mksh R40 Release Candidate 1

Testsuite:
• add new need-pass: {yes|no} attribute, default yes
• exit with 1 if a need-pass test failed unexpectedly
  idea by Kacper Kornet <draenog@pld-linux.org>
• mark utf8bom-2 as need-pass: no
Infrstructure:
• add housekeeping function for making a tty raw
• switch functions with unused results to void
• struct op: u.charflag contains last char of ;; in TPAT
• var.c:arraysearch is now a global function
Language:
• add ;& (fall through) and ;| (examine next) delimiters
  in addition to ;; (end case) as zsh extensions, because
  POSIX standardised on ;& already
• add -A (read into array), -N (read exactly n bytes),
  -n (read up to n bytes), -t (timeout) flags for read
  from ksh93
• allow read -N -1 or -n -1 to slurp the entire input
• add -a (read into array the input characters) extension
  specific to mksh to read, idea by David Korn
• add -e (exit with error if PWD was not set correctly
  after a physical cd) to cd builtin, mandated by next
  POSIX, and change error codes accordingly
Rewrites:
• full rewrite of read builtin and its manpage section
• add regression tetss for most of the new functionality
• duplicate hexdump demo tests for use of read -a
• use read -raN-1 in dot.mkshrc to get NUL safe base64,
  DJB cdb hash and Jenkins one-at-a-time hash functions

check.pl

@@ -1,4 +1,4 @@
-# $MirOS: src/bin/mksh/check.pl,v 1.26 2011/03/28 21:58:05 tg Exp $
+# $MirOS: src/bin/mksh/check.pl,v 1.27 2011/05/29 02:18:47 tg Exp $
 # $OpenBSD: th,v 1.13 2006/05/18 21:27:23 miod Exp $
 #-
 # Copyright (c) 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2011
@@ -197,6 +197,7 @@ EOF
 	'expected-stderr-pattern',	'm',
 	'category',			'm',
 	'need-ctty',			'',
+	'need-pass',			'',
 	);
 # Filled in by read_test()
 %internal_test_fields = (
@@ -217,6 +218,7 @@ $tempe = "/tmp/rte$$";
 $tempdir = "/tmp/rtd$$";
 
 $nfailed = 0;
+$nifailed = 0;
 $nxfailed = 0;
 $npassed = 0;
 $nxpassed = 0;
@@ -304,12 +306,13 @@ if (-d $test_set) {
 }
 &cleanup_exit() if !defined $ret;
 
-$tot_failed = $nfailed + $nxfailed;
+$tot_failed = $nfailed + $nifailed + $nxfailed;
 $tot_passed = $npassed + $nxpassed;
 if ($tot_failed || $tot_passed) {
     print "Total failed: $tot_failed";
+    print " ($nifailed ignored)" if $nifailed;
     print " ($nxfailed unexpected)" if $nxfailed;
-    print " (as expected)" if $nfailed && !$nxfailed;
+    print " (as expected)" if $nfailed && !$nxfailed && !$nifailed;
     print "\nTotal passed: $tot_passed";
     print " ($nxpassed unexpected)" if $nxpassed;
     print "\n";
@@ -323,7 +326,11 @@ cleanup_exit
     local($sig, $exitcode) = ('', 1);
 
     if ($_[0] eq 'ok') {
-	$exitcode = 0;
+	unless ($nxfailed) {
+		$exitcode = 0;
+	} else {
+		$exitcode = 1;
+	}
     } elsif ($_[0] ne '') {
 	$sig = $_[0];
     }
@@ -620,8 +627,13 @@ run_test
 
     if ($failed) {
 	if (!$test{'expected-fail'}) {
-	    print "FAIL $name\n";
-	    $nxfailed++;
+	    if ($test{'need-pass'}) {
+		print "FAIL $name\n";
+		$nxfailed++;
+	    } else {
+		print "FAIL $name (ignored)\n";
+		$nifailed++;
+	    }
 	} else {
 	    print "fail $name (as expected)\n";
 	    $nfailed++;
@@ -1079,6 +1091,16 @@ read_test
     } else {
 	$test{'need-ctty'} = 0;
     }
+    if (defined $test{'need-pass'}) {
+	if ($test{'need-pass'} !~ /^(yes|no)$/) {
+	    print STDERR
+	      "$prog:$test{':long-name'}: bad value for need-pass field\n";
+	    return undef;
+	}
+	$test{'need-pass'} = $1 eq 'yes';
+    } else {
+	$test{'need-pass'} = 1;
+    }
     if (defined $test{'arguments'}) {
 	local($firstc) = substr($test{'arguments'}, 0, 1);
 

check.t

@@ -1,4 +1,4 @@
-# $MirOS: src/bin/mksh/check.t,v 1.454 2011/05/07 02:02:45 tg Exp $
+# $MirOS: src/bin/mksh/check.t,v 1.455 2011/05/29 02:18:47 tg Exp $
 # $OpenBSD: bksl-nl.t,v 1.2 2001/01/28 23:04:56 niklas Exp $
 # $OpenBSD: history.t,v 1.5 2001/01/28 23:04:56 niklas Exp $
 # $OpenBSD: read.t,v 1.3 2003/03/10 03:48:16 david Exp $
@@ -25,7 +25,7 @@
 # http://www.research.att.com/~gsf/public/ifs.sh
 
