Fix default justification width for non-printable/non-ASCII chars

Reproducer:

    unset s
    s='コーンシェル'
    echo ${#s}
    typeset -L s
    typeset -p s
    echo ${#s}

Output:

    6
    typeset -L 18 s='コーンシェル      '
    12

Expected output:

    6
    typeset -L 12 s=コーンシェル
    6

Note that the typeset -L, -R or -Z option-argument values do not
represent numbers of characters or bytes. Their purpose is to
justify. They count horizontal terminal positions. Double-width
characters, like Japanese, occupy two horizontal terminal positions
each, instead of one, and ksh takes this into account. Thus, the
justification width above is 12 though the string length in
characters is 6, becasue each character occupies two terminal
positions.

Also, control characters have no width and do not count. This is
similarly broken.

src/cmd/ksh93/sh/name.c:
- nv_newattr(): The default justification width was calculated based
  on strlen(), which counts bytes, not terminal positions. Use a loop
  with mbchar() and mbwidth() (see ast.h) to calculate the correct
  default width in terminal positions.

Resolves: #189

Author: McDutchie

NEWS

@@ -2,6 +2,11 @@ This documents significant changes in the 1.0 branch of ksh 93u+m.
 For full details, see the git log at: https://github.com/ksh93/ksh/tree/1.0
 Uppercase BUG_* IDs are shell bug IDs as used by the Modernish shell library.
 
+2024-03-10:
+
+- Fixed a longstanding bug where the default terminal width for typeset -L, -R,
+  or -Z, if not given, was miscalculated for multibyte or control characters.
+
 2024-03-07:
 
 - Fixed a bug that caused some systems to corrupt the display of multibyte

src/cmd/ksh93/include/version.h

@@ -18,7 +18,7 @@
 
 #define SH_RELEASE_FORK	"93u+m"		/* only change if you develop a new ksh93 fork */
 #define SH_RELEASE_SVER	"1.0.9-beta"	/* semantic version number: https://semver.org */
-#define SH_RELEASE_DATE	"2024-03-07"	/* must be in this format for $((.sh.version)) */
+#define SH_RELEASE_DATE	"2024-03-10"	/* must be in this format for $((.sh.version)) */
 #define SH_RELEASE_CPYR	"(c) 2020-2024 Contributors to ksh " SH_RELEASE_FORK
 
 /* Scripts sometimes field-split ${.sh.version}, so don't change amount of whitespace. */

src/cmd/ksh93/sh/name.c

@@ -2996,7 +2996,20 @@ void nv_newattr (Namval_t *np, unsigned newatts, int size)
 					ap->nelem |= ARRAY_SCAN;
 			}
 			if(size==0 && !(newatts&NV_INTEGER) && (newatts&NV_HOST)!=NV_HOST && (newatts&(NV_LJUST|NV_RJUST|NV_ZFILL)))
+			{	/*
+				 * Calculate the default terminal width for -L, -R, -Z if no numeric option-argument was given.
+				 * Note: we count terminal positions, not characters (double-width adds 2, control char adds 0)
+				 */
+				char *cq = cp;
+				wchar_t c;
+				int w;
+				n = 0;
+				mbinit();
+				while(c = mbchar(cq))
+					if ((w = mbwidth(c)) > 0)
+						n += w;
 				size = n;
+			}
 		}
 		else if(!trans)
 			_nv_unset(np,NV_EXPORT);

src/cmd/ksh93/tests/attributes.sh

@@ -845,5 +845,13 @@ do	read -r -N6 var
 	[[ $got == "$exp" ]] || err_exit "loop optimization bug with 'typeset -b' variables (expected '$exp', got '$got')"
 done <<< 'twotowthreetfourro'
 
+# ======
+# control characters should not be counted for default justification` width
+# https://github.com/ksh93/ksh/issues/189
+exp='typeset -L 5 s=$'\''1\n2\a3\t4\x[0b]5'\'
+got=$(s=$'1\n2\a3\t4\v5'; typeset -L s; typeset -p s)
+[[ $got == "$exp" ]] || err_exit "default terminal width for typeset -L incorrect" \
+	"(expected $(printf %q "$exp"); got $(printf %q "$got"))"
+
 # ======
 exit $((Errors<125?Errors:125))

src/cmd/ksh93/tests/locale.sh

@@ -2,7 +2,7 @@
 #                                                                      #
 #               This software is part of the ast package               #
 #          Copyright (c) 1982-2012 AT&T Intellectual Property          #
-#          Copyright (c) 2020-2023 Contributors to ksh 93u+m           #
+#          Copyright (c) 2020-2024 Contributors to ksh 93u+m           #
 #                      and is licensed under the                       #
 #                 Eclipse Public License, Version 2.0                  #
 #                                                                      #
@@ -454,5 +454,19 @@ then	unset LANG "${!LC_@}" i
 	fi
 fi
 
+# ======
+# double-width characters should count for two for default justification width
+# https://github.com/ksh93/ksh/issues/189
+if	((SHOPT_MULTIBYTE))
+then	unset s "${!LC_@}"
+	LANG=C.UTF-8
+	s='コーンシェル'
+	typeset -L s
+	got=$(typeset -p s; echo ${#s})
+	exp=$'typeset -L 12 s=コーンシェル\n6'  # each double-width character counts for two terminal positions
+	[[ $got == "$exp" ]] || err_exit "default terminal width for typeset -L incorrect" \
+		"(expected $(printf %q "$exp"); got $(printf %q "$got"))"
+fi
+
 # ======
 exit $((Errors<125?Errors:125))