSH_EBCDIC) || defined(MKSH_FAUX_EBCDIC)
+ ebcdic_init();
+#endif
+ set_ifs(TC_IFSWS);
+
#ifdef __OS2__
for (i = 0; i < 3; ++i)
if (!isatty(i))
@@ -333,8 +338,6 @@
initvar();
- initctypes();
-
inittraps();
coproc_init();
@@ -409,12 +412,12 @@
/* override default PATH regardless of environment */
#ifdef MKSH_DEFPATH_OVERRIDE
- vp = global(TPATH);
- setstr(vp, MKSH_DEFPATH_OVERRIDE, KSH_RETURN_ERROR);
+ vp = global(TPATH);
+ setstr(vp, MKSH_DEFPATH_OVERRIDE, KSH_RETURN_ERROR);
#endif
/* for security */
- typeset("IFS= \t\n", 0, 0, 0, 0);
+ typeset(TinitIFS, 0, 0, 0, 0);
/* assign default shell variable values */
typeset("PATHSEP=" MKSH_PATHSEPS, 0, 0, 0, 0);
@@ -497,7 +500,7 @@
if (!(s->start = s->str = argv[argi++]))
errorf(Tf_optfoo, "", "", 'c', Treq_arg);
while (*s->str) {
- if (*s->str != ' ' && ctype(*s->str, C_QUOTE))
+ if (ctype(*s->str, C_QUOTE))
break;
s->str++;
}
@@ -1554,7 +1557,7 @@
goto illegal_fd_name;
if (name[0] == 'p')
return (coproc_getfd(mode, emsgp));
- if (!ksh_isdigit(name[0])) {
+ if (!ctype(name[0], C_DIGIT)) {
illegal_fd_name:
if (emsgp)
*emsgp = "illegal file descriptor name";
@@ -1893,7 +1896,7 @@
const struct tbl *a = *((const struct tbl * const *)p1);
const struct tbl *b = *((const struct tbl * const *)p2);
- return (strcmp(a->name, b->name));
+ return (ascstrcmp(a->name, b->name));
}
struct tbl **
diff --git a/src/misc.c b/src/misc.c
index 6957c22..1205072 100644
--- a/src/misc.c
+++ b/src/misc.c
@@ -5,6 +5,8 @@
* Copyright (c) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010,
* 2011, 2012, 2013, 2014, 2015, 2016, 2017
* mirabilos <m@mirbsd.org>
+ * Copyright (c) 2015
+ * Daniel Richard G. <skunk@iSKUNK.ORG>
*
* Provided that these terms and disclaimer and all copyright notices
* are retained or reproduced in an accompanying document, permission
@@ -30,7 +32,7 @@
#include <grp.h>
#endif
-__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.255 2017/04/12 16:46:22 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.279 2017/08/07 21:39:25 tg Exp $");
#define KSH_CHVT_FLAG
#ifdef MKSH_SMALL
@@ -47,7 +49,8 @@
static const unsigned char *pat_scan(const unsigned char *,
const unsigned char *, bool) MKSH_A_PURE;
static int do_gmatch(const unsigned char *, const unsigned char *,
- const unsigned char *, const unsigned char *) MKSH_A_PURE;
+ const unsigned char *, const unsigned char *,
+ const unsigned char *) MKSH_A_PURE;
static const unsigned char *gmatch_cclass(const unsigned char *, unsigned char)
MKSH_A_PURE;
#ifdef KSH_CHVT_CODE
@@ -68,37 +71,6 @@
#define DO_SETUID(func, argvec) func argvec
#endif
-/*
- * Fast character classes
- */
-void
-setctypes(const char *s, int t)
-{
- if (t & C_IFS) {
- unsigned int i = 0;
-
- while (++i <= UCHAR_MAX)
- chtypes[i] &= ~C_IFS;
- /* include '\0' in C_IFS */
- chtypes[0] |= C_IFS;
- }
- while (*s != 0)
- chtypes[(unsigned char)*s++] |= t;
-}
-
-void
-initctypes(void)
-{
- setctypes(letters_uc, C_ALPHX);
- setctypes(letters_lc, C_ALPHX);
- chtypes['_'] |= C_ALPHX;
- setctypes("0123456789", C_DIGIT);
- setctypes(TC_LEX1, C_LEX1);
- setctypes("*@#!$-?", C_VAR1);
- setctypes(TC_IFSWS, C_IFSWS);
- setctypes("=-+?", C_SUBOP1);
- setctypes("\t\n \"#$&'()*;<=>?[\\]`|", C_QUOTE);
-}
/* called from XcheckN() to grow buffer */
char *
@@ -147,7 +119,7 @@
{
size_t i = 0;
- if ((n[0] == '-' || n[0] == '+') && n[1] && !n[2])
+ if (ctype(n[0], C_MINUS | C_PLUS) && n[1] && !n[2])
while (i < NELEM(options)) {
if (OFC(i) == n[1])
return (i);
@@ -299,6 +271,11 @@
} else if ((f == FPOSIX || f == FSH) && newval) {
/* Turning on -o posix or -o sh? */
Flag(FBRACEEXPAND) = 0;
+ /* Turning on -o posix? */
+ if (f == FPOSIX) {
+ /* C locale required for compliance */
+ UTFMODE = 0;
+ }
} else if (f == FTALKING) {
/* Changing interactive flag? */
if ((what == OF_CMDLINE || what == OF_SET) && procpid == kshpid)
@@ -483,7 +460,7 @@
}
}
if (!(go.info & GI_MINUSMINUS) && argv[go.optind] &&
- (argv[go.optind][0] == '-' || argv[go.optind][0] == '+') &&
+ ctype(argv[go.optind][0], C_MINUS | C_PLUS) &&
argv[go.optind][1] == '\0') {
/* lone - clears -v and -x flags */
if (argv[go.optind][0] == '-') {
@@ -512,7 +489,7 @@
for (i = go.optind; argv[i]; i++)
;
qsort(&argv[go.optind], i - go.optind, sizeof(void *),
- xstrcmp);
+ ascpstrcmp);
}
if (arrayset)
go.optind += set_array(array, tobool(arrayset > 0),
@@ -533,7 +510,7 @@
do {
c = *s++;
- } while (ksh_isspace(c));
+ } while (ctype(c, C_SPACE));
switch (c) {
case '-':
@@ -545,7 +522,7 @@
}
do {
- if (!ksh_isdigit(c))
+ if (!ctype(c, C_DIGIT))
/* not numeric */
return (0);
if (num.u > 214748364U)
@@ -585,7 +562,7 @@
sp = cp;
simplify_gmatch_pat1a:
dp = cp;
- se = sp + strlen((const void *)sp);
+ se = strnul(sp);
while ((c = *sp++)) {
if (!ISMAGIC(c)) {
*dp++ = c;
@@ -657,29 +634,30 @@
if (s == NULL || p == NULL)
return (0);
- se = s + strlen(s);
- pe = p + strlen(p);
+ pe = strnul(p);
/*
* isfile is false iff no syntax check has been done on
- * the pattern. If check fails, just to a strcmp().
+ * the pattern. If check fails, just do a strcmp().
*/
- if (!isfile && !has_globbing(p, pe)) {
+ if (!isfile && !has_globbing(p)) {
size_t len = pe - p + 1;
char tbuf[64];
char *t = len <= sizeof(tbuf) ? tbuf : alloc(len, ATEMP);
debunk(t, p, len);
return (!strcmp(t, s));
}
+ se = strnul(s);
/*
* since the do_gmatch() engine sucks so much, we must do some
* pattern simplifications
*/
pnew = simplify_gmatch_pattern((const unsigned char *)p);
- pe = pnew + strlen(pnew);
+ pe = strnul(pnew);
rv = do_gmatch((const unsigned char *)s, (const unsigned char *)se,
- (const unsigned char *)pnew, (const unsigned char *)pe);
+ (const unsigned char *)pnew, (const unsigned char *)pe,
+ (const unsigned char *)s);
afree(pnew, ATEMP);
return (rv);
}
@@ -690,7 +668,7 @@
* Syntax errors are:
* - [ with no closing ]
* - imbalanced $(...) expression
- * - [...] and *(...) not nested (eg, [a$(b|]c), *(a[b|c]d))
+ * - [...] and *(...) not nested (eg, @(a[b|)c], *(a[b|c]d))
*/
/*XXX
* - if no magic,
@@ -701,76 +679,101 @@
* return ?
* - return ?
*/
-int
-has_globbing(const char *xp, const char *xpe)
+bool
+has_globbing(const char *pat)
{
- const unsigned char *p = (const unsigned char *) xp;
- const unsigned char *pe = (const unsigned char *) xpe;
- int c;
- int nest = 0, bnest = 0;
+ unsigned char c, subc;
bool saw_glob = false;
- /* inside [...] */
- bool in_bracket = false;
+ unsigned int nest = 0;
+ const unsigned char *p = (const unsigned char *)pat;
+ const unsigned char *s;
- for (; p < pe; p++) {
- if (!ISMAGIC(*p))
+ while ((c = *p++)) {
+ /* regular character? ok. */
+ if (!ISMAGIC(c))
continue;
- if ((c = *++p) == '*' || c == '?')
+ /* MAGIC + NUL? abort. */
+ if (!(c = *p++))
+ return (false);
+ /* some specials */
+ if (ord(c) == ord('*') || ord(c) == ord('?')) {
+ /* easy glob, accept */
saw_glob = true;
- else if (c == '[') {
- if (!in_bracket) {
- saw_glob = true;
- in_bracket = true;
- if (ISMAGIC(p[1]) && p[2] == '!')
- p += 2;
- if (ISMAGIC(p[1]) && p[2] == ']')
- p += 2;
+ } else if (ord(c) == ord('[')) {
+ /* bracket expression; eat negation and initial ] */
+ if (ISMAGIC(p[0]) && ord(p[1]) == ord('!'))
+ p += 2;
+ if (ISMAGIC(p[0]) && ord(p[1]) == ord(']'))
+ p += 2;
+ /* check next string part */
+ s = p;
+ while ((c = *s++)) {
+ /* regular chars are ok */
+ if (!ISMAGIC(c))
+ continue;
+ /* MAGIC + NUL cannot happen */
+ if (!(c = *s++))
+ return (false);
+ /* terminating bracket? */
+ if (ord(c) == ord(']')) {
+ /* accept and continue */
+ p = s;
+ saw_glob = true;
+ break;
+ }
+ /* sub-bracket expressions */
+ if (ord(c) == ord('[') && (
+ /* collating element? */
+ ord(*s) == ord('.') ||
+ /* equivalence class? */
+ ord(*s) == ord('=') ||
+ /* character class? */
+ ord(*s) == ord(':'))) {
+ /* must stop with exactly the same c */
+ subc = *s++;
+ /* arbitrarily many chars in betwixt */
+ while ((c = *s++))
+ /* but only this sequence... */
+ if (c == subc && ISMAGIC(*s) &&
+ ord(s[1]) == ord(']')) {
+ /* accept, terminate */
+ s += 2;
+ break;
+ }
+ /* EOS without: reject bracket expr */
+ if (!c)
+ break;
+ /* continue; */
+ }
+ /* anything else just goes on */
}
- /*XXX Do we need to check ranges here? POSIX Q */
- } else if (c == ']') {
- if (in_bracket) {
- if (bnest)
- /* [a*(b]) */
- return (0);
- in_bracket = false;
- }
- } else if ((c & 0x80) && vstrchr("*+?@! ", c & 0x7f)) {
+ } else if ((c & 0x80) && ctype(c & 0x7F, C_PATMO | C_SPC)) {
+ /* opening pattern */
saw_glob = true;
- if (in_bracket)
- bnest++;
- else
- nest++;
- } else if (c == '|') {
- if (in_bracket && !bnest)
- /* *(a[foo|bar]) */
- return (0);
- } else if (c == /*(*/ ')') {
- if (in_bracket) {
- if (!bnest--)
- /* *(a[b)c] */
- return (0);
- } else if (nest)
- nest--;
+ ++nest;
+ } else if (ord(c) == ord(/*(*/ ')')) {
+ /* closing pattern */
+ if (nest)
+ --nest;
}
- /*
- * else must be a MAGIC-MAGIC, or MAGIC-!,
- * MAGIC--, MAGIC-], MAGIC-{, MAGIC-, MAGIC-}
- */
}
- return (saw_glob && !in_bracket && !nest);
+ return (saw_glob && !nest);
}
/* Function must return either 0 or 1 (assumed by code for 0x80|'!') */
static int
do_gmatch(const unsigned char *s, const unsigned char *se,
- const unsigned char *p, const unsigned char *pe)
+ const unsigned char *p, const unsigned char *pe,
+ const unsigned char *smin)
{
- unsigned char sc, pc;
+ unsigned char sc, pc, sl = 0;
const unsigned char *prest, *psub, *pnext;
const unsigned char *srest;
if (s == NULL || p == NULL)
return (0);
+ if (s > smin && s <= se)
+ sl = s[-1];
while (p < pe) {
pc = *p++;
sc = s < se ? *s : '\0';
@@ -778,15 +781,39 @@
if (!ISMAGIC(pc)) {
if (sc != pc)
return (0);
+ sl = sc;
continue;
}
- switch (*p++) {
- case '[':
+ switch (ord(*p++)) {
+ case ord('['):
+ /* BSD cclass extension? */
+ if (ISMAGIC(p[0]) && ord(p[1]) == ord('[') &&
+ ord(p[2]) == ord(':') &&
+ ctype((pc = p[3]), C_ANGLE) &&
+ ord(p[4]) == ord(':') &&
+ ISMAGIC(p[5]) && ord(p[6]) == ord(']') &&
+ ISMAGIC(p[7]) && ord(p[8]) == ord(']')) {
+ /* zero-length match */
+ --s;
+ p += 9;
+ /* word begin? */
+ if (ord(pc) == ord('<') &&
+ !ctype(sl, C_ALNUX) &&
+ ctype(sc, C_ALNUX))
+ break;
+ /* word end? */
+ if (ord(pc) == ord('>') &&
+ ctype(sl, C_ALNUX) &&
+ !ctype(sc, C_ALNUX))
+ break;
+ /* neither */
+ return (0);
+ }
if (sc == 0 || (p = gmatch_cclass(p, sc)) == NULL)
return (0);
break;
- case '?':
+ case ord('?'):
if (sc == 0)
return (0);
if (UTFMODE) {
@@ -795,39 +822,39 @@
}
break;
- case '*':
+ case ord('*'):
if (p == pe)
return (1);
s--;
do {
- if (do_gmatch(s, se, p, pe))
+ if (do_gmatch(s, se, p, pe, smin))
return (1);
} while (s++ < se);
return (0);
/**
- * [*+?@!](pattern|pattern|..)