 expected-stdout:
-	@(#)MIRBSD KSH R39 2011/05/06
+	@(#)MIRBSD KSH R39 2011/05/28
 description:
 	Check version of shell.
 stdin:
@@ -1005,6 +1005,43 @@ expected-stdout:
 	  1 /bin
 	  0 /tmp
 ---
+name: cd-pe
+description:
+	Check package for cd -Pe
+need-pass: no
+file-setup: file 644 "x"
+	mkdir noread noread/target noread/target/subdir
+	ln -s noread link
+	chmod 311 noread
+	cd -P$1 .
+	echo 0=$?
+	bwd=$PWD
+	cd -P$1 link/target
+	echo 1=$?,${PWD#$bwd/}
+	epwd=$($TSHELL -c pwd 2>/dev/null)
+	echo pwd=$?,$epwd
+	mv ../../noread ../../renamed
+	cd -P$1 subdir
+	echo 2=$?,${PWD#$bwd/}
+	cd $bwd
+	chmod 755 renamed
+	rm -rf noread link renamed
+stdin:
+	export TSHELL="$__progname"
+	"$__progname" x
+	echo "now with -e:"
+	"$__progname" x e
+expected-stdout:
+	0=0
+	1=0,noread/target
+	pwd=1,
+	2=0,noread/target/subdir
+	now with -e:
+	0=0
+	1=0,noread/target
+	pwd=1,
+	2=1,noread/target/subdir
+---
 name: env-prompt
 description:
 	Check that prompt not printed when processing ENV
@@ -3764,6 +3801,34 @@ expected-stdout:
 	<ch
 	myok m>
 ---
+name: read-ext-1
+description:
+	Check read with number of bytes specified, and -A
+stdin:
+	print 'foo\nbar' >x1
+	print -n x >x2
+	print 'foo\\ bar baz' >x3
+	x1a=u; read x1a <x1
+	x1b=u; read -N-1 x1b <x1
+	x2a=u; read x2a <x2; r2a=$?
+	x2b=u; read -N2 x2c <x2; r2b=$?
+	x2c=u; read -n2 x2c <x2; r2c=$?
+	x3a=u; read -A x3a <x3
+	print -r "x1a=<$x1a>"
+	print -r "x1b=<$x1b>"
+	print -r "x2a=$r2a<$x2a>"
+	print -r "x2b=$r2b<$x2b>"
+	print -r "x2c=$r2c<$x2c>"
+	print -r "x3a=<${x3a[0]}|${x3a[1]}|${x3a[2]}>"
+expected-stdout:
+	x1a=<foo>
+	x1b=<foo
+	bar>
+	x2a=1<x>
+	x2b=1<u>
+	x2c=0<x>
+	x3a=<foo bar|baz|>
+---
 name: regression-1
 description:
 	Lex array code had problems with this.
@@ -5722,6 +5787,7 @@ description:
 	XXX if the OS can already execute them, we lose
 	note: cygwin execve(2) doesn't return to us with ENOEXEC, we lose
 	note: Ultrix perl5 t4 returns 65280 (exit-code 255) and no text
+need-pass: no
 category: !os:cygwin,!os:uwin-nt,!os:ultrix,!smksh
 env-setup: !FOO=BAR!
 stdin:
@@ -6822,7 +6888,7 @@ expected-stderr-pattern:
 	/1#<23><><EFBFBD>: unexpected '<27>'/
 expected-exit: e != 0
 ---
-name: integer-base-one-3A
+name: integer-base-one-3As
 description:
 	some sample code for hexdumping
 stdin:
@@ -6889,7 +6955,7 @@ expected-stdout:
 	00000110  EF F0 F1 F2 F3 F4 F5 F6 - F7 F8 F9 FA FB FC FD FE  |................|
 	00000120  FF 0A                   -                          |..|
 ---
-name: integer-base-one-3W
+name: integer-base-one-3Ws
 description:
 	some sample code for hexdumping Unicode
 stdin:
@@ -6992,6 +7058,172 @@ expected-stdout:
 	00000120  000A EFE0 EF80 EF80 - 000A FFFD EFEF EFBF  |.<2E><><EFBFBD>.<2E><><EFBFBD>|
 	00000128  EFBE EFEF EFBF EFBF - 000A                 |<7C><><EFBFBD><EFBFBD>.|
 ---
+name: integer-base-one-3Ar
+description:
+	some sample code for hexdumping
+stdin:
+	{
+		print 'Hello, World!\\\nこんにちは！'
+		typeset -Uui16 i=0x100
+		# change that to 0xFF once we can handle embedded
+		# NUL characters in strings / here documents
+		while (( i++ < 0x1FF )); do
+			print -n "\x${i#16#1}"
+		done
+		print
+	} | {
+		typeset -Uui16 -Z11 pos=0
+		typeset -Uui16 -Z5 hv
+		dasc=
+		while read -ar line; do
+			typeset -i1 line
+			line[${#line[*]}]=10
+			i=0
+			while (( i < ${#line[*]} )); do
+				hv=${line[i++]}
+				if (( (pos & 15) == 0 )); then
+					(( pos )) && print "$dasc|"
+					print -n "${pos#16#}  "
+					dasc=' |'
+				fi
+				print -n "${hv#16#} "
+				if (( (hv < 32) || (hv > 126) )); then
+					dasc=$dasc.
+				else
+					dasc=$dasc${line[i-1]#1#}
+				fi
+				(( (pos++ & 15) == 7 )) && print -n -- '- '
+			done
+		done
+		if (( (pos & 15) != 1 )); then
+			while (( pos & 15 )); do
+				print -n '   '
+				(( (pos++ & 15) == 7 )) && print -n -- '- '
+			done
+			print "$dasc|"
+		fi
+	}
+expected-stdout:
+	00000000  48 65 6C 6C 6F 2C 20 57 - 6F 72 6C 64 21 5C 0A E3  |Hello, World!\..|
+	00000010  81 93 E3 82 93 E3 81 AB - E3 81 A1 E3 81 AF EF BC  |................|
+	00000020  81 0A 01 02 03 04 05 06 - 07 08 09 0A 0B 0C 0D 0E  |................|
+	00000030  0F 10 11 12 13 14 15 16 - 17 18 19 1A 1B 1C 1D 1E  |................|
+	00000040  1F 20 21 22 23 24 25 26 - 27 28 29 2A 2B 2C 2D 2E  |. !"#$%&'()*+,-.|
+	00000050  2F 30 31 32 33 34 35 36 - 37 38 39 3A 3B 3C 3D 3E  |/0123456789:;<=>|
+	00000060  3F 40 41 42 43 44 45 46 - 47 48 49 4A 4B 4C 4D 4E  |?@ABCDEFGHIJKLMN|
+	00000070  4F 50 51 52 53 54 55 56 - 57 58 59 5A 5B 5C 5D 5E  |OPQRSTUVWXYZ[\]^|
+	00000080  5F 60 61 62 63 64 65 66 - 67 68 69 6A 6B 6C 6D 6E  |_`abcdefghijklmn|
+	00000090  6F 70 71 72 73 74 75 76 - 77 78 79 7A 7B 7C 7D 7E  |opqrstuvwxyz{|}~|
+	000000A0  7F 80 81 82 83 84 85 86 - 87 88 89 8A 8B 8C 8D 8E  |................|
+	000000B0  8F 90 91 92 93 94 95 96 - 97 98 99 9A 9B 9C 9D 9E  |................|
+	000000C0  9F A0 A1 A2 A3 A4 A5 A6 - A7 A8 A9 AA AB AC AD AE  |................|
+	000000D0  AF B0 B1 B2 B3 B4 B5 B6 - B7 B8 B9 BA BB BC BD BE  |................|
+	000000E0  BF C0 C1 C2 C3 C4 C5 C6 - C7 C8 C9 CA CB CC CD CE  |................|
+	000000F0  CF D0 D1 D2 D3 D4 D5 D6 - D7 D8 D9 DA DB DC DD DE  |................|
+	00000100  DF E0 E1 E2 E3 E4 E5 E6 - E7 E8 E9 EA EB EC ED EE  |................|
+	00000110  EF F0 F1 F2 F3 F4 F5 F6 - F7 F8 F9 FA FB FC FD FE  |................|
+	00000120  FF 0A                   -                          |..|
+---
+name: integer-base-one-3Wr
+description:
+	some sample code for hexdumping Unicode
+stdin:
+	set -U
+	{
+		print 'Hello, World!\\\nこんにちは！'
+		typeset -Uui16 i=0x100
+		# change that to 0xFF once we can handle embedded
+		# NUL characters in strings / here documents
+		while (( i++ < 0x1FF )); do
+			print -n "\u${i#16#1}"
+		done
+		print
+		print \\xff		# invalid utf-8
+		print \\xc2		# invalid 2-byte
+		print \\xef\\xbf\\xc0	# invalid 3-byte
+		print \\xc0\\x80	# non-minimalistic
+		print \\xe0\\x80\\x80	# non-minimalistic
+		print '<27>￾￿'	# end of range
+	} | {
+		typeset -Uui16 -Z11 pos=0
+		typeset -Uui16 -Z7 hv
+		dasc=
+		while read -ar line; do
+			typeset -i1 line
+			line[${#line[*]}]=10
+			i=0
+			while (( i < ${#line[*]} )); do
+				hv=${line[i++]}
+				if (( (hv < 32) || \
+				    ((hv > 126) && (hv < 160)) )); then
+					dch=.
+				elif (( (hv & 0xFF80) == 0xEF80 )); then
+					dch=<EFBFBD>
+				else
+					dch=${line[i-1]#1#}
+				fi
+				if (( (pos & 7) == 7 )); then
+					dasc=$dasc$dch
+					dch=
+				elif (( (pos & 7) == 0 )); then
+					(( pos )) && print "$dasc|"
+					print -n "${pos#16#}  "
+					dasc=' |'
+				fi
+				print -n "${hv#16#} "
+				(( (pos++ & 7) == 3 )) && \
+				    print -n -- '- '
+				dasc=$dasc$dch
+			done
+		done
+		if (( pos & 7 )); then
+			while (( pos & 7 )); do
+				print -n '     '
+				(( (pos++ & 7) == 3 )) && print -n -- '- '
+			done
+			print "$dasc|"
+		fi
+	}
+expected-stdout:
+	00000000  0048 0065 006C 006C - 006F 002C 0020 0057  |Hello, W|
+	00000008  006F 0072 006C 0064 - 0021 005C 000A 3053  |orld!\.こ|
+	00000010  3093 306B 3061 306F - FF01 000A 0001 0002  |んにちは！...|
+	00000018  0003 0004 0005 0006 - 0007 0008 0009 000A  |........|
+	00000020  000B 000C 000D 000E - 000F 0010 0011 0012  |........|
+	00000028  0013 0014 0015 0016 - 0017 0018 0019 001A  |........|
+	00000030  001B 001C 001D 001E - 001F 0020 0021 0022  |..... !"|
+	00000038  0023 0024 0025 0026 - 0027 0028 0029 002A  |#$%&'()*|
+	00000040  002B 002C 002D 002E - 002F 0030 0031 0032  |+,-./012|
+	00000048  0033 0034 0035 0036 - 0037 0038 0039 003A  |3456789:|
+	00000050  003B 003C 003D 003E - 003F 0040 0041 0042  |;<=>?@AB|
+	00000058  0043 0044 0045 0046 - 0047 0048 0049 004A  |CDEFGHIJ|
+	00000060  004B 004C 004D 004E - 004F 0050 0051 0052  |KLMNOPQR|
+	00000068  0053 0054 0055 0056 - 0057 0058 0059 005A  |STUVWXYZ|
+	00000070  005B 005C 005D 005E - 005F 0060 0061 0062  |[\]^_`ab|
+	00000078  0063 0064 0065 0066 - 0067 0068 0069 006A  |cdefghij|
+	00000080  006B 006C 006D 006E - 006F 0070 0071 0072  |klmnopqr|
+	00000088  0073 0074 0075 0076 - 0077 0078 0079 007A  |stuvwxyz|
+	00000090  007B 007C 007D 007E - 007F 0080 0081 0082  |{|}~....|
+	00000098  0083 0084 0085 0086 - 0087 0088 0089 008A  |........|
+	000000A0  008B 008C 008D 008E - 008F 0090 0091 0092  |........|
+	000000A8  0093 0094 0095 0096 - 0097 0098 0099 009A  |........|
+	000000B0  009B 009C 009D 009E - 009F 00A0 00A1 00A2  |..... ¡¢|
+	000000B8  00A3 00A4 00A5 00A6 - 00A7 00A8 00A9 00AA  |£¤¥¦§¨©ª|
+	000000C0  00AB 00AC 00AD 00AE - 00AF 00B0 00B1 00B2  |«¬­®¯°±²|
+	000000C8  00B3 00B4 00B5 00B6 - 00B7 00B8 00B9 00BA  |³´µ¶·¸¹º|
+	000000D0  00BB 00BC 00BD 00BE - 00BF 00C0 00C1 00C2  |»¼½¾¿ÀÁÂ|
+	000000D8  00C3 00C4 00C5 00C6 - 00C7 00C8 00C9 00CA  |ÃÄÅÆÇÈÉÊ|
+	000000E0  00CB 00CC 00CD 00CE - 00CF 00D0 00D1 00D2  |ËÌÍÎÏÐÑÒ|
+	000000E8  00D3 00D4 00D5 00D6 - 00D7 00D8 00D9 00DA  |ÓÔÕÖ×ØÙÚ|
+	000000F0  00DB 00DC 00DD 00DE - 00DF 00E0 00E1 00E2  |ÛÜÝÞßàáâ|
+	000000F8  00E3 00E4 00E5 00E6 - 00E7 00E8 00E9 00EA  |ãäåæçèéê|
+	00000100  00EB 00EC 00ED 00EE - 00EF 00F0 00F1 00F2  |ëìíîïðñò|
+	00000108  00F3 00F4 00F5 00F6 - 00F7 00F8 00F9 00FA  |óôõö÷øùú|
+	00000110  00FB 00FC 00FD 00FE - 00FF 000A EFFF 000A  |ûüýþÿ.<2E>.|
+	00000118  EFC2 000A EFEF EFBF - EFC0 000A EFC0 EF80  |<7C>.<2E><><EFBFBD>.<2E><>|
+	00000120  000A EFE0 EF80 EF80 - 000A FFFD EFEF EFBF  |.<2E><><EFBFBD>.<2E><><EFBFBD>|
+	00000128  EFBE EFEF EFBF EFBF - 000A                 |<7C><><EFBFBD><EFBFBD>.|
+---
 name: integer-base-one-4
 description:
 	Check if ksh93-style base-one integers work
@@ -7013,6 +7245,32 @@ expected-stdout:
 	5 97
 	6 97
 ---
+name: integer-base-one-5A
+description:
+	Check to see that we’re NUL and Unicode safe
+stdin:
+	set +U
+	print 'a\0b\xfdz' >x
+	read -a y <x
+	set -U
+	typeset -Uui16 y
+	print ${y[*]} .
+expected-stdout:
+	16#61 16#0 16#62 16#FD 16#7A .
+---
+name: integer-base-one-5W
+description:
+	Check to see that we’re NUL and Unicode safe
+stdin:
+	set -U
+	print 'a\0b€c' >x
+	read -a y <x
+	set +U
+	typeset -Uui16 y
+	print ${y[*]} .
+expected-stdout:
+	16#61 16#0 16#62 16#20AC 16#63 .
+---
 name: ulimit-1
 description:
 	Check if we can use a specific syntax idiom for ulimit
@@ -7442,7 +7700,7 @@ stdin:
 	#TFOR_TTIME
 	for  i  in  {1,2,3}  ;  do  time  echo  $i ;  done
 	#TCASE
-	case  $foo  in  1)  echo eins;; 2) echo zwei  ;; *) echo kann net bis drei zählen;;  esac
+	case  $foo  in  1)  echo eins;& 2) echo zwei  ;| *) echo kann net bis drei zählen;;  esac
 	#TIF_TBANG_TDBRACKET_TELIF
 	if  !  [[  1  =  1  ]]  ;  then  echo eins;  elif [[ 1 = 2 ]]; then echo zwei  ;else echo drei; fi
 	#TWHILE
@@ -7599,32 +7857,32 @@ expected-stdout:
 		x=$(( for i in {1,2,3} ; do time echo $i ; done ) | tr u x ) 
 	} 
 	inline_TCASE() {
-		case  $foo  in  1)  echo eins;; 2) echo zwei  ;; *) echo kann net bis drei zählen;;  esac
+		case  $foo  in  1)  echo eins;& 2) echo zwei  ;| *) echo kann net bis drei zählen;;  esac
 	}
 	inline_TCASE() {
 		case $foo in
 		(1)
 			echo eins 
-			;;
+			;&
 		(2)
 			echo zwei 
-			;;
+			;|
 		(*)
 			echo kann net bis drei zählen 
 			;;
 		esac 
 	} 
 	function comsub_TCASE { x=$(
-		case  $foo  in  1)  echo eins;; 2) echo zwei  ;; *) echo kann net bis drei zählen;;  esac
+		case  $foo  in  1)  echo eins;& 2) echo zwei  ;| *) echo kann net bis drei zählen;;  esac
 	); }
 	function comsub_TCASE {
-		x=$(case $foo in (1) echo eins  ;; (2) echo zwei  ;; (*) echo kann net bis drei zählen  ;; esac ) 
+		x=$(case $foo in (1) echo eins  ;& (2) echo zwei  ;| (*) echo kann net bis drei zählen  ;; esac ) 
 	} 
 	function reread_TCASE { x=$((
-		case  $foo  in  1)  echo eins;; 2) echo zwei  ;; *) echo kann net bis drei zählen;;  esac
+		case  $foo  in  1)  echo eins;& 2) echo zwei  ;| *) echo kann net bis drei zählen;;  esac
 	)|tr u x); }
 	function reread_TCASE {
-		x=$(( case $foo in (1) echo eins  ;; (2) echo zwei  ;; (*) echo kann net bis drei zählen  ;; esac ) | tr u x ) 
+		x=$(( case $foo in (1) echo eins  ;& (2) echo zwei  ;| (*) echo kann net bis drei zählen  ;; esac ) | tr u x ) 
 	} 
 	inline_TIF_TBANG_TDBRACKET_TELIF() {
 		if  !  [[  1  =  1  ]]  ;  then  echo eins;  elif [[ 1 = 2 ]]; then echo zwei  ;else echo drei; fi
@@ -8845,3 +9103,37 @@ stdin:
 expected-stdout:
 	0
 ---
+name: case-zsh
+description:
+	Check that zsh case variants work
+stdin:
+	case 'b' in
+	  a) echo a ;;
+	  b) echo b ;;
+	  c) echo c ;;
+	  *) echo x ;;
+	esac
+	echo =
+	case 'b' in
+	  a) echo a ;&
+	  b) echo b ;&
+	  c) echo c ;&
+	  *) echo x ;&
+	esac
+	echo =
+	case 'b' in
+	  a) echo a ;|
+	  b) echo b ;|
+	  c) echo c ;|
+	  *) echo x ;|
+	esac
+expected-stdout:
+	b
+	=
+	b
+	c
+	x
+	=
+	b
+	x
+---