+ * [+*?@!](pattern|pattern|..)
* This is also needed for ${..%..}, etc.
*/
/* matches one or more times */
- case 0x80|'+':
+ case 0x80|ord('+'):
/* matches zero or more times */
- case 0x80|'*':
+ case 0x80|ord('*'):
if (!(prest = pat_scan(p, pe, false)))
return (0);
s--;
/* take care of zero matches */
- if (p[-1] == (0x80 | '*') &&
- do_gmatch(s, se, prest, pe))
+ if (ord(p[-1]) == (0x80 | ord('*')) &&
+ do_gmatch(s, se, prest, pe, smin))
return (1);
for (psub = p; ; psub = pnext) {
pnext = pat_scan(psub, pe, true);
for (srest = s; srest <= se; srest++) {
- if (do_gmatch(s, srest, psub, pnext - 2) &&
- (do_gmatch(srest, se, prest, pe) ||
- (s != srest && do_gmatch(srest,
- se, p - 2, pe))))
+ if (do_gmatch(s, srest, psub, pnext - 2, smin) &&
+ (do_gmatch(srest, se, prest, pe, smin) ||
+ (s != srest &&
+ do_gmatch(srest, se, p - 2, pe, smin))))
return (1);
}
if (pnext == prest)
@@ -836,24 +863,24 @@
return (0);
/* matches zero or once */
- case 0x80|'?':
+ case 0x80|ord('?'):
/* matches one of the patterns */
- case 0x80|'@':
+ case 0x80|ord('@'):
/* simile for @ */
- case 0x80|' ':
+ case 0x80|ord(' '):
if (!(prest = pat_scan(p, pe, false)))
return (0);
s--;
/* Take care of zero matches */
- if (p[-1] == (0x80 | '?') &&
- do_gmatch(s, se, prest, pe))
+ if (ord(p[-1]) == (0x80 | ord('?')) &&
+ do_gmatch(s, se, prest, pe, smin))
return (1);
for (psub = p; ; psub = pnext) {
pnext = pat_scan(psub, pe, true);
srest = prest == pe ? se : s;
for (; srest <= se; srest++) {
- if (do_gmatch(s, srest, psub, pnext - 2) &&
- do_gmatch(srest, se, prest, pe))
+ if (do_gmatch(s, srest, psub, pnext - 2, smin) &&
+ do_gmatch(srest, se, prest, pe, smin))
return (1);
}
if (pnext == prest)
@@ -862,7 +889,7 @@
return (0);
/* matches none of the patterns */
- case 0x80|'!':
+ case 0x80|ord('!'):
if (!(prest = pat_scan(p, pe, false)))
return (0);
s--;
@@ -872,7 +899,7 @@
for (psub = p; ; psub = pnext) {
pnext = pat_scan(psub, pe, true);
if (do_gmatch(s, srest, psub,
- pnext - 2)) {
+ pnext - 2, smin)) {
matched = 1;
break;
}
@@ -880,7 +907,7 @@
break;
}
if (!matched &&
- do_gmatch(srest, se, prest, pe))
+ do_gmatch(srest, se, prest, pe, smin))
return (1);
}
return (0);
@@ -890,55 +917,245 @@
return (0);
break;
}
+ sl = sc;
}
return (s == se);
}
-static const unsigned char *
-gmatch_cclass(const unsigned char *p, unsigned char sub)
-{
- unsigned char c, d;
- bool notp, found = false;
- const unsigned char *orig_p = p;
+/*XXX this is a prime example for bsearch or a const hashtable */
+static const struct cclass {
+ const char *name;
+ uint32_t value;
+} cclasses[] = {
+ /* POSIX */
+ { "alnum", C_ALNUM },
+ { "alpha", C_ALPHA },
+ { "blank", C_BLANK },
+ { "cntrl", C_CNTRL },
+ { "digit", C_DIGIT },
+ { "graph", C_GRAPH },
+ { "lower", C_LOWER },
+ { "print", C_PRINT },
+ { "punct", C_PUNCT },
+ { "space", C_SPACE },
+ { "upper", C_UPPER },
+ { "xdigit", C_SEDEC },
+ /* BSD */
+ /* "<" and ">" are handled inline */
+ /* GNU bash */
+ { "ascii", C_ASCII },
+ { "word", C_ALNUX },
+ /* mksh */
+ { "sh_alias", C_ALIAS },
+ { "sh_edq", C_EDQ },
+ { "sh_ifs", C_IFS },
+ { "sh_ifsws", C_IFSWS },
+ { "sh_nl", C_NL },
+ { "sh_quote", C_QUOTE },
+ /* sentinel */
+ { NULL, 0 }
+};
- if ((notp = tobool(ISMAGIC(*p) && *++p == '!')))
- p++;
- do {
- c = *p++;
+static const unsigned char *
+gmatch_cclass(const unsigned char *pat, unsigned char sc)
+{
+ unsigned char c, subc, lc;
+ const unsigned char *p = pat, *s;
+ bool found = false;
+ bool negated = false;
+ char *subp;
+
+ /* check for negation */
+ if (ISMAGIC(p[0]) && ord(p[1]) == ord('!')) {
+ p += 2;
+ negated = true;
+ }
+ /* make initial ] non-MAGIC */
+ if (ISMAGIC(p[0]) && ord(p[1]) == ord(']'))
+ ++p;
+ /* iterate over bracket expression, debunk()ing on the fly */
+ while ((c = *p++)) {
+ nextc:
+ /* non-regular character? */
if (ISMAGIC(c)) {
- c = *p++;
- if ((c & 0x80) && !ISMAGIC(c)) {
- /* extended pattern matching: *+?@! */
- c &= 0x7F;
- /* XXX the ( char isn't handled as part of [] */
- if (c == ' ')
- /* simile for @: plain (..) */
- c = '(' /*)*/;
+ /* MAGIC + NUL cannot happen */
+ if (!(c = *p++))
+ break;
+ /* terminating bracket? */
+ if (ord(c) == ord(']')) {
+ /* accept and return */
+ return (found != negated ? p : NULL);
+ }
+ /* sub-bracket expressions */
+ if (ord(c) == ord('[') && (
+ /* collating element? */
+ ord(*p) == ord('.') ||
+ /* equivalence class? */
+ ord(*p) == ord('=') ||
+ /* character class? */
+ ord(*p) == ord(':'))) {
+ /* must stop with exactly the same c */
+ subc = *p++;
+ /* save away start of substring */
+ s = p;
+ /* arbitrarily many chars in betwixt */
+ while ((c = *p++))
+ /* but only this sequence... */
+ if (c == subc && ISMAGIC(*p) &&
+ ord(p[1]) == ord(']')) {
+ /* accept, terminate */
+ p += 2;
+ break;
+ }
+ /* EOS without: reject bracket expr */
+ if (!c)
+ break;
+ /* debunk substring */
+ strndupx(subp, s, p - s - 3, ATEMP);
+ debunk(subp, subp, p - s - 3 + 1);
+ cclass_common:
+ /* whither subexpression */
+ if (ord(subc) == ord(':')) {
+ const struct cclass *cls = cclasses;
+
+ /* search for name in cclass list */
+ while (cls->name)
+ if (!strcmp(subp, cls->name)) {
+ /* found, match? */
+ if (ctype(sc,
+ cls->value))
+ found = true;
+ /* break either way */
+ break;
+ } else
+ ++cls;
+ /* that's all here */
+ afree(subp, ATEMP);
+ continue;
+ }
+ /* collating element or equivalence class */
+ /* Note: latter are treated as former */
+ if (ctype(subp[0], C_ASCII) && !subp[1])
+ /* [.a.] where a is one ASCII char */
+ c = subp[0];
+ else
+ /* force no match */
+ c = 0;
+ /* no longer needed */
+ afree(subp, ATEMP);
+ } else if (!ISMAGIC(c) && (c & 0x80)) {
+ /* 0x80|' ' is plain (...) */
+ if ((c &= 0x7F) != ' ') {
+ /* check single match NOW */
+ if (sc == c)
+ found = true;
+ /* next character is (...) */
+ }
+ c = '(' /*)*/;
}
}
- if (c == '\0')
- /* No closing ] - act as if the opening [ was quoted */
- return (sub == '[' ? orig_p : NULL);
- if (ISMAGIC(p[0]) && p[1] == '-' &&
- (!ISMAGIC(p[2]) || p[3] != ']')) {
- /* MAGIC- */
- p += 2;
- d = *p++;
- if (ISMAGIC(d)) {
- d = *p++;
- if ((d & 0x80) && !ISMAGIC(d))
- d &= 0x7f;
- }
- /* POSIX says this is an invalid expression */
- if (c > d)
- return (NULL);
- } else
- d = c;
- if (c == sub || (c <= sub && sub <= d))
- found = true;
- } while (!(ISMAGIC(p[0]) && p[1] == ']'));
+ /* range expression? */
+ if (!(ISMAGIC(p[0]) && ord(p[1]) == ord('-') &&
+ /* not terminating bracket? */
+ (!ISMAGIC(p[2]) || ord(p[3]) != ord(']')))) {
+ /* no, check single match */
+ if (sc == c)
+ /* note: sc is never NUL */
+ found = true;
+ /* do the next "first" character */
+ continue;
+ }
+ /* save lower range bound */
+ lc = c;
+ /* skip over the range operator */
+ p += 2;
+ /* do the same shit as above... almost */
+ subc = 0;
+ if (!(c = *p++))
+ break;
+ /* non-regular character? */
+ if (ISMAGIC(c)) {
+ /* MAGIC + NUL cannot happen */
+ if (!(c = *p++))
+ break;
+ /* sub-bracket expressions */
+ if (ord(c) == ord('[') && (
+ /* collating element? */
+ ord(*p) == ord('.') ||
+ /* equivalence class? */
+ ord(*p) == ord('=') ||
+ /* character class? */
+ ord(*p) == ord(':'))) {
+ /* must stop with exactly the same c */
+ subc = *p++;
+ /* save away start of substring */
+ s = p;
+ /* arbitrarily many chars in betwixt */
+ while ((c = *p++))
+ /* but only this sequence... */
+ if (c == subc && ISMAGIC(*p) &&
+ ord(p[1]) == ord(']')) {
+ /* accept, terminate */
+ p += 2;
+ break;
+ }
+ /* EOS without: reject bracket expr */
+ if (!c)
+ break;
+ /* debunk substring */
+ strndupx(subp, s, p - s - 3, ATEMP);
+ debunk(subp, subp, p - s - 3 + 1);
+ /* whither subexpression */
+ if (ord(subc) == ord(':')) {
+ /* oops, not a range */
- return ((found != notp) ? p+2 : NULL);
+ /* match single previous char */
+ if (lc && (sc == lc))
+ found = true;
+ /* match hyphen-minus */
+ if (ord(sc) == ord('-'))
+ found = true;
+ /* handle cclass common part */
+ goto cclass_common;
+ }
+ /* collating element or equivalence class */
+ /* Note: latter are treated as former */
+ if (ctype(subp[0], C_ASCII) && !subp[1])
+ /* [.a.] where a is one ASCII char */
+ c = subp[0];
+ else
+ /* force no match */
+ c = 0;
+ /* no longer needed */
+ afree(subp, ATEMP);
+ /* other meaning below */
+ subc = 0;
+ } else if (c == (0x80 | ' ')) {
+ /* 0x80|' ' is plain (...) */
+ c = '(' /*)*/;
+ } else if (!ISMAGIC(c) && (c & 0x80)) {
+ c &= 0x7F;
+ subc = '(' /*)*/;
+ }
+ }
+ /* now do the actual range match check */
+ if (lc != 0 /* && c != 0 */ &&
+ asciibetical(lc) <= asciibetical(sc) &&
+ asciibetical(sc) <= asciibetical(c))
+ found = true;
+ /* forced next character? */
+ if (subc) {
+ c = subc;
+ goto nextc;
+ }
+ /* otherwise, just go on with the pattern string */
+ }
+ /* if we broke here, the bracket expression was invalid */
+ if (ord(sc) == ord('['))
+ /* initial opening bracket as literal match */
+ return (pat);
+ /* or rather no match */
+ return (NULL);
}
/* Look for next ) or | (if match_sep) in *(foo|bar) pattern */
@@ -953,16 +1170,30 @@
if ((*++p == /*(*/ ')' && nest-- == 0) ||
(*p == '|' && match_sep && nest == 0))
return (p + 1);
- if ((*p & 0x80) && vstrchr("*+?@! ", *p & 0x7f))
+ if ((*p & 0x80) && ctype(*p & 0x7F, C_PATMO | C_SPC))
nest++;
}
return (NULL);
}
int
-xstrcmp(const void *p1, const void *p2)
+ascstrcmp(const void *s1, const void *s2)
{
- return (strcmp(*(const char * const *)p1, *(const char * const *)p2));
+ const uint8_t *cp1 = s1, *cp2 = s2;
+
+ while (*cp1 == *cp2) {
+ if (*cp1++ == '\0')
+ return (0);
+ ++cp2;
+ }
+ return ((int)asciibetical(*cp1) - (int)asciibetical(*cp2));
+}
+
+int
+ascpstrcmp(const void *pstr1, const void *pstr2)
+{
+ return (ascstrcmp(*(const char * const *)pstr1,
+ *(const char * const *)pstr2));
}
/* Initialise a Getopt structure */
@@ -1032,7 +1263,7 @@
go->info |= flag == '-' ? GI_MINUS : GI_PLUS;
}
go->p++;
- if (c == '?' || c == ':' || c == ';' || c == ',' || c == '#' ||
+ if (ctype(c, C_QUEST | C_COLON | C_HASH) || c == ';' || c == ',' ||
!(o = cstrchr(optionsp, c))) {
if (optionsp[0] == ':') {
go->buf[0] = c;
@@ -1086,13 +1317,14 @@
* argument is missing.