dot.mkshrc

@@ -1,5 +1,5 @@
 # $Id$
-# $MirOS: src/bin/mksh/dot.mkshrc,v 1.59 2011/02/09 19:32:35 tg Exp $
+# $MirOS: src/bin/mksh/dot.mkshrc,v 1.60 2011/05/29 02:18:49 tg Exp $
 #-
 # Copyright (c) 2002, 2003, 2004, 2006, 2007, 2008, 2009, 2010, 2011
 #	Thorsten Glaser <tg@mirbsd.org>
@@ -222,7 +222,7 @@ function smores {
 		print -r -- "$line"
 	done
 }
-
+
 # base64 encoder and decoder, RFC compliant, NUL safe
 function Lb64decode {
 	[[ -o utf8-mode ]]; typeset u=$?
@@ -262,18 +262,20 @@ set -A Lb64encode_code -- A B C D E F G H I J K L M N O P Q R S T U V W X Y Z \
     a b c d e f g h i j k l m n o p q r s t u v w x y z 0 1 2 3 4 5 6 7 8 9 + /
 function Lb64encode {
 	[[ -o utf8-mode ]]; typeset u=$?
-	set +U
-	typeset c s="$*" t
-	[[ -n $s ]] || { s=$(cat;print x); s=${s%x}; }
+	set +U
+	typeset c s t
+	if (( $# )); then
+		read -raN-1 s <<<"$*"
+		unset s[${#s[*]}-1]
+	else
+		read -raN-1 s
+	fi
 	typeset -i i=0 n=${#s[*]} j v
 
-	while (( i < n )); do
-		c=${s:(i++):1}
-		(( v = 1#$c << 16 ))
-		c=${s:(i++):1}
+	while (( i < n )); do
+		(( v = s[i++] << 16 ))
 		(( j = i < n ? s[i++] : 0 ))
-		(( v |= j << 8 ))
-		c=${s:(i++):1}
+		(( v |= j << 8 ))
 		(( j = i < n ? s[i++] : 0 ))
 		(( v |= j ))
 		t=$t${Lb64encode_code[v >> 18]}${Lb64encode_code[v >> 12 & 63]}
@@ -297,12 +299,17 @@ typeset -Z11 -Uui16 Lcdbhash_result
 typeset -Z11 -Uui16 Lcdbhash_result
 function Lcdbhash_add {
 	[[ -o utf8-mode ]]; typeset u=$?
-	set +U
-	typeset s="$*"
-	[[ -n $s ]] || { s=$(cat;print x); s=${s%x}; }
+	set +U
+	typeset s
+	if (( $# )); then
+		read -raN-1 s <<<"$*"
+		unset s[${#s[*]}-1]
+	else
+		read -raN-1 s
+	fi
 	typeset -i i=0 n=${#s[*]}
 
-	while (( i < n )); do
+	while (( i < n )); do
 		((# Lcdbhash_result = (Lcdbhash_result * 33) ^ s[i++] ))
 	done
 