*/
if (argv[go->optind - 1][go->p]) {
- if (ksh_isdigit(argv[go->optind - 1][go->p])) {
+ if (ctype(argv[go->optind - 1][go->p], C_DIGIT)) {
go->optarg = argv[go->optind - 1] + go->p;
go->p = 0;
} else
go->optarg = NULL;
} else {
- if (argv[go->optind] && ksh_isdigit(argv[go->optind][0])) {
+ if (argv[go->optind] &&
+ ctype(argv[go->optind][0], C_DIGIT)) {
go->optarg = argv[go->optind++];
go->p = 0;
} else
@@ -1115,8 +1347,8 @@
bool inquote = true;
/* first, check whether any quotes are needed */
- while ((c = *p++) >= 32)
- if (ctype(c, C_QUOTE))
+ while (rtt2asc(c = *p++) >= 32)
+ if (ctype(c, C_QUOTE | C_SPC))
inquote = false;
p = (const unsigned char *)s;
@@ -1154,6 +1386,7 @@
shf_putc('$', shf);
shf_putc('\'', shf);
while ((c = *p) != 0) {
+#ifndef MKSH_EBCDIC
if (c >= 0xC2) {
n = utf_mbtowc(&wc, (const char *)p);
if (n != (size_t)-1) {
@@ -1162,10 +1395,11 @@
continue;
}
}
+#endif
++p;
switch (c) {
/* see unbksl() in this file for comments */
- case 7:
+ case KSH_BEL:
c = 'a';
if (0)
/* FALLTHROUGH */
@@ -1189,11 +1423,11 @@
c = 't';
if (0)
/* FALLTHROUGH */
- case 11:
+ case KSH_VTAB:
c = 'v';
if (0)
/* FALLTHROUGH */
- case '\033':
+ case KSH_ESC:
/* take E not e because \e is \ in *roff */
c = 'E';
/* FALLTHROUGH */
@@ -1203,7 +1437,12 @@
if (0)
/* FALLTHROUGH */
default:
- if (c < 32 || c > 0x7E) {
+#if defined(MKSH_EBCDIC) || defined(MKSH_FAUX_EBCDIC)
+ if (ksh_isctrl(c))
+#else
+ if (!ctype(c, C_PRINT))
+#endif
+ {
/* FALLTHROUGH */
case '\'':
shf_fprintf(shf, "\\%03o", c);
@@ -2154,13 +2393,7 @@
fc = (*fg)();
switch (fc) {
case 'a':
- /*
- * according to the comments in pdksh, \007 seems
- * to be more portable than \a (due to HP-UX cc,
- * Ultrix cc, old pcc, etc.) so we avoid the escape
- * sequence altogether in mksh and assume ASCII
- */
- wc = 7;
+ wc = KSH_BEL;
break;
case 'b':
wc = '\b';
@@ -2169,11 +2402,11 @@
if (!cstyle)
goto unknown_escape;
c = (*fg)();
- wc = CTRL(c);
+ wc = ksh_toctrl(c);
break;
case 'E':
case 'e':
- wc = 033;
+ wc = KSH_ESC;
break;
case 'f':
wc = '\f';
@@ -2188,8 +2421,7 @@
wc = '\t';
break;
case 'v':
- /* assume ASCII here as well */
- wc = 11;
+ wc = KSH_VTAB;
break;
case '1':
case '2':
@@ -2212,7 +2444,7 @@
wc = 0;
i = 3;
while (i--)
- if ((c = (*fg)()) >= ord('0') && c <= ord('7'))
+ if (ctype((c = (*fg)()), C_OCTAL))
wc = (wc << 3) + ksh_numdig(c);
else {
(*fp)(c);
@@ -2240,17 +2472,17 @@
n = 0;
while (n < i || i == -1) {
wc <<= 4;
- if ((c = (*fg)()) >= ord('0') && c <= ord('9'))
- wc += ksh_numdig(c);
- else if (c >= ord('A') && c <= ord('F'))
- wc += ksh_numuc(c) + 10;
- else if (c >= ord('a') && c <= ord('f'))
- wc += ksh_numlc(c) + 10;
- else {
+ if (!ctype((c = (*fg)()), C_SEDEC)) {
wc >>= 4;
(*fp)(c);
break;
}
+ if (ctype(c, C_DIGIT))
+ wc += ksh_numdig(c);
+ else if (ctype(c, C_UPPER))
+ wc += ksh_numuc(c) + 10;
+ else
+ wc += ksh_numlc(c) + 10;
++n;
}
if (!n)
diff --git a/src/mksh.1 b/src/mksh.1
index 6a2609a..aa67ac9 100644
--- a/src/mksh.1
+++ b/src/mksh.1
@@ -1,4 +1,4 @@
-.\" $MirOS: src/bin/mksh/mksh.1,v 1.442 2017/04/12 18:30:58 tg Exp $
+.\" $MirOS: src/bin/mksh/mksh.1,v 1.451 2017/08/16 21:40:14 tg Exp $
.\" $OpenBSD: ksh.1,v 1.160 2015/07/04 13:27:04 feinerer Exp $
.\"-
.\" Copyright © 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009,
@@ -76,7 +76,7 @@
.\" with -mandoc, it might implement .Mx itself, but we want to
.\" use our own definition. And .Dd must come *first*, always.
.\"
-.Dd $Mdocdate: April 12 2017 $
+.Dd $Mdocdate: August 16 2017 $
.\"
.\" Check which macro package we use, and do other -mdoc setup.
.\"
@@ -2091,7 +2091,7 @@
which did print the delimiter character so you were out of luck
if you did not have any non-printing characters.
.Pp
-Since Backslashes and other special characters may be
+Since backslashes and other special characters may be
interpreted by the shell, to set
.Ev PS1
either escape the backslash itself
@@ -2106,7 +2106,7 @@
.Pq colour would work, too ,
in the prompt string:
.Bd -literal -offset indent
-x=$(print \e\e001)
+x=$(print \e\e001) # otherwise unused char
PS1="$x$(print \e\er)$x$(tput so)$x\e$PWD$x$(tput se)$x\*(Gt "
.Ed
.Pp
@@ -3066,11 +3066,13 @@
.Ic alias
lists all aliases.
For any name without a value, the existing alias is listed.
-Any name with a value defines an alias (see
+Any name with a value defines an alias; see
.Sx Aliases
-above).
-.Li \&[A\-Za\-z0\-9_!%,@\-]
-are valid in names except they may not begin with a hyphen-minus.
+above.
+.Li \&[][A\-Za\-z0\-9_!%,.@:\-]
+are valid in names, except they may not begin with a hyphen-minus, and
+.Ic \&[[
+is not a valid alias name.
.Pp
When listing aliases, one of two formats is used.
Normally, aliases are listed as
@@ -3162,7 +3164,8 @@
.Pp
Control characters may be written using caret notation
i.e. \*(haX represents Ctrl-X.
-Note that although only two prefix characters (usually ESC and \*(haX)
+The caret itself can be escaped by a backslash, which also escapes itself.
+Note that although only three prefix characters (usually ESC, \*(haX and NUL)
are supported, some multi-character sequences can be supported.
.Pp
The following default bindings show how the arrow keys, the home, end and
@@ -4305,9 +4308,11 @@
.Dq sh
and this autodetection feature is compiled in
.Pq not in MirBSD .
-As a side effect, setting this flag turns off
+As a side effect, setting this flag turns off the
.Ic braceexpand
-mode, which can be turned back on manually, and
+and
+.Ic utf8\-mode
+flags, which can be turned back on manually, and
.Ic sh
mode (unless both are enabled at the same time).
.It Fl o Ic sh
@@ -5382,6 +5387,11 @@
.Ic getopts
does not accept options with a leading
.Ql + .
+.It
+.Ic exec
+skips builtins, functions and other commands and uses a
+.Ev PATH
+search to determine the utility to execute.
.El
.Ss SH mode
Compatibility mode; intended for use with legacy scripts that
@@ -5537,7 +5547,7 @@
.No INTR Pq \*(haC ,
.No \*(haG
.Xc
-Abort the current command, empty the line buffer and
+Abort the current command, save it to the history, empty the line buffer and
set the exit state to interrupted.
.It auto\-insert: Op Ar n
Simply causes the character to appear as literal input.
@@ -5572,7 +5582,8 @@
words, leaving the cursor past the end of the last word.
.It clear\-screen: \*(ha[\*(haL
Prints a compile-time configurable sequence to clear the screen and home
-the cursor, redraws the entire prompt and the currently edited input line.
+the cursor, redraws the last line of the prompt string and the currently
+edited input line.
The default sequence works for almost all standard terminals.
.It comment: \*(ha[#
If the current line does not begin with a comment character, one is added at
@@ -6434,7 +6445,7 @@
They move as expected, both in insert and command mode.
.It Ar intr No and Ar quit
The interrupt and quit terminal characters cause the current line to be
-deleted and a new prompt to be printed.
+removed to the history and a new prompt to be printed.
.El
.Sh FILES
.Bl -tag -width XetcXsuid_profile -compact
@@ -6584,7 +6595,7 @@
.An Michael Rendell .
The effort of several projects, such as Debian and OpenBSD, and other
contributors including our users, to improve the shell is appreciated.
-See the documentation, web site and CVS for details.
+See the documentation, website and source code (CVS) for details.
.Pp
.Nm mksh\-os2
is developed by
@@ -6594,6 +6605,10 @@
is developed by
.An Michael Langguth Aq Mt lan@scalaris.com .
.Pp
+.Nm mksh Ns / Ns Tn z/OS
+is contributed by
+.An Daniel Richard G. Aq Mt skunk@iSKUNK.ORG .
+.Pp
The BSD daemon is Copyright \(co Marshall Kirk McKusick.
The complete legalese is at:
.Pa http://www.mirbsd.org/TaC\-mksh.txt
@@ -6633,12 +6648,14 @@
locale.
.Nm mksh Ns 's
.Ic utf8\-mode
+.Em must
+be disabled in POSIX mode, and it
only supports the Unicode BMP (Basic Multilingual Plane) and maps
raw octets into the U+EF80..U+EFFF wide character range; compare
.Sx Arithmetic expressions .
The following
.Tn POSIX
-.Nm sh
+.Nm sh Ns -compatible
code toggles the
.Ic utf8\-mode
option dependent on the current
@@ -6680,7 +6697,7 @@
.Xr memmove 3 .
.Pp
This document attempts to describe
-.Nm mksh\ R55
+.Nm mksh\ R56
and up,
.\" with vendor patches from insert-your-name-here,
compiled without any options impacting functionality, such as
@@ -6881,3 +6898,9 @@
.Nm
separates the shortcuts: Cursor Up goes up one command
and PgUp searches the history as described above.
+.Ss "My question is not answered here!"
+Check
+.Pa http://www.mirbsd.org/mksh\-faq.htm
+which contains a collection of frequently asked questions about
+.Nm
+in general, for packagers, etc. while these above are in user scope.
diff --git a/src/os2.c b/src/os2.c
index 5d39630..fc27d5a 100644
--- a/src/os2.c
+++ b/src/os2.c
@@ -1,6 +1,8 @@
/*-
* Copyright (c) 2015
* KO Myung-Hun <komh@chollian.net>
+ * Copyright (c) 2017
+ * mirabilos <m@mirbsd.org>
*
* Provided that these terms and disclaimer and all copyright notices
* are retained or reproduced in an accompanying document, permission
@@ -28,7 +30,7 @@
#include <unistd.h>
#include <process.h>
-__RCSID("$MirOS: src/bin/mksh/os2.c,v 1.1 2017/04/02 15:00:44 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/os2.c,v 1.2 2017/04/29 22:04:29 tg Exp $");
static char *remove_trailing_dots(char *);
static int access_stat_ex(int (*)(), const char *, void *);
@@ -247,9 +249,9 @@
static char *
remove_trailing_dots(char *name)
{
- char *p;
+ char *p = strnul(name);
- for (p = name + strlen(name); --p > name && *p == '.'; )
+ while (--p > name && *p == '.')
/* nothing */;
if (*p != '.' && *p != '/' && *p != '\\' && *p != ':')
diff --git a/src/sh.h b/src/sh.h
index 5b36378..88883cb 100644
--- a/src/sh.h
+++ b/src/sh.h
@@ -112,6 +112,13 @@
#include <wchar.h>
#endif
+/* monkey-patch known-bad offsetof versions to quell a warning */
+#if (defined(__KLIBC__) || defined(__dietlibc__)) && \
+ ((defined(__GNUC__) && (__GNUC__ > 3)) || defined(__NWCC__))
+#undef offsetof
+#define offsetof(s, e) __builtin_offsetof(s, e)
+#endif
+
#undef __attribute__
#if HAVE_ATTRIBUTE_BOUNDED
#define MKSH_A_BOUNDED(x,y,z) __attribute__((__bounded__(x, y, z)))
@@ -175,9 +182,9 @@
#endif
#ifdef EXTERN
-__RCSID("$MirOS: src/bin/mksh/sh.h,v 1.808 2017/04/12 17:38:46 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/sh.h,v 1.841 2017/08/29 13:38:31 tg Exp $");
#endif
-#define MKSH_VERSION "R55 2017/04/12"
+#define MKSH_VERSION "R56 2017/08/29"
/* arithmetic types: C implementation */
#if !HAVE_CAN_INTTYPES
@@ -257,6 +264,23 @@
#ifndef MKSH_INCLUDES_ONLY
+/* EBCDIC fun */
+
+/* see the large comment in shf.c for an EBCDIC primer */
+
+#if defined(MKSH_FOR_Z_OS) && defined(__MVS__) && defined(__IBMC__) && defined(__CHARSET_LIB)
+# if !__CHARSET_LIB && !defined(MKSH_EBCDIC)
+# error "Please compile with Build.sh -E for EBCDIC!"
+# endif
+# if __CHARSET_LIB && defined(MKSH_EBCDIC)
+# error "Please compile without -E argument to Build.sh for ASCII!"
+# endif
+# if __CHARSET_LIB && !defined(_ENHANCED_ASCII_EXT)
+ /* go all-out on ASCII */
+# define _ENHANCED_ASCII_EXT 0xFFFFFFFF
+# endif
+#endif
+
/* extra types */
/* getrusage does not exist on OS/2 kLIBC */
@@ -349,6 +373,8 @@
#define ksh_NSIG (_SIGMAX + 1)
#elif defined(NSIG_MAX)
#define ksh_NSIG (NSIG_MAX)
+#elif defined(MKSH_FOR_Z_OS)
+#define ksh_NSIG 40
#else
# error Please have your platform define NSIG.
#endif
@@ -487,6 +513,23 @@
#define ISTRIP 0
#endif
+#ifdef MKSH_EBCDIC
+#define KSH_BEL '\a'
+#define KSH_ESC 047
+#define KSH_ESC_STRING "\047"
+#define KSH_VTAB '\v'
+#else
+/*
+ * According to the comments in pdksh, \007 seems to be more portable
+ * than \a (HP-UX cc, Ultrix cc, old pcc, etc.) so we avoid the escape
+ * sequence if ASCII can be assumed.
+ */
+#define KSH_BEL 7
+#define KSH_ESC 033
+#define KSH_ESC_STRING "\033"
+#define KSH_VTAB 11
+#endif
+
/* some useful #defines */
#ifdef EXTERN
@@ -498,16 +541,22 @@
#endif
/* define bit in flag */
-#define BIT(i) (1 << (i))
+#define BIT(i) (1U << (i))
#define NELEM(a) (sizeof(a) / sizeof((a)[0]))
/*
* Make MAGIC a char that might be printed to make bugs more obvious, but
* not a char that is used often. Also, can't use the high bit as it causes
* portability problems (calling strchr(x, 0x80 | 'x') is error prone).
+ *
+ * MAGIC can be followed by MAGIC (to escape the octet itself) or one of:
+ * ' !)*,-?[]{|}' 0x80|' !*+?@' (probably… hysteric raisins abound)
+ *
+ * The |0x80 is likely unsafe on EBCDIC :( though the listed chars are
+ * low-bit7 at least on cp1047 so YMMV
*/
-#define MAGIC (7) /* prefix for *?[!{,} during expand */
-#define ISMAGIC(c) ((unsigned char)(c) == MAGIC)
+#define MAGIC KSH_BEL /* prefix for *?[!{,} during expand */
+#define ISMAGIC(c) (ord(c) == ord(MAGIC))
EXTERN const char *safe_prompt; /* safe prompt if PS1 substitution fails */
@@ -521,17 +570,21 @@
#else
#define KSH_VERSIONNAME_TEXTMODE ""
#endif
+#ifdef MKSH_EBCDIC
+#define KSH_VERSIONNAME_EBCDIC " +EBCDIC"
+#else
+#define KSH_VERSIONNAME_EBCDIC ""
+#endif
#ifndef KSH_VERSIONNAME_VENDOR_EXT
#define KSH_VERSIONNAME_VENDOR_EXT ""
#endif
EXTERN const char initvsn[] E_INIT("KSH_VERSION=@(#)" KSH_VERSIONNAME_ISLEGACY \
- " KSH " MKSH_VERSION KSH_VERSIONNAME_TEXTMODE KSH_VERSIONNAME_VENDOR_EXT);
+ " KSH " MKSH_VERSION KSH_VERSIONNAME_EBCDIC KSH_VERSIONNAME_TEXTMODE \
+ KSH_VERSIONNAME_VENDOR_EXT);
#define KSH_VERSION (initvsn + /* "KSH_VERSION=@(#)" */ 16)
EXTERN const char digits_uc[] E_INIT("0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ");
EXTERN const char digits_lc[] E_INIT("0123456789abcdefghijklmnopqrstuvwxyz");
-#define letters_uc (digits_uc + 10)
-#define letters_lc (digits_lc + 10)
/*
* Evil hack for const correctness due to API brokenness
@@ -585,15 +638,12 @@
#endif
#if defined(DEBUG) || defined(__COVERITY__)
-#define mkssert(e) do { if (!(e)) exit(255); } while (/* CONSTCOND */ 0)
#ifndef DEBUG_LEAKS
#define DEBUG_LEAKS
#endif
-#else
-#define mkssert(e) do { } while (/* CONSTCOND */ 0)
#endif
-#if (!defined(MKSH_BUILDMAKEFILE4BSD) && !defined(MKSH_BUILDSH)) || (MKSH_BUILD_R != 551)
+#if (!defined(MKSH_BUILDMAKEFILE4BSD) && !defined(MKSH_BUILDSH)) || (MKSH_BUILD_R != 562)
#error Must run Build.sh to compile this.
extern void thiswillneverbedefinedIhope(void);
int
@@ -605,7 +655,7 @@
#endif
/* use this ipv strchr(s, 0) but no side effects in s! */
-#define strnul(s) ((s) + strlen(s))
+#define strnul(s) ((s) + strlen((const void *)s))
#define utf_ptradjx(src, dst) do { \
(dst) = (src) + utf_ptradj(src); \
@@ -621,7 +671,7 @@
#else
/* be careful to evaluate arguments only once! */
#define strdupx(d, s, ap) do { \
- const char *strdup_src = (s); \
+ const char *strdup_src = (const void *)(s); \
char *strdup_dst = NULL; \
\
if (strdup_src != NULL) { \
@@ -632,7 +682,7 @@
(d) = strdup_dst; \
} while (/* CONSTCOND */ 0)
#define strndupx(d, s, n, ap) do { \
- const char *strdup_src = (s); \
+ const char *strdup_src = (const void *)(s); \
char *strdup_dst = NULL; \
\
if (strdup_src != NULL) { \
@@ -753,8 +803,8 @@
struct sretrace_info;
struct yyrecursive_state;
-EXTERN struct sretrace_info *retrace_info E_INIT(NULL);
-EXTERN int subshell_nesting_type E_INIT(0);
+EXTERN struct sretrace_info *retrace_info;
+EXTERN int subshell_nesting_type;
extern struct env {
ALLOC_ITEM alloc_INT; /* internal, do not touch */
@@ -865,8 +915,8 @@
EXTERN const char T4spaces[] E_INIT(" ");
#define T1space (Treal_sp2 + 5)
#define Tcolsp (Tf_sD_ + 2)
-EXTERN const char TC_LEX1[] E_INIT("|&;<>() \t\n");
-#define TC_IFSWS (TC_LEX1 + 7)
+#define TC_IFSWS (TinitIFS + 4)
+EXTERN const char TinitIFS[] E_INIT("IFS= \t\n");
EXTERN const char TFCEDIT_dollaru[] E_INIT("${FCEDIT:-/bin/ed} $_");
#define Tspdollaru (TFCEDIT_dollaru + 18)
EXTERN const char Tsgdot[] E_INIT("*=.");
@@ -1026,8 +1076,8 @@
#define T4spaces " "
#define T1space " "
#define Tcolsp ": "
-#define TC_LEX1 "|&;<>() \t\n"
#define TC_IFSWS " \t\n"
+#define TinitIFS "IFS= \t\n"
#define TFCEDIT_dollaru "${FCEDIT:-/bin/ed} $_"
#define Tspdollaru " $_"
#define Tsgdot "*=."
@@ -1277,7 +1327,7 @@
TMOUT_LEAVING /* have timed out */
};
EXTERN unsigned int ksh_tmout;
-EXTERN enum tmout_enum ksh_tmout_state E_INIT(TMOUT_EXECUTING);
+EXTERN enum tmout_enum ksh_tmout_state;
/* For "You have stopped jobs" message */
EXTERN bool really_exit;
@@ -1285,39 +1335,178 @@
/*
* fast character classes
*/
-#define C_ALPHX BIT(0) /* A-Za-z_ */
-#define C_DIGIT BIT(1) /* 0-9 */
-#define C_LEX1 BIT(2) /* \t \n\0|&;<>() */
-#define C_VAR1 BIT(3) /* *@#!$-? */
-#define C_IFSWS BIT(4) /* \t \n (IFS white space) */
-#define C_SUBOP1 BIT(5) /* "=-+?" */
-#define C_QUOTE BIT(6) /* \t\n "#$&'()*;<=>?[\]`| (needing quoting) */
-#define C_IFS BIT(7) /* $IFS */
-extern unsigned char chtypes[];
+/* internal types, do not reference */
-#define ctype(c, t) tobool(chtypes[(unsigned char)(c)] & (t))
-#define ord(c) ((int)(unsigned char)(c))
-#define ksh_issubop2(c) tobool((c) == ord('#') || (c) == ord('%'))
-#define ksh_isalias(c) (ctype((c), C_ALPHX | C_DIGIT) || (c) == ord('!') || \
- (c) == ord('%') || (c) == ord(',') || \
- (c) == ord('@') || (c) == ord('-'))
-#define ksh_isalpha(c) (ctype((c), C_ALPHX) && (c) != ord('_'))
-#define ksh_isalphx(c) ctype((c), C_ALPHX)
-#define ksh_isalnux(c) ctype((c), C_ALPHX | C_DIGIT)
-#define ksh_isdigit(c) ctype((c), C_DIGIT)
-#define ksh_islower(c) (((c) >= 'a') && ((c) <= 'z'))
-#define ksh_isupper(c) (((c) >= 'A') && ((c) <= 'Z'))
-#define ksh_tolower(c) (ksh_isupper(c) ? (c) - 'A' + 'a' : (c))
-#define ksh_toupper(c) (ksh_islower(c) ? (c) - 'a' + 'A' : (c))
-#define ksh_isdash(s) (((s)[0] == '-') && ((s)[1] == '\0'))
-#define ksh_isspace(c) ((((c) >= 0x09) && ((c) <= 0x0D)) || ((c) == 0x20))
-#define ksh_eq(c,u,l) (((c) | 0x20) == (l))
-#define ksh_numdig(c) ((c) - ord('0'))
-#define ksh_numuc(c) ((c) - ord('A'))
-#define ksh_numlc(c) ((c) - ord('a'))
+/* initially empty — filled at runtime from $IFS */
+#define CiIFS BIT(0)
+#define CiCNTRL BIT(1) /* \x01‥\x08\x0E‥\x1F\x7F */
+#define CiUPPER BIT(2) /* A‥Z */
+#define CiLOWER BIT(3) /* a‥z */
+#define CiHEXLT BIT(4) /* A‥Fa‥f */
+#define CiOCTAL BIT(5) /* 0‥7 */
+#define CiQCL BIT(6) /* &();| */
+#define CiALIAS BIT(7) /* !,.@ */
+#define CiQCX BIT(8) /* *[\\ */
+#define CiVAR1 BIT(9) /* !*@ */
+#define CiQCM BIT(10) /* /^~ */
+#define CiDIGIT BIT(11) /* 89 */
+#define CiQC BIT(12) /* "' */
+#define CiSPX BIT(13) /* \x0B\x0C */
+#define CiCURLY BIT(14) /* {} */
+#define CiANGLE BIT(15) /* <> */
+#define CiNUL BIT(16) /* \x00 */
+#define CiTAB BIT(17) /* \x09 */
+#define CiNL BIT(18) /* \x0A */