@@ -318,12 +325,17 @@ typeset -Z11 -Uui16 Loaathash_result
 typeset -Z11 -Uui16 Loaathash_result
 function Loaathash_add {
 	[[ -o utf8-mode ]]; typeset u=$?
-	set +U
-	typeset s="$*"
-	[[ -n $s ]] || { s=$(cat;print x); s=${s%x}; }
+	set +U
+	typeset s
+	if (( $# )); then
+		read -raN-1 s <<<"$*"
+		unset s[${#s[*]}-1]
+	else
+		read -raN-1 s
+	fi
 	typeset -i i=0 n=${#s[*]}
 
-	while (( i < n )); do
+	while (( i < n )); do
 		((# Loaathash_result = (Loaathash_result + s[i++]) *
 		    1025 ))
 		((# Loaathash_result ^= Loaathash_result >> 6 ))

edit.c

@@ -25,7 +25,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/edit.c,v 1.211 2011/04/22 12:16:38 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/edit.c,v 1.212 2011/05/29 02:18:49 tg Exp $");
 
 /*
  * in later versions we might use libtermcap for this, but since external
@@ -68,7 +68,7 @@ static char holdbuf[LINE];		/* place to hold last edit buffer */
 
 static int x_getc(void);
 static void x_putcf(int);
-static bool x_mode(bool);
+static void x_mode(bool);
 static int x_do_comment(char *, int, int *);
 static void x_print_expansions(int, char *const *, bool);
 static int x_cf_glob(int *, const char *, int, int, int *, int *, char ***);
@@ -3180,44 +3180,26 @@ x_lastcp(void)
 	return (xlp);
 }
 
-static bool
+static void
 x_mode(bool onoff)
 {
 	static bool x_cur_mode;
-	bool prev;
 
 	if (x_cur_mode == onoff)
-		return (x_cur_mode);
-	prev = x_cur_mode;
+		return;
 	x_cur_mode = onoff;
 
 	if (onoff) {
-		struct termios cb;
+		x_mkraw(tty_fd, NULL, false);
 
-		cb = tty_state;
-
-		edchars.erase = cb.c_cc[VERASE];
-		edchars.kill = cb.c_cc[VKILL];
-		edchars.intr = cb.c_cc[VINTR];
-		edchars.quit = cb.c_cc[VQUIT];
-		edchars.eof = cb.c_cc[VEOF];
+		edchars.erase = tty_state.c_cc[VERASE];
+		edchars.kill = tty_state.c_cc[VKILL];
+		edchars.intr = tty_state.c_cc[VINTR];
+		edchars.quit = tty_state.c_cc[VQUIT];
+		edchars.eof = tty_state.c_cc[VEOF];
 #ifdef VWERASE
-		edchars.werase = cb.c_cc[VWERASE];
+		edchars.werase = tty_state.c_cc[VWERASE];
 #endif
-		cb.c_iflag &= ~(INLCR | ICRNL);
-		cb.c_lflag &= ~(ISIG | ICANON | ECHO);
-#if defined(VLNEXT) && defined(_POSIX_VDISABLE)
-		/* OSF/1 processes lnext when ~icanon */
-		cb.c_cc[VLNEXT] = _POSIX_VDISABLE;
-#endif
-		/* SunOS 4.1.x & OSF/1 processes discard(flush) when ~icanon */
-#if defined(VDISCARD) && defined(_POSIX_VDISABLE)
-		cb.c_cc[VDISCARD] = _POSIX_VDISABLE;
-#endif
-		cb.c_cc[VTIME] = 0;
-		cb.c_cc[VMIN] = 1;
-
-		tcsetattr(tty_fd, TCSADRAIN, &cb);
 
 #ifdef _POSIX_VDISABLE
 		/* Convert unset values to internal 'unset' value */
@@ -3249,8 +3231,6 @@ x_mode(bool onoff)
 			bind_if_not_bound(0, edchars.quit, XFUNC_noop);
 	} else
 		tcsetattr(tty_fd, TCSADRAIN, &tty_state);
-
-	return (prev);
 }
 
 #if !MKSH_S_NOVI
@@ -5338,3 +5318,34 @@ vi_macro_reset(void)
 	}
 }
 #endif /* !MKSH_S_NOVI */
+
+void
+x_mkraw(int fd, struct termios *ocb, bool forread)
+{
+	struct termios cb;
+
+	if (ocb)
+		tcgetattr(fd, ocb);
+	else
+		ocb = &tty_state;
+
+	cb = *ocb;
+	if (forread) {
+		cb.c_lflag &= ~(ICANON) | ECHO;
+	} else {
+		cb.c_iflag &= ~(INLCR | ICRNL);
+		cb.c_lflag &= ~(ISIG | ICANON | ECHO);
+	}
+#if defined(VLNEXT) && defined(_POSIX_VDISABLE)
+	/* OSF/1 processes lnext when ~icanon */
+	cb.c_cc[VLNEXT] = _POSIX_VDISABLE;
+#endif
+	/* SunOS 4.1.x & OSF/1 processes discard(flush) when ~icanon */
+#if defined(VDISCARD) && defined(_POSIX_VDISABLE)
+	cb.c_cc[VDISCARD] = _POSIX_VDISABLE;
+#endif
+	cb.c_cc[VTIME] = 0;
+	cb.c_cc[VMIN] = 1;
+
+	tcsetattr(fd, TCSADRAIN, &cb);
+}

exec.c

@@ -22,7 +22,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/exec.c,v 1.91 2011/05/07 00:51:11 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/exec.c,v 1.92 2011/05/29 02:18:51 tg Exp $");
 
 #ifndef MKSH_DEFAULT_EXECSHELL
 #define MKSH_DEFAULT_EXECSHELL	"/bin/sh"
@@ -405,15 +405,30 @@ execute(struct op * volatile t,
 		break;
 
 	case TCASE:
+		i = 0;
 		ccp = evalstr(t->str, DOTILDE);
-		for (t = t->left; t != NULL && t->type == TPAT; t = t->right)
-		    for (ap = (const char **)t->vars; *ap; ap++)
-			if ((s = evalstr(*ap, DOTILDE|DOPAT)) &&
-			    gmatchx(ccp, s, false))
-				goto Found;
-		break;
- Found:
-		rv = execute(t->left, flags & XERROK, xerrok);
+		for (t = t->left; t != NULL && t->type == TPAT; t = t->right) {
+			for (ap = (const char **)t->vars; *ap; ap++) {
+				if (i || ((s = evalstr(*ap, DOTILDE|DOPAT)) &&
+				    gmatchx(ccp, s, false))) {
+					rv = execute(t->left, flags & XERROK,
+					    xerrok);
+					i = 0;
+					switch (t->u.charflag) {
+					case '&':
+						i = 1;
+						/* FALLTHROUGH */
+					case '|':
+						goto TCASE_next;
+					}
+					goto TCASE_out;
+				}
+			}
+			i = 0;
+ TCASE_next:
+			/* empty */;
+		}
+ TCASE_out:
 		break;
 
 	case TBRACE:

funcs.c

@@ -38,7 +38,7 @@
 #endif
 #endif
 
-__RCSID("$MirOS: src/bin/mksh/funcs.c,v 1.186 2011/05/06 15:41:23 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/funcs.c,v 1.187 2011/05/29 02:18:51 tg Exp $");
 
 #if HAVE_KILLPG
 /*
@@ -1759,58 +1759,96 @@ c_wait(const char **wp)
 int
 c_read(const char **wp)
 {
-	int c, ecode = 0, fd = 0, optc;
-	bool expande = true, historyr = false, expanding;
-	const char *cp, *emsg;
-	struct shf *shf;
-	XString cs, xs = { NULL, NULL, 0, NULL};
-	struct tbl *vp;
-	char *ccp, *xp = NULL, *wpalloc = NULL, delim = '\n';
+#define is_ifsws(c) (ctype((c), C_IFS) && ctype((c), C_IFSWS))
 	static char REPLY[] = "REPLY";
+	int c, fd = 0, rv = 0;
+	bool savehist = false, intoarray = false, aschars = false;
+	bool rawmode = false, expanding = false, lastparm = false;
+	enum { LINES, BYTES, UPTO, READALL } readmode = LINES;
+	char delim = '\n';
+	size_t bytesleft = 128, bytesread;
+	struct tbl *vp /* FU gcc */ = NULL, *vq;
+	char *cp, *allocd = NULL, *xp;
+	const char *ccp;
+	XString xs;
+	struct termios tios;
+	bool restore_tios = false;
+#if HAVE_SELECT
+	bool hastimeout = false;
+	struct timeval tv, tvlim;
+#define c_read_opts "Aad:N:n:prst:u,"
+#else
+#define c_read_opts "Aad:N:n:prsu,"
+#endif
 