+#define CiCR BIT(19) /* \x0D */
+#define CiSP BIT(20) /* \x20 */
+#define CiHASH BIT(21) /* # */
+#define CiSS BIT(22) /* $ */
+#define CiPERCT BIT(23) /* % */
+#define CiPLUS BIT(24) /* + */
+#define CiMINUS BIT(25) /* - */
+#define CiCOLON BIT(26) /* : */
+#define CiEQUAL BIT(27) /* = */
+#define CiQUEST BIT(28) /* ? */
+#define CiBRACK BIT(29) /* ] */
+#define CiUNDER BIT(30) /* _ */
+#define CiGRAVE BIT(31) /* ` */
+/* out of space, but one for *@ would make sense, possibly others */
-EXTERN int ifs0 E_INIT(' '); /* for "$*" */
+/* compile-time initialised, ASCII only */
+extern const uint32_t tpl_ctypes[128];
+/* run-time, contains C_IFS as well, full 2⁸ octet range */
+EXTERN uint32_t ksh_ctypes[256];
+/* first octet of $IFS, for concatenating "$*" */
+EXTERN char ifs0;
+
+/* external types */
+
+/* !%,-.0‥9:@A‥Z[]_a‥z valid characters in alias names */
+#define C_ALIAS (CiALIAS | CiBRACK | CiCOLON | CiDIGIT | CiLOWER | CiMINUS | CiOCTAL | CiPERCT | CiUNDER | CiUPPER)
+/* 0‥9A‥Za‥z alphanumerical */
+#define C_ALNUM (CiDIGIT | CiLOWER | CiOCTAL | CiUPPER)
+/* 0‥9A‥Z_a‥z alphanumerical plus underscore (“word character”) */
+#define C_ALNUX (CiDIGIT | CiLOWER | CiOCTAL | CiUNDER | CiUPPER)
+/* A‥Za‥z alphabetical (upper plus lower) */
+#define C_ALPHA (CiLOWER | CiUPPER)
+/* A‥Z_a‥z alphabetical plus underscore (identifier lead) */
+#define C_ALPHX (CiLOWER | CiUNDER | CiUPPER)
+/* \x01‥\x7F 7-bit ASCII except NUL */
+#define C_ASCII (CiALIAS | CiANGLE | CiBRACK | CiCNTRL | CiCOLON | CiCR | CiCURLY | CiDIGIT | CiEQUAL | CiGRAVE | CiHASH | CiLOWER | CiMINUS | CiNL | CiOCTAL | CiPERCT | CiPLUS | CiQC | CiQCL | CiQCM | CiQCX | CiQUEST | CiSP | CiSPX | CiSS | CiTAB | CiUNDER | CiUPPER)
+/* \x09\x20 tab and space */
+#define C_BLANK (CiSP | CiTAB)
+/* \x09\x20"' separator for completion */
+#define C_CFS (CiQC | CiSP | CiTAB)
+/* \x00‥\x1F\x7F POSIX control characters */
+#define C_CNTRL (CiCNTRL | CiCR | CiNL | CiNUL | CiSPX | CiTAB)
+/* 0‥9 decimal digits */
+#define C_DIGIT (CiDIGIT | CiOCTAL)
+/* &();`| editor x_locate_word() command */
+#define C_EDCMD (CiGRAVE | CiQCL)
+/* \x09\x0A\x20"&'():;<=>`| editor non-word characters */
+#define C_EDNWC (CiANGLE | CiCOLON | CiEQUAL | CiGRAVE | CiNL | CiQC | CiQCL | CiSP | CiTAB)
+/* "#$&'()*:;<=>?[\\`{|} editor quotes for tab completion */
+#define C_EDQ (CiANGLE | CiCOLON | CiCURLY | CiEQUAL | CiGRAVE | CiHASH | CiQC | CiQCL | CiQCX | CiQUEST | CiSS)
+/* !‥~ POSIX graphical (alphanumerical plus punctuation) */
+#define C_GRAPH (C_PUNCT | CiDIGIT | CiLOWER | CiOCTAL | CiUPPER)
+/* A‥Fa‥f hex letter */
+#define C_HEXLT CiHEXLT
+/* \x00 + $IFS IFS whitespace, IFS non-whitespace, NUL */
+#define C_IFS (CiIFS | CiNUL)
+/* \x09\x0A\x20 IFS whitespace */
+#define C_IFSWS (CiNL | CiSP | CiTAB)
+/* \x09\x0A\x20&();<>| (for the lexer) */
+#define C_LEX1 (CiANGLE | CiNL | CiQCL | CiSP | CiTAB)
+/* a‥z lowercase letters */
+#define C_LOWER CiLOWER
+/* not alnux or dollar separator for motion */
+#define C_MFS (CiALIAS | CiANGLE | CiBRACK | CiCNTRL | CiCOLON | CiCR | CiCURLY | CiEQUAL | CiGRAVE | CiHASH | CiMINUS | CiNL | CiNUL | CiPERCT | CiPLUS | CiQC | CiQCL | CiQCM | CiQCX | CiQUEST | CiSP | CiSPX | CiTAB)
+/* 0‥7 octal digit */
+#define C_OCTAL CiOCTAL
+/* !*+?@ pattern magical operator, except space */
+#define C_PATMO (CiPLUS | CiQUEST | CiVAR1)
+/* \x20‥~ POSIX printable characters (graph plus space) */
+#define C_PRINT (C_GRAPH | CiSP)
+/* !"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~ POSIX punctuation */
+#define C_PUNCT (CiALIAS | CiANGLE | CiBRACK | CiCOLON | CiCURLY | CiEQUAL | CiGRAVE | CiHASH | CiMINUS | CiPERCT | CiPLUS | CiQC | CiQCL | CiQCM | CiQCX | CiQUEST | CiSS | CiUNDER)
+/* \x09\x0A"#$&'()*;<=>?[\\]`| characters requiring quoting, minus space */
+#define C_QUOTE (CiANGLE | CiBRACK | CiEQUAL | CiGRAVE | CiHASH | CiNL | CiQC | CiQCL | CiQCX | CiQUEST | CiSS | CiTAB)
+/* 0‥9A‥Fa‥f hexadecimal digit */
+#define C_SEDEC (CiDIGIT | CiHEXLT | CiOCTAL)
+/* \x09‥\x0D\x20 POSIX space class */
+#define C_SPACE (CiCR | CiNL | CiSP | CiSPX | CiTAB)
+/* +-=? substitution operations with word */
+#define C_SUB1 (CiEQUAL | CiMINUS | CiPLUS | CiQUEST)
+/* #% substitution operations with pattern */
+#define C_SUB2 (CiHASH | CiPERCT)
+/* A‥Z uppercase letters */
+#define C_UPPER CiUPPER
+/* !#$*-?@ substitution parameters, other than positional */
+#define C_VAR1 (CiHASH | CiMINUS | CiQUEST | CiSS | CiVAR1)
+
+/* individual chars you might like */
+#define C_ANGLE CiANGLE /* <> angle brackets */
+#define C_COLON CiCOLON /* : colon */
+#define C_CR CiCR /* \x0D ASCII carriage return */
+#define C_DOLAR CiSS /* $ dollar sign */
+#define C_EQUAL CiEQUAL /* = equals sign */
+#define C_GRAVE CiGRAVE /* ` accent gravis */
+#define C_HASH CiHASH /* # hash sign */
+#define C_LF CiNL /* \x0A ASCII line feed */
+#define C_MINUS CiMINUS /* - hyphen-minus */
+#ifdef MKSH_WITH_TEXTMODE
+#define C_NL (CiNL | CiCR) /* CR or LF under OS/2 TEXTMODE */
+#else
+#define C_NL CiNL /* LF only like under Unix */
+#endif
+#define C_NUL CiNUL /* \x00 ASCII NUL */
+#define C_PLUS CiPLUS /* + plus sign */
+#define C_QC CiQC /* "' quote characters */
+#define C_QUEST CiQUEST /* ? question mark */
+#define C_SPC CiSP /* \x20 ASCII space */
+#define C_TAB CiTAB /* \x09 ASCII horizontal tabulator */
+#define C_UNDER CiUNDER /* _ underscore */
+
+/* identity transform of octet */
+#define ord(c) ((unsigned int)(unsigned char)(c))
+#if defined(MKSH_EBCDIC) || defined(MKSH_FAUX_EBCDIC)
+EXTERN unsigned short ebcdic_map[256];
+EXTERN unsigned char ebcdic_rtt_toascii[256];
+EXTERN unsigned char ebcdic_rtt_fromascii[256];
+extern void ebcdic_init(void);
+/* one-way to-ascii-or-high conversion, for POSIX locale ordering */
+#define asciibetical(c) ((unsigned int)ebcdic_map[(unsigned char)(c)])
+/* two-way round-trip conversion, for general use */
+#define rtt2asc(c) ebcdic_rtt_toascii[(unsigned char)(c)]
+#define asc2rtt(c) ebcdic_rtt_fromascii[(unsigned char)(c)]
+/* case-independent char comparison */
+#define ksh_eq(c,u,l) (ord(c) == ord(u) || ord(c) == ord(l))
+#else
+#define asciibetical(c) ord(c)
+#define rtt2asc(c) ((unsigned char)(c))
+#define asc2rtt(c) ((unsigned char)(c))
+#define ksh_eq(c,u,l) ((ord(c) | 0x20) == ord(l))
+#endif
+/* control character foo */
+#ifdef MKSH_EBCDIC
+#define ksh_isctrl(c) (ord(c) < 0x40 || ord(c) == 0xFF)
+#else
+#define ksh_isctrl(c) ((ord(c) & 0x7F) < 0x20 || (c) == 0x7F)
+#endif
+/* new fast character classes */
+#define ctype(c,t) tobool(ksh_ctypes[ord(c)] & (t))
+/* helper functions */
+#define ksh_isdash(s) tobool(ord((s)[0]) == '-' && ord((s)[1]) == '\0')
+/* invariant distance even in EBCDIC */
+#define ksh_tolower(c) (ctype(c, C_UPPER) ? (c) - 'A' + 'a' : (c))
+#define ksh_toupper(c) (ctype(c, C_LOWER) ? (c) - 'a' + 'A' : (c))
+/* strictly speaking rtt2asc() here, but this works even in EBCDIC */
+#define ksh_numdig(c) (ord(c) - ord('0'))
+#define ksh_numuc(c) (rtt2asc(c) - rtt2asc('A'))
+#define ksh_numlc(c) (rtt2asc(c) - rtt2asc('a'))
+#define ksh_toctrl(c) asc2rtt(ord(c) == ord('?') ? 0x7F : rtt2asc(c) & 0x9F)
+#define ksh_unctrl(c) asc2rtt(rtt2asc(c) ^ 0x40U)
/* Argument parsing for built-in commands and getopts command */
@@ -1990,12 +2179,77 @@
#define HERES 10 /* max number of << in line */
-#undef CTRL
-#define CTRL(x) ((x) == '?' ? 0x7F : (x) & 0x1F) /* ASCII */
-#define UNCTRL(x) ((x) ^ 0x40) /* ASCII */
-#define ISCTRL(x) (((signed char)((uint8_t)(x) + 1)) < 33)
+#ifdef MKSH_EBCDIC
+#define CTRL_AT (0x00U)
+#define CTRL_A (0x01U)
+#define CTRL_B (0x02U)
+#define CTRL_C (0x03U)
+#define CTRL_D (0x37U)
+#define CTRL_E (0x2DU)
+#define CTRL_F (0x2EU)
+#define CTRL_G (0x2FU)
+#define CTRL_H (0x16U)
+#define CTRL_I (0x05U)
+#define CTRL_J (0x15U)
+#define CTRL_K (0x0BU)
+#define CTRL_L (0x0CU)
+#define CTRL_M (0x0DU)
+#define CTRL_N (0x0EU)
+#define CTRL_O (0x0FU)
+#define CTRL_P (0x10U)
+#define CTRL_Q (0x11U)
+#define CTRL_R (0x12U)
+#define CTRL_S (0x13U)
+#define CTRL_T (0x3CU)
+#define CTRL_U (0x3DU)
+#define CTRL_V (0x32U)
+#define CTRL_W (0x26U)
+#define CTRL_X (0x18U)
+#define CTRL_Y (0x19U)
+#define CTRL_Z (0x3FU)
+#define CTRL_BO (0x27U)
+#define CTRL_BK (0x1CU)
+#define CTRL_BC (0x1DU)
+#define CTRL_CA (0x1EU)
+#define CTRL_US (0x1FU)
+#define CTRL_QM (0x07U)
+#else
+#define CTRL_AT (0x00U)
+#define CTRL_A (0x01U)
+#define CTRL_B (0x02U)
+#define CTRL_C (0x03U)
+#define CTRL_D (0x04U)
+#define CTRL_E (0x05U)
+#define CTRL_F (0x06U)
+#define CTRL_G (0x07U)
+#define CTRL_H (0x08U)
+#define CTRL_I (0x09U)
+#define CTRL_J (0x0AU)
+#define CTRL_K (0x0BU)
+#define CTRL_L (0x0CU)
+#define CTRL_M (0x0DU)
+#define CTRL_N (0x0EU)
+#define CTRL_O (0x0FU)
+#define CTRL_P (0x10U)
+#define CTRL_Q (0x11U)
+#define CTRL_R (0x12U)
+#define CTRL_S (0x13U)
+#define CTRL_T (0x14U)
+#define CTRL_U (0x15U)
+#define CTRL_V (0x16U)
+#define CTRL_W (0x17U)
+#define CTRL_X (0x18U)
+#define CTRL_Y (0x19U)
+#define CTRL_Z (0x1AU)
+#define CTRL_BO (0x1BU)
+#define CTRL_BK (0x1CU)
+#define CTRL_BC (0x1DU)
+#define CTRL_CA (0x1EU)
+#define CTRL_US (0x1FU)
+#define CTRL_QM (0x7FU)
+#endif
-#define IDENT 64
+#define IDENT 64
EXTERN Source *source; /* yyparse/yylex source */
EXTERN YYSTYPE yylval; /* result from yylex */
@@ -2273,8 +2527,6 @@
MKSH_A_FORMAT(__printf__, 1, 2);
#endif
/* misc.c */
-void setctypes(const char *, int);
-void initctypes(void);
size_t option(const char *) MKSH_A_PURE;
char *getoptions(void);
void change_flag(enum sh_flag, int, bool);
@@ -2282,8 +2534,9 @@
int parse_args(const char **, int, bool *);
int getn(const char *, int *);
int gmatchx(const char *, const char *, bool);
-int has_globbing(const char *, const char *) MKSH_A_PURE;
-int xstrcmp(const void *, const void *) MKSH_A_PURE;
+bool has_globbing(const char *) MKSH_A_PURE;
+int ascstrcmp(const void *, const void *) MKSH_A_PURE;
+int ascpstrcmp(const void *, const void *) MKSH_A_PURE;
void ksh_getopt_reset(Getopt *, int);
int ksh_getopt(const char **, Getopt *, const char *);
void print_value_quoted(struct shf *, const char *);
@@ -2346,6 +2599,7 @@
MKSH_A_FORMAT(__printf__, 1, 2);
ssize_t shf_vfprintf(struct shf *, const char *, va_list)
MKSH_A_FORMAT(__printf__, 2, 0);
+void set_ifs(const char *);
/* syn.c */
void initkeywords(void);
struct op *compile(Source *, bool, bool);
@@ -2483,7 +2737,7 @@
#define mksh_abspath(s) __extension__({ \
const char *mksh_abspath_s = (s); \
(mksh_cdirsep(mksh_abspath_s[0]) || \
- (ksh_isalpha(mksh_abspath_s[0]) && \
+ (ctype(mksh_abspath_s[0], C_ALPHA) && \
mksh_abspath_s[1] == ':')); \
})
#define mksh_cdirsep(c) __extension__({ \
@@ -2492,15 +2746,15 @@
})
#define mksh_sdirsep(s) __extension__({ \
const char *mksh_sdirsep_s = (s); \
- ((char *)((ksh_isalphx(mksh_sdirsep_s[0]) && \
+ ((char *)((ctype(mksh_sdirsep_s[0], C_ALPHA) && \
mksh_sdirsep_s[1] == ':' && \
!mksh_cdirsep(mksh_sdirsep_s[2])) ? \
(mksh_sdirsep_s + 1) : strpbrk(mksh_sdirsep_s, "/\\"))); \
})
#define mksh_vdirsep(s) (mksh_sdirsep((s)) != NULL)
#else
-#define mksh_abspath(s) ((s)[0] == '/')
-#define mksh_cdirsep(c) ((c) == '/')
+#define mksh_abspath(s) (ord((s)[0]) == ord('/'))
+#define mksh_cdirsep(c) (ord(c) == ord('/'))
#define mksh_sdirsep(s) strchr((s), '/')
#define mksh_vdirsep(s) vstrchr((s), '/')
#endif
diff --git a/src/shf.c b/src/shf.c
index 09cc7c3..7e53352 100644
--- a/src/shf.c
+++ b/src/shf.c
@@ -4,6 +4,8 @@
* Copyright (c) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2011,
* 2012, 2013, 2015, 2016, 2017
* mirabilos <m@mirbsd.org>
+ * Copyright (c) 2015
+ * Daniel Richard G. <skunk@iSKUNK.ORG>
*
* Provided that these terms and disclaimer and all copyright notices
* are retained or reproduced in an accompanying document, permission
@@ -25,7 +27,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/shf.c,v 1.79 2017/04/12 17:08:49 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/shf.c,v 1.95 2017/05/05 22:45:58 tg Exp $");
/* flags to shf_emptybuf() */
#define EB_READSW 0x01 /* about to switch to reading */
@@ -874,11 +876,11 @@
flags |= FL_SIZET;
continue;
}
- if (ksh_isdigit(c)) {
+ if (ctype(c, C_DIGIT)) {
bool overflowed = false;
tmp = ksh_numdig(c);
- while (c = *fmt++, ksh_isdigit(c))
+ while (ctype((c = *fmt++), C_DIGIT))
if (notok2mul(2147483647, tmp, 10))
overflowed = true;
else
@@ -899,7 +901,7 @@
/* nasty format */
break;
- if (ksh_isupper(c)) {
+ if (ctype(c, C_UPPER)) {
flags |= FL_UPPER;
c = ksh_tolower(c);
}
@@ -1029,8 +1031,7 @@
if (!(flags & FL_RIGHT)) {
/* skip past sign or 0x when padding with 0 */
if ((flags & FL_ZERO) && (flags & FL_NUMBER)) {
- if (*s == '+' || *s == '-' ||
- *s == ' ') {
+ if (ctype(*s, C_SPC | C_PLUS | C_MINUS)) {
shf_putc(*s, shf);
s++;
precision--;
@@ -1158,3 +1159,163 @@
}
}
#endif
+
+/* fast character classes */
+const uint32_t tpl_ctypes[128] = {
+ /* 0x00 */
+ CiNUL, CiCNTRL, CiCNTRL, CiCNTRL,
+ CiCNTRL, CiCNTRL, CiCNTRL, CiCNTRL,
+ CiCNTRL, CiTAB, CiNL, CiSPX,
+ CiSPX, CiCR, CiCNTRL, CiCNTRL,
+ /* 0x10 */
+ CiCNTRL, CiCNTRL, CiCNTRL, CiCNTRL,
+ CiCNTRL, CiCNTRL, CiCNTRL, CiCNTRL,
+ CiCNTRL, CiCNTRL, CiCNTRL, CiCNTRL,
+ CiCNTRL, CiCNTRL, CiCNTRL, CiCNTRL,
+ /* 0x20 */
+ CiSP, CiALIAS | CiVAR1, CiQC, CiHASH,
+ CiSS, CiPERCT, CiQCL, CiQC,
+ CiQCL, CiQCL, CiQCX | CiVAR1, CiPLUS,
+ CiALIAS, CiMINUS, CiALIAS, CiQCM,
+ /* 0x30 */
+ CiOCTAL, CiOCTAL, CiOCTAL, CiOCTAL,
+ CiOCTAL, CiOCTAL, CiOCTAL, CiOCTAL,
+ CiDIGIT, CiDIGIT, CiCOLON, CiQCL,
+ CiANGLE, CiEQUAL, CiANGLE, CiQUEST,
+ /* 0x40 */
+ CiALIAS | CiVAR1, CiUPPER | CiHEXLT,
+ CiUPPER | CiHEXLT, CiUPPER | CiHEXLT,
+ CiUPPER | CiHEXLT, CiUPPER | CiHEXLT,
+ CiUPPER | CiHEXLT, CiUPPER,
+ CiUPPER, CiUPPER, CiUPPER, CiUPPER,
+ CiUPPER, CiUPPER, CiUPPER, CiUPPER,
+ /* 0x50 */
+ CiUPPER, CiUPPER, CiUPPER, CiUPPER,
+ CiUPPER, CiUPPER, CiUPPER, CiUPPER,
+ CiUPPER, CiUPPER, CiUPPER, CiQCX | CiBRACK,
+ CiQCX, CiBRACK, CiQCM, CiUNDER,
+ /* 0x60 */
+ CiGRAVE, CiLOWER | CiHEXLT,
+ CiLOWER | CiHEXLT, CiLOWER | CiHEXLT,
+ CiLOWER | CiHEXLT, CiLOWER | CiHEXLT,
+ CiLOWER | CiHEXLT, CiLOWER,
+ CiLOWER, CiLOWER, CiLOWER, CiLOWER,
+ CiLOWER, CiLOWER, CiLOWER, CiLOWER,
+ /* 0x70 */
+ CiLOWER, CiLOWER, CiLOWER, CiLOWER,
+ CiLOWER, CiLOWER, CiLOWER, CiLOWER,
+ CiLOWER, CiLOWER, CiLOWER, CiCURLY,
+ CiQCL, CiCURLY, CiQCM, CiCNTRL
+};
+
+void
+set_ifs(const char *s)
+{
+#if defined(MKSH_EBCDIC) || defined(MKSH_FAUX_EBCDIC)
+ int i = 256;
+
+ memset(ksh_ctypes, 0, sizeof(ksh_ctypes));
+ while (i--)
+ if (ebcdic_map[i] < 0x80U)
+ ksh_ctypes[i] = tpl_ctypes[ebcdic_map[i]];
+#else
+ memcpy(ksh_ctypes, tpl_ctypes, sizeof(tpl_ctypes));
+ memset((char *)ksh_ctypes + sizeof(tpl_ctypes), '\0',
+ sizeof(ksh_ctypes) - sizeof(tpl_ctypes));
+#endif
+ ifs0 = *s;
+ while (*s)
+ ksh_ctypes[ord(*s++)] |= CiIFS;
+}
+
+#if defined(MKSH_EBCDIC) || defined(MKSH_FAUX_EBCDIC)
+#include <locale.h>
+
+/*
+ * Many headaches with EBCDIC:
+ * 1. There are numerous EBCDIC variants, and it is not feasible for us
+ * to support them all. But we can support the EBCDIC code pages that
+ * contain all (most?) of the characters in ASCII, and these
+ * usually tend to agree on the code points assigned to the ASCII
+ * subset. If you need a representative example, look at EBCDIC 1047,
+ * which is first among equals in the IBM MVS development
+ * environment: https://en.wikipedia.org/wiki/EBCDIC_1047
+ * Unfortunately, the square brackets are not consistently mapped,
+ * and for certain reasons, we need an unambiguous bijective
+ * mapping between EBCDIC and "extended ASCII".
+ * 2. Character ranges that are contiguous in ASCII, like the letters
+ * in [A-Z], are broken up into segments (i.e. [A-IJ-RS-Z]), so we
+ * can't implement e.g. islower() as { return c >= 'a' && c <= 'z'; }
+ * because it will also return true for a handful of extraneous
+ * characters (like the plus-minus sign at 0x8F in EBCDIC 1047, a
+ * little after 'i'). But at least '_' is not one of these.
+ * 3. The normal [0-9A-Za-z] characters are at codepoints beyond 0x80.
+ * Not only do they require all 8 bits instead of 7, if chars are
+ * signed, they will have negative integer values! Something like
+ * (c - 'A') could actually become (c + 63)! Use the ord() macro to
+ * ensure you're getting a value in [0, 255].
+ * 4. '\n' is actually NL (0x15, U+0085) instead of LF (0x25, U+000A).
+ * EBCDIC has a proper newline character instead of "emulating" one
+ * with line feeds, although this is mapped to LF for our purposes.
+ * 5. Note that it is possible to compile programs in ASCII mode on IBM
+ * mainframe systems, using the -qascii option to the XL C compiler.
+ * We can determine the build mode by looking at __CHARSET_LIB:
+ * 0 == EBCDIC, 1 == ASCII
+ */
+
+void
+ebcdic_init(void)
+{
+ int i = 256;
+ unsigned char t;
+ bool mapcache[256];
+
+ while (i--)
+ ebcdic_rtt_toascii[i] = i;
+ memset(ebcdic_rtt_fromascii, 0xFF, sizeof(ebcdic_rtt_fromascii));
+ setlocale(LC_ALL, "");
+#ifdef MKSH_EBCDIC
+ if (__etoa_l(ebcdic_rtt_toascii, 256) != 256) {
+ write(2, "mksh: could not map EBCDIC to ASCII\n", 36);
+ exit(255);
+ }
+#endif
+
+ memset(mapcache, 0, sizeof(mapcache));
+ i = 256;
+ while (i--) {
+ t = ebcdic_rtt_toascii[i];
+ /* ensure unique round-trip capable mapping */
+ if (mapcache[t]) {
+ write(2, "mksh: duplicate EBCDIC to ASCII mapping\n", 40);
+ exit(255);
+ }
+ /*
+ * since there are 256 input octets, this also ensures
+ * the other mapping direction is completely filled
+ */
+ mapcache[t] = true;
+ /* fill the complete round-trip map */
+ ebcdic_rtt_fromascii[t] = i;
+ /*
+ * Only use the converted value if it's in the range
+ * [0x00; 0x7F], which I checked; the "extended ASCII"
+ * characters can be any encoding, not just Latin1,
+ * and the C1 control characters other than NEL are
+ * hopeless, but we map EBCDIC NEL to ASCII LF so we
+ * cannot even use C1 NEL.
+ * If ever we map to Unicode, bump the table width to
+ * an unsigned int, and or the raw unconverted EBCDIC
+ * values with 0x01000000 instead.