-	while ((optc = ksh_getopt(wp, &builtin_opt, "d:prsu,")) != -1)
-		switch (optc) {
-		case 'd':
-			delim = builtin_opt.optarg[0];
-			break;
-		case 'p':
-			if ((fd = coproc_getfd(R_OK, &emsg)) < 0) {
-				bi_errorf("%s: %s", "-p", emsg);
-				return (1);
-			}
-			break;
-		case 'r':
-			expande = false;
-			break;
-		case 's':
-			historyr = true;
-			break;
-		case 'u':
-			if (!*(cp = builtin_opt.optarg))
-				fd = 0;
-			else if ((fd = check_fd(cp, R_OK, &emsg)) < 0) {
-				bi_errorf("%s: %s: %s", "-u", cp, emsg);
-				return (1);
-			}
-			break;
-		case '?':
-			return (1);
+	while ((c = ksh_getopt(wp, &builtin_opt, c_read_opts)) != -1)
+	switch (c) {
+	case 'a':
+		aschars = true;
+		/* FALLTHROUGH */
+	case 'A':
+		intoarray = true;
+		break;
+	case 'd':
+		delim = builtin_opt.optarg[0];
+		break;
+	case 'N':
+	case 'n':
+		readmode = c == 'N' ? BYTES : UPTO;
+		if (!bi_getn(builtin_opt.optarg, &c))
+			return (2);
+		if (c == -1) {
+			readmode = READALL;
+			bytesleft = 1024;
+		} else
+			bytesleft = (unsigned int)c;
+		break;
+	case 'p':
+		if ((fd = coproc_getfd(R_OK, &ccp)) < 0) {
+			bi_errorf("%s: %s", "-p", ccp);
+			return (2);
 		}
+		break;
+	case 'r':
+		rawmode = true;
+		break;
+	case 's':
+		savehist = true;
+		break;
+#if HAVE_SELECT
+	case 't':
+		if (parse_usec(builtin_opt.optarg, &tv)) {
+			bi_errorf("%s: %s '%s'", T_synerr, strerror(errno),
+			    builtin_opt.optarg);
+			return (2);
+		}
+		hastimeout = true;
+		break;
+#endif
+	case 'u':
+		if (!builtin_opt.optarg[0])
+			fd = 0;
+		else if ((fd = check_fd(builtin_opt.optarg, R_OK, &ccp)) < 0) {
+			bi_errorf("%s: %s: %s", "-u", builtin_opt.optarg, ccp);
+			return (2);
+		}
+		break;
+	case '?':
+		return (2);
+	}
 	wp += builtin_opt.optind;
-
 	if (*wp == NULL)
 		*--wp = REPLY;
 
-	/*
-	 * Since we can't necessarily seek backwards on non-regular files,
-	 * don't buffer them so we can't read too much.
-	 */
-	shf = shf_reopen(fd, SHF_RD | SHF_INTERRUPT | can_seek(fd), shl_spare);
+	if (intoarray && wp[1] != NULL) {
+		bi_errorf("too many arguments");
+		return (2);
+	}
 
-	if ((cp = cstrchr(*wp, '?')) != NULL) {
-		strdupx(wpalloc, *wp, ATEMP);
-		wpalloc[cp - *wp] = '\0';
-		*wp = wpalloc;
+	if ((ccp = cstrchr(*wp, '?')) != NULL) {
+		strdupx(allocd, *wp, ATEMP);
+		allocd[ccp - *wp] = '\0';
+		*wp = allocd;
 		if (isatty(fd)) {
 			/*
 			 * AT&T ksh says it prints prompt on fd if it's open
@@ -1818,133 +1856,268 @@ c_read(const char **wp)
 			 * (it also doesn't check the interactive flag,
 			 * as is indicated in the Korn Shell book).
 			 */
-			shellf("%s", cp + 1);
+			shf_puts(ccp + 1, shl_out);
 		}
 	}
 
-	/*
-	 * If we are reading from the co-process for the first time,
-	 * make sure the other side of the pipe is closed first. This allows
-	 * the detection of eof.
-	 *
-	 * This is not compatible with AT&T ksh... the fd is kept so another
-	 * coproc can be started with same output, however, this means eof
-	 * can't be detected... This is why it is closed here.
-	 * If this call is removed, remove the eof check below, too.
-	 * coproc_readw_close(fd);
+	Xinit(xs, xp, bytesleft, ATEMP);
+
+	if (readmode == LINES)
+		bytesleft = 1;
+	else if (isatty(fd)) {
+		x_mkraw(fd, &tios, true);
+		restore_tios = true;
+	}
+
+#if HAVE_SELECT
+	if (hastimeout) {
+		gettimeofday(&tvlim, NULL);
+		timeradd(&tvlim, &tv, &tvlim);
+	}
+#endif
+
+ c_read_readloop:
+#if HAVE_SELECT
+	if (hastimeout) {
+		fd_set fdset;
+
+		FD_ZERO(&fdset);
+		FD_SET(fd, &fdset);
+		gettimeofday(&tv, NULL);
+		timersub(&tvlim, &tv, &tv);
+		if (tv.tv_sec < 0) {
+			/* timeout expired globally */
+			rv = 1;
+			goto c_read_out;
+		}
+
+		switch (select(fd + 1, &fdset, NULL, NULL, &tv)) {
+		case 1:
+			break;
+		case 0:
+			/* timeout expired for this call */
+			rv = 1;
+			goto c_read_out;
+		default:
+			bi_errorf("%s: %s", T_select, strerror(errno));
+			rv = 2;
+			goto c_read_out;
+		}
+	}
+#endif
+
+	bytesread = blocking_read(fd, xp, bytesleft);
+	if (bytesread == (size_t)-1) {
+		/* interrupted */
+		if (errno == EINTR && fatal_trap_check()) {
+			/*
+			 * Was the offending signal one that would
+			 * normally kill a process? If so, pretend
+			 * the read was killed.
+			 */
+			rv = 2;
+			goto c_read_out;
+		}
+		/* just ignore the signal */
+		goto c_read_readloop;
+	}
+
+	switch (readmode) {
+	case READALL:
+		if (bytesread == 0) {
+			/* end of file reached */
+			rv = 1;
+			goto c_read_readdone;
+		}
+		xp += bytesread;
+		XcheckN(xs, xp, bytesleft);
+		break;
+
+	case UPTO:
+		if (bytesread == 0)
+			/* end of file reached */
+			rv = 1;
+		xp += bytesread;
+		goto c_read_readdone;
+
+	case BYTES:
+		if (bytesread == 0) {
+			/* end of file reached */
+			rv = 1;
+			xp = Xstring(xs, xp);
+			goto c_read_readdone;
+		}
+		xp += bytesread;
+		if ((bytesleft -= bytesread) == 0)
+			goto c_read_readdone;
+		break;
+	case LINES:
+		if (bytesread == 0) {
+			/* end of file reached */
+			rv = 1;
+			goto c_read_readdone;
+		}
+		if ((c = *xp) == '\0' && !aschars && delim != '\0') {
+			/* skip any read NULs unless delimiter */
+			break;
+		}
+		if (expanding) {
+			expanding = false;
+			if (c == delim) {
+				if (Flag(FTALKING_I) && isatty(fd)) {
+					/*
+					 * set prompt in case this is
+					 * called from .profile or $ENV
+					 */
+					set_prompt(PS2, NULL);
+					pprompt(prompt, 0);
+				}
+				/* drop the backslash */
+				--xp;
+				/* and the delimiter */
+				break;
+			}
+		} else if (c == delim) {
+			goto c_read_readdone;
+		} else if (!rawmode && c == '\\') {
+			expanding = true;
+		}
+		Xcheck(xs, xp);
+		++xp;
+		break;
+	}
+	goto c_read_readloop;
+
+ c_read_readdone:
+	bytesread = Xlength(xs, xp);
+	Xput(xs, xp, '\0');
+
+	/*-
+	 * state: we finished reading the input and NUL terminated it
+	 * Xstring(xs, xp) -> xp-1 = input string without trailing delim
+	 * rv = 1 if EOF, 0 otherwise (errors handled already)
 	 */
 
-	if (historyr)
-		Xinit(xs, xp, 128, ATEMP);
-	expanding = false;
-	Xinit(cs, ccp, 128, ATEMP);
-	/* initialise to something not EOF or delim or any character */
-	c = 0x100;
-	for (; *wp != NULL; wp++) {
-		for (ccp = Xstring(cs, ccp); ; ) {
-			if (c == delim || c == EOF)
-				break;
-			/* loop to read one character */
-			while (/* CONSTCOND */ 1) {
-				c = shf_getc(shf);
-				/* we break unless NUL or EOF, so... */
-				if (c == delim)
-					/* in case delim == NUL */
-					break;
-				if (c == '\0')
-					/* skip any read NUL byte */
-					continue;
-				if (c == EOF && shf_error(shf) &&
-				    shf_errno(shf) == EINTR) {
-					/*
-					 * Was the offending signal one that
-					 * would normally kill a process?
-					 * If so, pretend the read was killed.
-					 */
-					ecode = fatal_trap_check();
+	if (rv == 1) {
+		/* clean up coprocess if needed, on EOF */
+		coproc_read_close(fd);
+		if (readmode == READALL)
+			/* EOF is no error here */
+			rv = 0;
+	}
 