+ */
+ if (t < 0x80U)
+ ebcdic_map[i] = (unsigned short)ord(t);
+ else
+ ebcdic_map[i] = (unsigned short)(0x100U | ord(i));
+ }
+ if (ebcdic_rtt_toascii[0] || ebcdic_rtt_fromascii[0] || ebcdic_map[0]) {
+ write(2, "mksh: NUL not at position 0\n", 28);
+ exit(255);
+ }
+}
+#endif
diff --git a/src/signames.inc b/src/signames.inc
deleted file mode 100644
index 07811fd..0000000
--- a/src/signames.inc
+++ /dev/null
@@ -1,31 +0,0 @@
- { "ABRT", 6 },
- { "FPE", 8 },
- { "ILL", 4 },
- { "INT", 2 },
- { "SEGV", 11 },
- { "TERM", 15 },
- { "ALRM", 14 },
- { "BUS", 7 },
- { "CHLD", 17 },
- { "CONT", 18 },
- { "HUP", 1 },
- { "KILL", 9 },
- { "PIPE", 13 },
- { "QUIT", 3 },
- { "STOP", 19 },
- { "TSTP", 20 },
- { "TTIN", 21 },
- { "TTOU", 22 },
- { "USR1", 10 },
- { "USR2", 12 },
- { "POLL", 29 },
- { "PROF", 27 },
- { "SYS", 31 },
- { "TRAP", 5 },
- { "URG", 23 },
- { "VTALRM", 26 },
- { "XCPU", 24 },
- { "XFSZ", 25 },
- { "WINCH", 28 },
- { "PWR", 30 },
- { "STKFLT", 16 },
diff --git a/src/syn.c b/src/syn.c
index 0454488..c50c2ab 100644
--- a/src/syn.c
+++ b/src/syn.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/syn.c,v 1.120 2017/04/06 01:59:57 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/syn.c,v 1.124 2017/05/05 22:53:31 tg Exp $");
struct nesting_state {
int start_token; /* token than began nesting (eg, FOR) */
@@ -91,7 +91,7 @@
c = tpeek(0);
if (c == 0 && !outtree)
outtree = newtp(TEOF);
- else if (c != '\n' && c != 0)
+ else if (!ctype(c, C_LF | C_NUL))
syntaxerr(NULL);
}
@@ -330,7 +330,7 @@
XPput(args, yylval.cp);
break;
- case '(' /*)*/:
+ case ord('(' /*)*/):
if (XPsize(args) == 0 && XPsize(vars) == 1 &&
is_wdvarassign(yylval.cp)) {
char *tcp;
@@ -373,7 +373,7 @@
XPsize(vars) != 0)
syntaxerr(NULL);
ACCEPT;
- musthave(/*(*/')', 0);
+ musthave(/*(*/ ')', 0);
t = function_body(XPptrv(args)[0],
sALIAS, false);
}
@@ -386,18 +386,18 @@
Leave:
break;
- case '(': /*)*/ {
+ case ord('(' /*)*/): {
int subshell_nesting_type_saved;
Subshell:
subshell_nesting_type_saved = subshell_nesting_type;
- subshell_nesting_type = ')';
- t = nested(TPAREN, '(', ')', sALIAS);
+ subshell_nesting_type = ord(')');
+ t = nested(TPAREN, ord('('), ord(')'), sALIAS);
subshell_nesting_type = subshell_nesting_type_saved;
break;
}
- case '{': /*}*/
- t = nested(TBRACE, '{', '}', sALIAS);
+ case ord('{' /*}*/):
+ t = nested(TBRACE, ord('{'), ord('}'), sALIAS);
break;
case MDPAREN:
@@ -407,8 +407,8 @@
switch (token(LETEXPR)) {
case LWORD:
break;
- case '(': /*)*/
- c = '(';
+ case ord('(' /*)*/):
+ c = ord('(');
goto Subshell;
default:
syntaxerr(NULL);
@@ -554,8 +554,8 @@
*/
if (c == DO)
c = DONE;
- else if (c == '{')
- c = '}';
+ else if (c == ord('{'))
+ c = ord('}');
else
syntaxerr(NULL);
list = c_list(sALIAS, true);
@@ -610,8 +610,8 @@
/* A {...} can be used instead of in...esac for case statements */
if (c == IN)
c = ESAC;
- else if (c == '{')
- c = '}';
+ else if (c == ord('{'))
+ c = ord('}');
else
syntaxerr(NULL);
t = tl = NULL;
@@ -636,17 +636,18 @@
XPinit(ptns, 16);
t = newtp(TPAT);
/* no ALIAS here */
- if (token(CONTIN | KEYWORD) != '(')
+ if (token(CONTIN | KEYWORD) != ord('('))
REJECT;
do {
switch (token(0)) {
case LWORD:
break;
- case '}':
+ case ord('}'):
case ESAC:
if (symbol != endtok) {
strdupx(yylval.cp,
- symbol == '}' ? Tcbrace : Tesac, ATEMP);
+ symbol == ord('}') ? Tcbrace : Tesac,
+ ATEMP);
break;
}
/* FALLTHROUGH */
@@ -658,23 +659,23 @@
REJECT;
XPput(ptns, NULL);
t->vars = (char **)XPclose(ptns);
- musthave(')', 0);
+ musthave(ord(')'), 0);
t->left = c_list(sALIAS, true);
/* initialise to default for ;; or omitted */
- t->u.charflag = ';';
+ t->u.charflag = ord(';');
/* SUSv4 requires the ;; except in the last casepart */
if ((tpeek(CONTIN|KEYWORD|sALIAS)) != endtok)
switch (symbol) {
default:
syntaxerr(NULL);
case BRKEV:
- t->u.charflag = '|';
+ t->u.charflag = ord('|');
if (0)
/* FALLTHROUGH */
case BRKFT:
- t->u.charflag = '&';
+ t->u.charflag = ord('&');
/* FALLTHROUGH */
case BREAK:
/* initialised above, but we need to eat the token */
@@ -697,10 +698,10 @@
* only allow [a-zA-Z_0-9] but this allows more as old pdkshs
* have allowed more; the following were never allowed:
* NUL TAB NL SP " $ & ' ( ) ; < = > \ ` |
- * C_QUOTE covers all but adds # * ? [ ]
+ * C_QUOTE|C_SPC covers all but adds # * ? [ ]
*/
for (p = sname; *p; p++)
- if (ctype(*p, C_QUOTE))
+ if (ctype(*p, C_QUOTE | C_SPC))
yyerror(Tinvname, sname, Tfunction);
/*
@@ -710,14 +711,14 @@
* only accepts an open-brace.
*/
if (ksh_func) {
- if (tpeek(CONTIN|KEYWORD|sALIAS) == '(' /*)*/) {
+ if (tpeek(CONTIN|KEYWORD|sALIAS) == ord('(' /*)*/)) {
/* function foo () { //}*/
ACCEPT;
- musthave(')', 0);
+ musthave(ord(/*(*/ ')'), 0);
/* degrade to POSIX function */
ksh_func = false;
}
- musthave('{' /*}*/, CONTIN|KEYWORD|sALIAS);
+ musthave(ord('{' /*}*/), CONTIN|KEYWORD|sALIAS);
REJECT;
}
@@ -809,8 +810,8 @@
{ "in", IN, true },
{ Tfunction, FUNCTION, true },
{ Ttime, TIME, true },
- { "{", '{', true },
- { Tcbrace, '}', true },
+ { "{", ord('{'), true },
+ { Tcbrace, ord('}'), true },
{ "!", BANG, true },
{ "[[", DBRACKET, true },
/* Lexical tokens (0[EOF], LWORD and REDIR handled specially) */
@@ -822,7 +823,7 @@
{ "((", MDPAREN, false },
{ "|&", COPROC, false },
/* and some special cases... */
- { "newline", '\n', false },
+ { "newline", ord('\n'), false },
{ NULL, 0, false }
};
@@ -997,9 +998,9 @@
ret = (uqword && !strcmp(yylval.cp,
dbtest_tokens[(int)TM_NOT])) ? TO_NONNULL : TO_NONOP;
else if (meta == TM_OPAREN)
- ret = c == '(' /*)*/ ? TO_NONNULL : TO_NONOP;
+ ret = c == ord('(') /*)*/ ? TO_NONNULL : TO_NONOP;
else if (meta == TM_CPAREN)
- ret = c == /*(*/ ')' ? TO_NONNULL : TO_NONOP;
+ ret = c == /*(*/ ord(')') ? TO_NONNULL : TO_NONOP;
else if (meta == TM_UNOP || meta == TM_BINOP) {
if (meta == TM_BINOP && c == REDIR &&
(yylval.iop->ioflag == IOREAD ||
@@ -1079,7 +1080,7 @@
tv->tv_sec = 0;
/* parse integral part */
- while (ksh_isdigit(*s)) {
+ while (ctype(*s, C_DIGIT)) {
tt.tv_sec = tv->tv_sec * 10 + ksh_numdig(*s++);
/*XXX this overflow check maybe UB */
if (tt.tv_sec / 10 != tv->tv_sec) {
@@ -1101,14 +1102,14 @@
/* parse decimal fraction */
i = 100000;
- while (ksh_isdigit(*s)) {
+ while (ctype(*s, C_DIGIT)) {
tv->tv_usec += i * ksh_numdig(*s++);
if (i == 1)
break;
i /= 10;
}
/* check for junk after fractional part */
- while (ksh_isdigit(*s))
+ while (ctype(*s, C_DIGIT))
++s;
if (*s) {
errno = EINVAL;
@@ -1133,11 +1134,11 @@
int stok, etok;
if (subtype != COMSUB) {
- stok = '{';
- etok = '}';
+ stok = ord('{');
+ etok = ord('}');
} else {
- stok = '(';
- etok = ')';
+ stok = ord('(');
+ etok = ord(')');
}
ys = alloc(sizeof(struct yyrecursive_state), ATEMP);
diff --git a/src/tree.c b/src/tree.c
index 1fd8f2a..1062feb 100644
--- a/src/tree.c
+++ b/src/tree.c
@@ -23,7 +23,7 @@
#include "sh.h"
-__RCSID("$MirOS: src/bin/mksh/tree.c,v 1.89 2017/04/12 16:46:23 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/tree.c,v 1.93 2017/05/05 22:53:32 tg Exp $");
#define INDENT 8
@@ -329,34 +329,34 @@
case EOS:
return (--wp);
case ADELIM:
- if (*wp == /*{*/'}') {
+ if (ord(*wp) == ord(/*{*/ '}')) {
++wp;
goto wdvarput_csubst;
}
/* FALLTHROUGH */
case CHAR:
- c = *wp++;
+ c = ord(*wp++);
shf_putc(c, shf);
break;
case QCHAR:
- c = *wp++;
+ c = ord(*wp++);
if (opmode & WDS_TPUTS)
switch (c) {
- case '\n':
+ case ord('\n'):
if (quotelevel == 0) {
- c = '\'';
+ c = ord('\'');
shf_putc(c, shf);
- shf_putc('\n', shf);
+ shf_putc(ord('\n'), shf);
}
break;
default:
if (quotelevel == 0)
/* FALLTHROUGH */
- case '"':
- case '`':
- case '$':
- case '\\':
- shf_putc('\\', shf);
+ case ord('"'):
+ case ord('`'):
+ case ord('$'):
+ case ord('\\'):
+ shf_putc(ord('\\'), shf);
break;
}
shf_putc(c, shf);
@@ -365,7 +365,7 @@
case COMSUB:
shf_puts("$(", shf);
cs = ")";
- if (*wp == '(' /*)*/)
+ if (ord(*wp) == ord('(' /*)*/))
shf_putc(' ', shf);
pSUB:
while ((c = *wp++) != 0)
@@ -374,11 +374,11 @@
break;
case FUNASUB:
case FUNSUB:
- c = ' ';
+ c = ord(' ');
if (0)
/* FALLTHROUGH */
case VALSUB:
- c = '|';
+ c = ord('|');
shf_putc('$', shf);
shf_putc('{', shf);
shf_putc(c, shf);
@@ -403,14 +403,14 @@
break;
case OSUBST:
shf_putc('$', shf);
- if (*wp++ == '{')
+ if (ord(*wp++) == ord('{'))
shf_putc('{', shf);
while ((c = *wp++) != 0)
shf_putc(c, shf);
wp = wdvarput(shf, wp, 0, opmode);
break;
case CSUBST:
- if (*wp++ == '}') {
+ if (ord(*wp++) == ord('}')) {
wdvarput_csubst:
shf_putc('}', shf);
}
@@ -420,11 +420,11 @@
shf_putc('(', shf);
break;
case SPAT:
- c = '|';
+ c = ord('|');
if (0)
/* FALLTHROUGH */
case CPAT:
- c = /*(*/ ')';
+ c = ord(/*(*/ ')');
shf_putc(c, shf);
break;
}
@@ -467,39 +467,39 @@
{
int c;
- while ((c = *fmt++)) {
+ while ((c = ord(*fmt++))) {
if (c == '%') {
- switch ((c = *fmt++)) {
- case 'c':
+ switch ((c = ord(*fmt++))) {
+ case ord('c'):
/* character (octet, probably) */
shf_putchar(va_arg(va, int), shf);
break;
- case 's':
+ case ord('s'):
/* string */
shf_puts(va_arg(va, char *), shf);
break;
- case 'S':
+ case ord('S'):
/* word */
wdvarput(shf, va_arg(va, char *), 0, WDS_TPUTS);
break;
- case 'd':
+ case ord('d'):
/* signed decimal */
shf_fprintf(shf, Tf_d, va_arg(va, int));
break;
- case 'u':
+ case ord('u'):
/* unsigned decimal */
shf_fprintf(shf, "%u", va_arg(va, unsigned int));
break;
- case 'T':
+ case ord('T'):
/* format tree */
ptree(va_arg(va, struct op *), indent, shf);
goto dont_trash_prevent_semicolon;
- case ';':
+ case ord(';'):
/* newline or ; */
- case 'N':