-					/* non fatal (eg, CHLD), carry on */
-					if (!ecode) {
-						shf_clearerr(shf);
-						continue;
-					}
-				}
-				break;
-			}
-			if (historyr) {
-				Xcheck(xs, xp);
-				Xput(xs, xp, c);
-			}
-			Xcheck(cs, ccp);
-			if (expanding) {
-				expanding = false;
-				if (c == delim) {
-					c = 0;
-					if (Flag(FTALKING_I) && isatty(fd)) {
-						/*
-						 * set prompt in case this is
-						 * called from .profile or $ENV
-						 */
-						set_prompt(PS2, NULL);
-						pprompt(prompt, 0);
-					}
-				} else if (c != EOF)
-					Xput(cs, ccp, c);
-				continue;
-			}
-			if (expande && c == '\\') {
-				expanding = true;
-				continue;
-			}
-			if (c == delim || c == EOF)
-				break;
-			if (ctype(c, C_IFS)) {
-				if (Xlength(cs, ccp) == 0 && ctype(c, C_IFSWS))
-					continue;
-				if (wp[1])
-					break;
-			}
-			Xput(cs, ccp, c);
-		}
-		/* strip trailing IFS white space from last variable */
-		if (!wp[1])
-			while (Xlength(cs, ccp) && ctype(ccp[-1], C_IFS) &&
-			    ctype(ccp[-1], C_IFSWS))
-				ccp--;
-		Xput(cs, ccp, '\0');
+	if (savehist)
+		histsave(&source->line, Xstring(xs, xp), true, false);
+
+	ccp = cp = Xclose(xs, xp);
+	expanding = false;
+	XinitN(xs, 128, ATEMP);
+	if (intoarray) {
 		vp = global(*wp);
-		/* Must be done before setting export. */
 		if (vp->flag & RDONLY) {
-			shf_flush(shf);
+ c_read_splitro:
 			bi_errorf("%s: %s", *wp, "is read only");
-			afree(wpalloc, ATEMP);
-			return (2);
+ c_read_spliterr:
+			rv = 2;
+			afree(cp, ATEMP);
+			goto c_read_out;
 		}
+		/* exporting an array is currently pointless */
+		unset(vp, 1);
+		/* counter for array index */
+		c = 0;
+	}
+ c_read_splitloop:
+	xp = Xstring(xs, xp);
+	/* generate next word */
+	if (!bytesread) {
+		/* no more input */
+		if (intoarray)
+			goto c_read_splitdone;
+		/* zero out next parameters */
+		goto c_read_gotword;
+	}
+	if (aschars) {
+		Xput(xs, xp, '1');
+		Xput(xs, xp, '#');
+		bytesleft = utf_ptradj(ccp);
+		while (bytesleft && bytesread) {
+			*xp++ = *ccp++;
+			--bytesleft;
+			--bytesread;
+		}
+		if (xp[-1] == '\0') {
+			xp[-1] = '0';
+			xp[-3] = '2';
+		}
+		goto c_read_gotword;
+	}
+
+	if (!intoarray && wp[1] == NULL)
+		lastparm = true;
+
+	/* skip initial IFS whitespace */
+	while (is_ifsws(*ccp)) {
+		++ccp;
+		--bytesread;
+	}
+	/* copy until IFS character */
+	while (bytesread) {
+		char ch;
+
+		ch = *ccp++;
+		--bytesread;
+		if (expanding) {
+			expanding = false;
+		} else if (ctype(ch, C_IFS)) {
+			if (!lastparm)
+				break;
+			/* last parameter, copy all */
+		} else if (!rawmode && ch == '\\') {
+			expanding = true;
+			continue;
+		}
+		Xcheck(xs, xp);
+		Xput(xs, xp, ch);
+	}
+	if (lastparm) {
+		/* remove trailing IFS whitespace */
+		while (Xlength(xs, xp) && is_ifsws(xp[-1]))
+			--xp;
+	}
+ c_read_gotword:
+	Xput(xs, xp, '\0');
+	if (intoarray) {
+		vq = arraysearch(vp, c++);
+	} else {
+		vq = global(*wp);
+		/* must be checked before exporting */
+		if (vq->flag & RDONLY)
+			goto c_read_splitro;
 		if (Flag(FEXPORT))
 			typeset(*wp, EXPORT, 0, 0, 0);
-		if (!setstr(vp, Xstring(cs, ccp), KSH_RETURN_ERROR)) {
-			shf_flush(shf);
-			afree(wpalloc, ATEMP);
-			return (1);
-		}
 	}
-
-	shf_flush(shf);
-	if (historyr) {
-		Xput(xs, xp, '\0');
-		histsave(&source->line, Xstring(xs, xp), true, false);
-		Xfree(xs, xp);
+	if (!setstr(vq, Xstring(xs, xp), KSH_RETURN_ERROR))
+		goto c_read_spliterr;
+	if (aschars) {
+		setint_v(vq, vq, false);
+		/* protect from UTFMODE changes */
+		vq->type = 0;
 	}
-	/*
-	 * if this is the co-process fd, close the file descriptor
-	 * (can get eof if and only if all processes are have died,
-	 * i.e. coproc.njobs is 0 and the pipe is closed).
-	 */
-	if (c == EOF && !ecode)
-		coproc_read_close(fd);
+	if (intoarray || *++wp != NULL)
+		goto c_read_splitloop;
 
-	afree(wpalloc, ATEMP);
-	return (ecode ? ecode : c == EOF);
+ c_read_splitdone:
+	/* free up */
+	afree(cp, ATEMP);
+
+ c_read_out:
+	afree(allocd, ATEMP);
+	Xfree(xs, xp);
+	if (restore_tios)
+		tcsetattr(fd, TCSADRAIN, &tios);
+	return (rv);
+#undef is_ifsws
 }
 
 int

lex.c

@@ -22,7 +22,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/lex.c,v 1.150 2011/05/07 00:51:12 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/lex.c,v 1.151 2011/05/29 02:18:52 tg Exp $");
 
 /*
  * states while lexing word
@@ -1018,6 +1018,10 @@ yylex(int cf)
 				    /* c == '(' ) */ MDPAREN;
 			else if (c == '|' && c2 == '&')
 				c = COPROC;
+			else if (c == ';' && c2 == '|')
+				c = BRKEV;
+			else if (c == ';' && c2 == '&')
+				c = BRKFT;
 			else
 				ungetsc(c2);
 		} else if (c == '\n') {

misc.c

@@ -29,7 +29,7 @@
 #include <grp.h>
 #endif
 
-__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.165 2011/05/04 23:16:02 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/misc.c,v 1.166 2011/05/29 02:18:53 tg Exp $");
 
 /* type bits for unsigned char */
 unsigned char chtypes[UCHAR_MAX + 1];
@@ -1641,14 +1641,17 @@ c_cd(const char **wp)
 	bool physical = tobool(Flag(FPHYSICAL));
 	/* was a node from cdpath added in? */
 	int cdnode;
-	/* print where we cd'd? */
-	bool printpath = false;
+	/* show where we went?, error for $PWD */
+	bool printpath = false, eflag = false;
 	struct tbl *pwd_s, *oldpwd_s;
 	XString xs;
 	char *dir, *allocd = NULL, *tryp, *pwd, *cdpath;
 
-	while ((optc = ksh_getopt(wp, &builtin_opt, "LP")) != -1)
+	while ((optc = ksh_getopt(wp, &builtin_opt, "eLP")) != -1)
 		switch (optc) {
+		case 'e':
+			eflag = true;
+			break;
 		case 'L':
 			physical = false;
 			break;
@@ -1656,13 +1659,13 @@ c_cd(const char **wp)
 			physical = true;
 			break;
 		case '?':
-			return (1);
+			return (2);
 		}
 	wp += builtin_opt.optind;
 
 	if (Flag(FRESTRICTED)) {
 		bi_errorf("restricted shell - can't cd");
-		return (1);
+		return (2);
 	}
 
 	pwd_s = global("PWD");
@@ -1672,7 +1675,7 @@ c_cd(const char **wp)
 		/* No arguments - go home */
 		if ((dir = str_val(global("HOME"))) == null) {
 			bi_errorf("no home directory (HOME not set)");
-			return (1);
+			return (2);
 		}
 	} else if (!wp[1]) {
 		/* One argument: - or dir */
@@ -1683,7 +1686,7 @@ c_cd(const char **wp)
 			dir = str_val(oldpwd_s);
 			if (dir == null) {
 				bi_errorf("no OLDPWD");
-				return (1);
+				return (2);
 			}
 			printpath = true;
 		}
@@ -1694,7 +1697,7 @@ c_cd(const char **wp)
 
 		if (!current_wd[0]) {
 			bi_errorf("can't determine current directory");
-			return (1);
+			return (2);
 		}
 		/*
 		 * substitute arg1 for arg2 in current path.
@@ -1704,7 +1707,7 @@ c_cd(const char **wp)
 		 */
 		if ((cp = strstr(current_wd, wp[0])) == NULL) {
 			bi_errorf("bad substitution");
-			return (1);
+			return (2);
 		}
 		/*-
 		 * ilen = part of current_wd before wp[0]
@@ -1723,7 +1726,7 @@ c_cd(const char **wp)
 		printpath = true;
 	} else {
 		bi_errorf("too many arguments");
-		return (1);
+		return (2);
 	}
 
 #ifdef NO_PATH_MAX
@@ -1750,9 +1753,12 @@ c_cd(const char **wp)
 		else
 			bi_errorf("%s: %s", tryp, strerror(errno));
 		afree(allocd, ATEMP);
-		return (1);
+		Xfree(xs, xp);
+		return (2);
 	}
 
+	rv = 0;
+
 	/* allocd (above) => dir, which is no longer used */
 	afree(allocd, ATEMP);
 	allocd = NULL;
@@ -1770,8 +1776,14 @@ c_cd(const char **wp)
 
 	if (Xstring(xs, xp)[0] != '/') {
 		pwd = NULL;
-	} else if (!physical || !(pwd = allocd = do_realpath(Xstring(xs, xp))))
+	} else if (!physical) {
+		goto norealpath_PWD;
+	} else if ((pwd = allocd = do_realpath(Xstring(xs, xp))) == NULL) {
+		if (eflag)
+			rv = 1;
+ norealpath_PWD:
 		pwd = Xstring(xs, xp);
+	}
 
 	/* Set PWD */
 	if (pwd) {
@@ -1784,12 +1796,15 @@ c_cd(const char **wp)
 		set_current_wd(null);
 		pwd = Xstring(xs, xp);
 		/* XXX unset $PWD? */
+		if (eflag)
+			rv = 1;
 	}
 	if (printpath || cdnode)
 		shprintf("%s\n", pwd);
 
 	afree(allocd, ATEMP);
-	return (0);
+	Xfree(xs, xp);
+	return (rv);
 }
 
 

mksh.1

@@ -1,4 +1,4 @@
-.\" $MirOS: src/bin/mksh/mksh.1,v 1.260 2011/05/06 15:41:24 tg Exp $
+.\" $MirOS: src/bin/mksh/mksh.1,v 1.261 2011/05/29 02:18:54 tg Exp $
 .\" $OpenBSD: ksh.1,v 1.140 2011/04/23 10:14:59 sobrado Exp $
 .\"-
 .\" Copyright © 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009,
@@ -72,7 +72,7 @@
 .\" with -mandoc, it might implement .Mx itself, but we want to
 .\" use our own definition. And .Dd must come *first*, always.
 .\"
-.Dd $Mdocdate: May 6 2011 $
+.Dd $Mdocdate: May 29 2011 $
 .\"
 .\" Check which macro package we use
 .\"
@@ -391,8 +391,11 @@ is used to create asynchronous pipelines;
 and
 .Ql \*(Ba\*(Ba
 are used to specify conditional execution;
-.Ql ;;
-is used in
+.Ql ;; ,
+.Ql ;&\&
+and
+.Ql ;\*(Ba\&
+are used in
 .Ic case
 statements;
 .Ql \&(( .. ))
@@ -642,10 +645,12 @@ and
 are reserved words, not meta-characters.
 .It Xo case Ar word No in
 .Oo Op \&(
-.Ar \ pattern
-.Op \*(Ba Ar pattern
-.No ... No )
-.Ar list No ;;\ \& Oc ... esac
+.Ar pattern
+.Op \*(Ba Ar pat
+.No ... Ns )
+.Ar list
+.Op ;; \*(Ba ;&\& \*(Ba ;\*(Ba\ \&
+.Oc ... esac
 .Xc
 The
 .Ic case
@@ -669,12 +674,22 @@ stripped; any space within a pattern must be quoted.
 Both the word and the
 patterns are subject to parameter, command, and arithmetic substitution, as
 well as tilde substitution.
+.Pp
 For historical reasons, open and close braces may be used instead of
 .Ic in
 and
 .Ic esac
 e.g.\&
-.Ic case $foo { *) echo bar; } .
+.Ic case $foo { *) echo bar;; } .
+.Pp
+The list terminators are
+.Ql ;;
+(terminate after the list),
+.Ql ;&\&
+(fall through into the next list) and
+.Ql ;\*(Ba\&
+(evaluate the remaining pattern-list tuples).
+.Pp
 The exit status of a
 .Ic case
 statement is that of the executed
@@ -3059,12 +3074,17 @@ option is supported as a no-op.
 .Pp
 .It Xo
 .Ic cd
-.Op Fl LP
+.Op Fl L
+.Op Ar dir
+.Xc
+.It Xo
+.Ic cd
+.Fl P Op Fl e
 .Op Ar dir
 .Xc
 .It Xo
 .Ic chdir
-.Op Fl LP
+.Op Fl eLP
 .Op Ar dir
 .Xc
 Set the working directory to
@@ -3120,15 +3140,21 @@ and
 .Ev OLDPWD
 parameters are updated to reflect the current and old working directory,
 respectively.
+If the
+.Fl e
+option is set for physical filesystem traversal, and
+.Ev PWD
+could not be set, the exit code is 1; greater than 1 if an
+error occurred, 0 otherwise.
 .Pp
 .It Xo
 .Ic cd
-.Op Fl LP
+.Op Fl eLP
 .Ar old new
 .Xc
 .It Xo
 .Ic chdir
-.Op Fl LP
+.Op Fl eLP
 .Ar old new
 .Xc
 The string
@@ -3627,36 +3653,96 @@ directories to the root directory) is printed.
 .Pp
 .It Xo
 .Ic read
-.Op Fl d Ar delimiter
-.Op Fl prsu Ns Op Ar n
-.Op Ar parameter ...
+.Op Fl A | Fl a
+.Op Fl d Ar x
+.Oo Fl N Ar z \*(Ba
+.Fl n Ar z Oc
+.Oo Fl p \*(Ba
+.Fl u Ns Op Ar n
+.Oc Op Fl t Ar n
+.Op Fl rs
+.Op Ar p ...
 .Xc
-Reads a line of input from the standard input, separates the line into fields
-using the
+Reads a line of input, separates the input into fields using the
 .Ev IFS
 parameter (see
 .Sx Substitution
-above), and assigns each field to the specified parameters.
-Lines are delimited by the first character of
-.Ar delimiter ,
-.Dv NUL
-if empty, if
-.Fl d
-was used, a newline otherwise.
-If there are more parameters than fields, the extra parameters are set to
-.Dv NULL ,
-or alternatively, if there are more fields than parameters, the last parameter
-is assigned the remaining fields (inclusive of any separating spaces).
+above), and assigns each field to the specified parameters
+.Ar p .
 If no parameters are specified, the
 .Ev REPLY
-parameter is used.
-If the input line ends in a backslash and the
-.Fl r
-option was not used, the backslash and the newline are stripped and more input
-is read.
-If no input is read,
+parameter is used to store the result.
+With the
+.Fl A
+and
+.Fl a
+options, only no or one parameter is accepted.
+If there are more parameters than fields, the extra parameters are set to
+the empty string or 0; if there are more fields than parameters, the last
+parameter is assigned the remaining fields (including the word separators).
+.Pp
+The options are as follows:
+.Bl -tag -width XuXnX
+.It Fl A
+Store the result into the parameter
+.Ar p
+(or
+.Ev REPLY )
+as array of words.
+.It Fl a
+Store the result without word splitting into the parameter
+.Ar p
+(or
+.Ev REPLY )
+as array of characters (wide characters if the
+.Ic utf8\-mode
+option is enacted, octets otherwise).
+.It Fl d Ar x
+Use the first byte of
+.Ar x ,
+.Dv NUL
+if empty, instead of the ASCII newline character as input line delimiter.
+.It Fl N Ar z
+Instead of reading till end-of-line, read exactly
+.Ar z
+bytes; less if EOF or a timeout occurs.
+.It Fl n Ar z
+Instead of reading till end-of-line, read up to
+.Ar z
+bytes but return as soon as any bytes are read, e.g.\& from a
+slow terminal device, or if EOF or a timeout occurs.
+.It Fl p
+Read from the currently active co-process, see
+.Sx Co-processes
+above for details on this.
+.It Fl u Ns Op Ar n
+Read from the file descriptor
+.Ar n
+(defaults to 0, i.e.\& standard input).
+The argument must immediately follow the option character.
+.It Fl t Ar n
+Interrupt reading after
+.Ar n
+seconds (specified as positive decimal value with an optional fractional part).
+.It Fl r
+Normally, the ASCII backslash character escapes the special
+meaning of the following character and is stripped from the input;
 .Ic read
-exits with a non-zero status.
+does not stop when encountering a backslash-newline sequence and
+does not store that newline in the result.
+This option enables raw mode, in which backslashes are not processed.
+.It Fl s
+The input line is saved to the history.
+.El
+.Pp
+If the input is a terminal, both the
+.Fl N
+and
+.Fl n
+options set it into raw mode;
+they read an entire file if \-1 is passed as
+.Ar z
+argument.
 .Pp
 The first parameter may have a question mark and a string appended to it, in
 which case the string is used as a prompt (printed to standard error before
@@ -3665,20 +3751,9 @@ any input is read) if the input is a
 (e.g.\&
 .Ic read nfoo?\*(aqnumber of foos: \*(aq ) .
 .Pp
-The
-.Fl u Ns Ar n
-and
-.Fl p
-options cause input to be read from file descriptor
-.Ar n
-.Pf ( Ar n
-defaults to 0 if omitted)
-or the current co-process (see
-.Sx Co-processes
-above for comments on this), respectively.
-If the
-.Fl s
-option is used, input is saved to the history file.
+If no input is read or a timeout occurred,
+.Ic read
+exits with a non-zero status.
 .Pp
 Another handy set of tricks:
 If
@@ -3689,6 +3764,17 @@ then leading whitespace will be removed (IFS) and backslashes processed.
 You might want to use
 .Ic while IFS= read \-r foo; do ...; done
 for pristine I/O.
+Similarily, when using the
+.Fl a
+option, use of the
+.Fl r
+option might be prudent; the same applies for:
+.Bd -literal -offset indent
+find . -type f -print0 \*(Ba \e
+    while IFS= read \-d \*(aq\*(aq \-r filename; do
+	print \-r \-\- "found <${filename#./}>"
+done
+.Ed
 .Pp
 The inner loop will be executed in a subshell and variable changes
 cannot be propagated if executed in a pipeline:
@@ -6153,7 +6239,7 @@ $ /bin/sleep 666 && echo fubar
 .Ed
 .Pp
 This document attempts to describe
-.Nm mksh\ R39c+CVS
+.Nm mksh\ R40~rc
 and up,
 compiled without any options impacting functionality, such as
 .Dv MKSH_SMALL ,