+ case ord('N'):
/* newline or space */
if (shf->flags & SHF_STRING) {
- if (c == ';' && !prevent_semicolon)
+ if (c == ord(';') && !prevent_semicolon)
shf_putc(';', shf);
shf_putc(' ', shf);
} else {
@@ -515,7 +515,7 @@
shf_putc(' ', shf);
}
break;
- case 'R':
+ case ord('R'):
/* I/O redirection */
pioact(shf, va_arg(va, struct ioword *));
break;
@@ -613,7 +613,7 @@
case ADELIM:
if (c == ADELIM && nest == 0)
return (wp + 1);
- if (*wp == /*{*/'}')
+ if (ord(*wp) == ord(/*{*/ '}'))
goto wdscan_csubst;
/* FALLTHROUGH */
case CHAR:
@@ -795,20 +795,20 @@
*dst++ = *cp++;
goto vist_loop;
}
- if (--sz == 0 || (c = (unsigned char)(*cp++)) == 0)
+ if (--sz == 0 || (c = ord(*cp++)) == 0)
/* NUL or not enough free space */
goto vist_out;
- if (ISCTRL(c & 0x7F)) {
+ if (ksh_isctrl(c)) {
/* C0 or C1 control character or DEL */
if (--sz == 0)
/* not enough free space for two chars */
goto vist_out;
- *dst++ = (c & 0x80) ? '$' : '^';
- c = UNCTRL(c & 0x7F);
- } else if (UTFMODE && c > 0x7F) {
+ *dst++ = '^';
+ c = ksh_unctrl(c);
+ } else if (UTFMODE && rtt2asc(c) > 0x7F) {
/* better not try to display broken multibyte chars */
/* also go easy on the Unicode: no U+FFFD here */
- c = '?';
+ c = ord('?');
}
*dst++ = c;
goto vist_loop;
@@ -822,10 +822,10 @@
void
dumpchar(struct shf *shf, int c)
{
- if (ISCTRL(c & 0x7F)) {
+ if (ksh_isctrl(c)) {
/* C0 or C1 control character or DEL */
- shf_putc((c & 0x80) ? '$' : '^', shf);
- c = UNCTRL(c & 0x7F);
+ shf_putc('^', shf);
+ c = ksh_unctrl(c);
}
shf_putc(c, shf);
}
@@ -842,7 +842,7 @@
shf_puts("EOS", shf);
return (--wp);
case ADELIM:
- if (*wp == /*{*/'}') {
+ if (ord(*wp) == ord(/*{*/ '}')) {
shf_puts(/*{*/ "]ADELIM(})", shf);
return (wp + 1);
}
@@ -855,9 +855,9 @@
break;
case QCHAR:
shf_puts("QCHAR<", shf);
- c = *wp++;
- if (quotelevel == 0 ||
- (c == '"' || c == '`' || c == '$' || c == '\\'))
+ c = ord(*wp++);
+ if (quotelevel == 0 || c == ord('"') ||
+ c == ord('\\') || ctype(c, C_DOLAR | C_GRAVE))
shf_putc('\\', shf);
dumpchar(shf, c);
goto closeandout;
diff --git a/src/var.c b/src/var.c
index b83977f..a53fae8 100644
--- a/src/var.c
+++ b/src/var.c
@@ -28,7 +28,7 @@
#include <sys/sysctl.h>
#endif
-__RCSID("$MirOS: src/bin/mksh/var.c,v 1.214 2017/04/02 16:47:43 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/var.c,v 1.220 2017/07/26 23:02:28 tg Exp $");
/*-
* Variables
@@ -183,7 +183,7 @@
*arrayp = false;
redo_from_ref:
p = skip_varname(n, false);
- if (innermost_refflag == SRF_NOP && (p != n) && ksh_isalphx(n[0])) {
+ if (innermost_refflag == SRF_NOP && (p != n) && ctype(n[0], C_ALPHX)) {
struct tbl *vp;
char *vn;
@@ -204,7 +204,7 @@
}
innermost_refflag = SRF_NOP;
- if (p != n && *p == '[' && (len = array_ref_len(p))) {
+ if (p != n && ord(*p) == ord('[') && (len = array_ref_len(p))) {
char *sub, *tmp;
mksh_ari_t rval;
@@ -249,14 +249,14 @@
vn = array_index_calc(n, &array, &val);
h = hash(vn);
c = (unsigned char)vn[0];
- if (!ksh_isalphx(c)) {
+ if (!ctype(c, C_ALPHX)) {
if (array)
errorf(Tbadsubst);
vp = vtemp;
vp->flag = DEFINED;
vp->type = 0;
vp->areap = ATEMP;
- if (ksh_isdigit(c)) {
+ if (ctype(c, C_DIGIT)) {
if (getn(vn, &c)) {
/* main.c:main_init() says 12 */
shf_snprintf(vp->name, 12, Tf_d, c);
@@ -339,7 +339,7 @@
*/
vn = array_index_calc(n, &array, &val);
h = hash(vn);
- if (!ksh_isalphx(*vn)) {
+ if (!ctype(*vn, C_ALPHX)) {
vp = vtemp;
vp->flag = DEFINED|RDONLY;
vp->type = 0;
@@ -414,9 +414,11 @@
*(s = strbuf) = '1';
s[1] = '#';
- if (!UTFMODE || ((n & 0xFF80) == 0xEF80))
+ if (!UTFMODE)
+ s[2] = (unsigned char)n;
+ else if ((n & 0xFF80) == 0xEF80)
/* OPTU-16 -> raw octet */
- s[2] = n & 0xFF;
+ s[2] = asc2rtt(n & 0xFF);
else
sz = utf_wctomb(s + 2, n);
s[2 + sz] = '\0';
@@ -464,7 +466,7 @@
#ifndef MKSH_SMALL
/* debugging */
if (s >= vq->val.s &&
- s <= vq->val.s + strlen(vq->val.s)) {
+ s <= strnul(vq->val.s)) {
internal_errorf(
"setstr: %s=%s: assigning to self",
vq->name, s);
@@ -532,7 +534,7 @@
do {
c = (unsigned char)*s++;
- } while (ksh_isspace(c));
+ } while (ctype(c, C_SPACE));
switch (c) {
case '-':
@@ -549,7 +551,7 @@
base = 16;
++s;
goto getint_c_style_base;
- } else if (Flag(FPOSIX) && ksh_isdigit(s[0]) &&
+ } else if (Flag(FPOSIX) && ctype(s[0], C_DIGIT) &&
!(vp->flag & ZEROFIL)) {
/* interpret as octal (deprecated) */
base = 8;
@@ -577,7 +579,7 @@
* the same as 1#\x80 does, thus is
* not round-tripping correctly XXX)
*/
- wc = 0xEF00 + *(const unsigned char *)s;
+ wc = 0xEF00 + rtt2asc(*s);
nump->u = (mksh_uari_t)wc;
return (1);
} else if (base > 36)
@@ -586,11 +588,11 @@
have_base = true;
continue;
}
- if (ksh_isdigit(c))
+ if (ctype(c, C_DIGIT))
c = ksh_numdig(c);
- else if (ksh_isupper(c))
+ else if (ctype(c, C_UPPER))
c = ksh_numuc(c) + 10;
- else if (ksh_islower(c))
+ else if (ctype(c, C_LOWER))
c = ksh_numlc(c) + 10;
else
return (-1);
@@ -670,7 +672,7 @@
qq = utf_skipcols(s, slen, &slen);
/* strip trailing spaces (AT&T uses qq[-1] == ' ') */
- while (qq > s && ksh_isspace(qq[-1])) {
+ while (qq > s && ctype(qq[-1], C_SPACE)) {
--qq;
--slen;
}
@@ -700,7 +702,7 @@
"%.*s", slen, s);
} else {
/* strip leading spaces/zeros */
- while (ksh_isspace(*s))
+ while (ctype(*s, C_SPACE))
s++;
if (vp->flag & ZEROFIL)
while (*s == '0')
@@ -778,7 +780,7 @@
/* no variable name given */
return (NULL);
}
- if (*val == '[') {
+ if (ord(*val) == ord('[')) {
if (new_refflag != SRF_NOP)
errorf(Tf_sD_s, var,
"reference variable can't be an array");
@@ -796,18 +798,18 @@
size_t i;
for (i = 1; i < len - 1; i++)
- if (!ksh_isdigit(val[i]))
+ if (!ctype(val[i], C_DIGIT))
return (NULL);
}
val += len;
}
- if (val[0] == '=') {
+ if (ord(val[0]) == ord('=')) {
strndupx(tvar, var, val - var, ATEMP);
++val;
} else if (set & IMPORT) {
/* environment invalid variable name or no assignment */
return (NULL);
- } else if (val[0] == '+' && val[1] == '=') {
+ } else if (ord(val[0]) == ord('+') && ord(val[1]) == ord('=')) {
strndupx(tvar, var, val - var, ATEMP);
val += 2;
vappend = true;
@@ -820,8 +822,9 @@
val = NULL;
/* handle foo[*] => foo (whole array) mapping for R39b */
len = strlen(tvar);
- if (len > 3 && tvar[len - 3] == '[' && tvar[len - 2] == '*' &&
- tvar[len - 1] == ']')
+ if (len > 3 && ord(tvar[len - 3]) == ord('[') &&
+ ord(tvar[len - 2]) == ord('*') &&
+ ord(tvar[len - 1]) == ord(']'))
tvar[len - 3] = '\0';
}
@@ -845,7 +848,7 @@
if (!(c = (unsigned char)qval[0]))
goto nameref_empty;
- else if (ksh_isdigit(c) && getn(qval, &c))
+ else if (ctype(c, C_DIGIT) && getn(qval, &c))
goto nameref_rhs_checked;
else if (qval[1] == '\0') switch (c) {
case '$':
@@ -858,7 +861,7 @@
nameref_empty:
errorf(Tf_sD_s, var, "empty nameref target");
}
- len = (*ccp == '[') ? array_ref_len(ccp) : 0;
+ len = (ord(*ccp) == ord('[')) ? array_ref_len(ccp) : 0;
if (ccp[len]) {
/*
* works for cases "no array", "valid array with
@@ -914,12 +917,12 @@
vpbase = (vp->flag & ARRAY) ? global(arrayname(tvar)) : vp;
/*
- * only allow export flag to be set; AT&T ksh allows any
- * attribute to be changed which means it can be truncated or
- * modified (-L/-R/-Z/-i)
+ * only allow export and readonly flag to be set; AT&T ksh
+ * allows any attribute to be changed which means it can be
+ * truncated or modified (-L/-R/-Z/-i)
*/
if ((vpbase->flag & RDONLY) &&
- (val || clr || (set & ~EXPORT)))
+ (val || clr || (set & ~(EXPORT | RDONLY))))
/* XXX check calls - is error here ok by POSIX? */
errorfx(2, Tf_ro, tvar);
afree(tvar, ATEMP);
@@ -1064,11 +1067,11 @@
{
size_t alen;
- if (s && ksh_isalphx(*s)) {
+ if (s && ctype(*s, C_ALPHX)) {
do {
++s;
- } while (ksh_isalnux(*s));
- if (aok && *s == '[' && (alen = array_ref_len(s)))
+ } while (ctype(*s, C_ALNUX));
+ if (aok && ord(*s) == ord('[') && (alen = array_ref_len(s)))
s += alen;
}
return (s);
@@ -1080,11 +1083,11 @@
/* skip array de-reference? */
bool aok)
{
- if (s[0] == CHAR && ksh_isalphx(s[1])) {
+ if (s[0] == CHAR && ctype(s[1], C_ALPHX)) {
do {
s += 2;
- } while (s[0] == CHAR && ksh_isalnux(s[1]));
- if (aok && s[0] == CHAR && s[1] == '[') {
+ } while (s[0] == CHAR && ctype(s[1], C_ALNUX));
+ if (aok && s[0] == CHAR && ord(s[1]) == ord('[')) {
/* skip possible array de-reference */
const char *p = s;
char c;
@@ -1095,9 +1098,9 @@
break;
c = p[1];
p += 2;
- if (c == '[')
+ if (ord(c) == ord('['))
depth++;
- else if (c == ']' && --depth == 0) {
+ else if (ord(c) == ord(']') && --depth == 0) {
s = p;
break;
}
@@ -1307,8 +1310,7 @@
return;
#endif
case V_IFS:
- setctypes(s = str_val(vp), C_IFS);
- ifs0 = *s;
+ set_ifs(str_val(vp));
return;
case V_PATH:
afree(path, APERM);
@@ -1436,8 +1438,7 @@
return;
#endif
case V_IFS:
- setctypes(TC_IFSWS, C_IFS);
- ifs0 = ' ';
+ set_ifs(TC_IFSWS);
break;
case V_PATH:
afree(path, APERM);
@@ -1527,8 +1528,8 @@
char c;
int depth = 0;
- while ((c = *s++) && (c != ']' || --depth))
- if (c == '[')
+ while ((c = *s++) && (ord(c) != ord(']') || --depth))
+ if (ord(c) == ord('['))
depth++;
if (!c)
return (0);
@@ -1600,17 +1601,18 @@
}
while ((ccp = vals[i])) {
#if 0 /* temporarily taken out due to regression */
- if (*ccp == '[') {
+ if (ord(*ccp) == ord('[')) {
int level = 0;
while (*ccp) {
- if (*ccp == ']' && --level == 0)
+ if (ord(*ccp) == ord(']') && --level == 0)
break;
- if (*ccp == '[')
+ if (ord(*ccp) == ord('['))
++level;
++ccp;
}
- if (*ccp == ']' && level == 0 && ccp[1] == '=') {
+ if (ord(*ccp) == ord(']') && level == 0 &&
+ ord(ccp[1]) == ord('=')) {
strndupx(cp, vals[i] + 1, ccp - (vals[i] + 1),
ATEMP);
evaluate(substitute(cp, 0), (mksh_ari_t *)&j,