sh.h

@@ -151,9 +151,9 @@
 #endif
 
 #ifdef EXTERN
-__RCSID("$MirOS: src/bin/mksh/sh.h,v 1.469 2011/05/06 15:41:25 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/sh.h,v 1.470 2011/05/29 02:18:55 tg Exp $");
 #endif
-#define MKSH_VERSION "R39 2011/05/06"
+#define MKSH_VERSION "R39 2011/05/28"
 
 #ifndef MKSH_INCLUDES_ONLY
 
@@ -1111,10 +1111,14 @@ struct op {
 					 */
 	int lineno;			/* TCOM/TFUNC: LINENO for this */
 	short type;			/* operation type, see below */
+	/* WARNING: newtp(), tcopy() use evalflags = 0 to clear union */
 	union {
-		/* WARNING: newtp(), tcopy() use evalflags = 0 to clear union */
-		short evalflags;	/* TCOM: arg expansion eval() flags */
-		short ksh_func;		/* TFUNC: function x (vs x()) */
+		/* TCOM: arg expansion eval() flags */
+		short evalflags;
+		/* TFUNC: function x (vs x()) */
+		short ksh_func;
+		/* TPAT: termination character */
+		char charflag;
 	} u;
 };
 
@@ -1392,6 +1396,8 @@ typedef union {
 #define BANG		278	/* ! */
 #define DBRACKET	279	/* [[ .. ]] */
 #define COPROC		280	/* |& */
+#define BRKEV		281	/* ;| */
+#define BRKFT		282	/* ;& */
 #define YYERRCODE	300
 
 /* flags to yylex */
@@ -1484,6 +1490,7 @@ int x_bind(const char *, const char *, bool, bool);
 int x_bind(const char *, const char *, bool);
 #endif
 void x_init(void);
+void x_mkraw(int, struct termios *, bool);
 int x_read(char *, size_t);
 /* eval.c */
 char *substitute(const char *, int);
@@ -1779,6 +1786,7 @@ const char *skip_varname(const char *, int);
 const char *skip_wdvarname(const char *, int);
 int is_wdvarname(const char *, int);
 int is_wdvarassign(const char *);
+struct tbl *arraysearch(struct tbl *, uint32_t);
 char **makenv(void);
 void change_winsz(void);
 int array_ref_len(const char *);

syn.c

@@ -22,7 +22,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/syn.c,v 1.64 2011/05/07 00:51:12 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/syn.c,v 1.65 2011/05/29 02:18:57 tg Exp $");
 
 extern short subshell_nesting_level;
 
@@ -619,7 +619,17 @@ casepart(int endtok)
 	t->left = c_list(true);
 	/* Note: POSIX requires the ;; */
 	if ((tpeek(CONTIN|KEYWORD|ALIAS)) != endtok)
-		musthave(BREAK, CONTIN|KEYWORD|ALIAS);
+		switch (symbol) {
+		default:
+			syntaxerr(NULL);
+		case BREAK:
+		case BRKEV:
+		case BRKFT:
+			t->u.charflag =
+			    (symbol == BRKEV) ? '|' :
+			    (symbol == BRKFT) ? '&' : ';';
+			ACCEPT;
+		}
 	return (t);
 }
 
@@ -768,6 +778,8 @@ const struct tokeninfo {
 	{ "&&",		LOGAND,	false },
 	{ "||",		LOGOR,	false },
 	{ ";;",		BREAK,	false },
+	{ ";|",		BRKEV,	false },
+	{ ";&",		BRKFT,	false },
 	{ "((",		MDPAREN, false },
 	{ "|&",		COPROC,	false },
 	/* and some special cases... */
@@ -782,8 +794,8 @@ initkeywords(void)
 	struct tbl *p;
 
 	ktinit(&keywords, APERM,
-	    /* must be 80% of 2^n (currently 20 keywords) */
-	    32);
+	    /* must be 80% of 2^n (currently 28 keywords) */
+	    64);
 	for (tt = tokentab; tt->name; tt++) {
 		if (tt->reserved) {
 			p = ktenter(&keywords, tt->name, hash(tt->name));

tree.c

@@ -22,7 +22,7 @@
 
 #include "sh.h"
 
-__RCSID("$MirOS: src/bin/mksh/tree.c,v 1.48 2011/05/07 00:24:35 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/tree.c,v 1.49 2011/05/29 02:18:57 tg Exp $");
 
 #define INDENT	8
 
@@ -100,7 +100,7 @@ ptree(struct op *t, int indent, struct shf *shf)
 	case TSELECT:
 	case TFOR:
 		fptreef(shf, indent, "%s %s ",
-		    (t->type == TFOR) ? "for" : "select", t->str);
+		    (t->type == TFOR) ? "for" : T_select, t->str);
 		if (t->vars != NULL) {
 			shf_puts("in ", shf);
 			w = (const char **)t->vars;
@@ -121,7 +121,8 @@ ptree(struct op *t, int indent, struct shf *shf)
 				    (w[1] != NULL) ? '|' : ')');
 				++w;
 			}
-			fptreef(shf, indent + INDENT, "%N%T%N;;", t1->left);
+			fptreef(shf, indent + INDENT, "%N%T%N;%c", t1->left,
+			    t1->u.charflag);
 		}
 		fptreef(shf, indent, "%Nesac ");
 		break;
@@ -949,7 +950,7 @@ dumptree(struct shf *shf, struct op *t)
 			shf_putc(')', shf);
 			shf_putc('\n', shf);
 			dumptree(shf, t1->left);
-			shf_fprintf(shf, " /%d]", i++);
+			shf_fprintf(shf, " ;%c/%d]", t1->u.charflag, i++);
 		}
 		break;
 	OPEN(TWHILE)

var.c

@@ -26,7 +26,7 @@
 #include <sys/sysctl.h>
 #endif
 
-__RCSID("$MirOS: src/bin/mksh/var.c,v 1.121 2011/05/07 02:02:47 tg Exp $");
+__RCSID("$MirOS: src/bin/mksh/var.c,v 1.122 2011/05/29 02:18:57 tg Exp $");
 
 /*
  * Variables
@@ -50,7 +50,6 @@ static void setspec(struct tbl *);
 static void unsetspec(struct tbl *);
 static int getint(struct tbl *, mksh_ari_t *, bool);
 static mksh_ari_t intval(struct tbl *);
-static struct tbl *arraysearch(struct tbl *, uint32_t);
 static const char *array_index_calc(const char *, bool *, uint32_t *);
 
 uint8_t set_refflag = 0;
@@ -348,6 +347,8 @@ str_val(struct tbl *vp)
 			n = (vp->val.i < 0) ? -vp->val.i : vp->val.i;
 		base = (vp->type == 0) ? 10 : vp->type;
 
+		if (base == 1 && n == 0)
+			base = 2;
 		if (base == 1) {
 			size_t sz = 1;
 
@@ -478,8 +479,6 @@ getint(struct tbl *vp, mksh_ari_t *nump, bool arith)
 		return (vp->type);
 	}
 	s = vp->val.s + vp->type;
-	if (s == NULL)	/* redundant given initial test */
-		s = null;
 	base = 10;
 	num = 0;
 	neg = 0;
@@ -553,10 +552,12 @@ setint_v(struct tbl *vq, struct tbl *vp, bool arith)
 		return (NULL);
 	if (!(vq->flag & INTEGER) && (vq->flag & ALLOC)) {
 		vq->flag &= ~ALLOC;
+		vq->type = 0;
 		afree(vq->val.s, vq->areap);
 	}
 	vq->val.i = num;
-	if (vq->type == 0) /* default base */
+	if (vq->type == 0)
+		/* default base */
 		vq->type = base;
 	vq->flag |= ISSET|INTEGER;
 	if (vq->flag&SPECIAL)
@@ -1240,7 +1241,7 @@ unsetspec(struct tbl *vp)
  * Search for (and possibly create) a table entry starting with
  * vp, indexed by val.
  */
-static struct tbl *
+struct tbl *
 arraysearch(struct tbl *vp, uint32_t val)
 {
 	struct tbl *prev, *curr, *news;