mdocml: Autodetect a suitable locale for -Tutf8 mode, and allow

mdocml: Autodetect a suitable locale for -Tutf8 mode, and allow

[schwarze]

Log Message:
-----------
Autodetect a suitable locale for -Tutf8 mode,
and allow overriding it manually.
Based on a patch from Svyatoslav Mishyn <juef at openmailbox dot org>
tweaked by me.
The idea originally came up in a conversation with Markus Waldeck.

Modified Files:
--------------
    mdocml:
        TODO
        configure
        configure.local.example
        term_ascii.c
        test-wchar.c

Revision Data
-------------
Index: term_ascii.c
===================================================================
RCS file: /home/cvs/mdocml/mdocml/term_ascii.c,v
retrieving revision 1.53
retrieving revision 1.54
diff -Lterm_ascii.c -Lterm_ascii.c -u -p -r1.53 -r1.54
--- term_ascii.c
+++ term_ascii.c
@@ -98,7 +98,7 @@ ascii_init(enum termenc enc, const struc
 
 		v = TERMENC_LOCALE == enc ?
 		    setlocale(LC_CTYPE, "") :
-		    setlocale(LC_CTYPE, "en_US.UTF-8");
+		    setlocale(LC_CTYPE, UTF8_LOCALE);
 		if (NULL != v && MB_CUR_MAX > 1) {
 			p->enc = enc;
 			p->advance = locale_advance;
Index: test-wchar.c
===================================================================
RCS file: /home/cvs/mdocml/mdocml/test-wchar.c,v
retrieving revision 1.3
retrieving revision 1.4
diff -Ltest-wchar.c -Ltest-wchar.c -u -p -r1.3 -r1.4
--- test-wchar.c
+++ test-wchar.c
@@ -35,9 +35,9 @@ main(void)
 		return 1;
 	}
 
-	if (setlocale(LC_CTYPE, "en_US.UTF-8") == NULL) {
-		fputs("setlocale(LC_CTYPE, \"en_US.UTF-8\") failed\n",
-		    stderr);
+	if (setlocale(LC_CTYPE, UTF8_LOCALE) == NULL) {
+		fprintf(stderr, "setlocale(LC_CTYPE, \"%s\") failed\n",
+		    UTF8_LOCALE);
 		return 1;
 	}
 
Index: configure
===================================================================
RCS file: /home/cvs/mdocml/mdocml/configure,v
retrieving revision 1.47
retrieving revision 1.48
diff -Lconfigure -Lconfigure -u -p -r1.47 -r1.48
--- configure
+++ configure
@@ -35,6 +35,7 @@ echo "config.log: writing..."
 
 MANPATH_DEFAULT="/usr/share/man:/usr/X11R6/man:/usr/local/man"
 OSNAME=
+UTF8_LOCALE=
 
 CC=`printf "all:\\n\\t@echo \\\$(CC)\\n" | env -i make -f -`
 CFLAGS="-g -W -Wall -Wmissing-prototypes -Wstrict-prototypes -Wwrite-strings"
@@ -176,6 +177,23 @@ runtest() {
 	return 1
 }
 
+# Select a UTF-8 locale.
+get_locale() {
+	[ -n "${HAVE_WCHAR}" ] && [ "${HAVE_WCHAR}" -eq 0 ] && return 0
+	ismanual UTF8_LOCALE "$UTF8_LOCALE" && return 0
+	echo "UTF8_LOCALE: testing..." 1>&3
+	UTF8_LOCALE=`locale -a | grep -i '^en_US\.UTF-*8$' | head -n 1`
+	if [ -z "${UTF8_LOCALE}" ]; then
+		UTF8_LOCALE=`locale -a | grep -i '\.UTF-*8' | head -n 1`
+		[ -n "${UTF8_LOCALE}" ] || return 1
+	fi
+	echo "UTF8_LOCALE=${UTF8_LOCALE}" 1>&2
+	echo "UTF8_LOCALE=${UTF8_LOCALE}" 1>&3
+	echo 1>&3
+	return 0;
+}
+
+
 # --- library functions ---
 runtest dirent-namlen	DIRENT_NAMLEN	|| true
 runtest EFTYPE		EFTYPE		|| true
@@ -199,7 +217,15 @@ runtest strptime	STRPTIME	|| true
 runtest strsep		STRSEP		|| true
 runtest strtonum	STRTONUM	|| true
 runtest vasprintf	VASPRINTF	|| true
-runtest wchar		WCHAR		|| true
+
+# --- wide character and locale support ---
+if get_locale; then
+	runtest wchar WCHAR -DUTF8_LOCALE=\"${UTF8_LOCALE}\" || true
+else
+	HAVE_WCHAR=0
+	echo "wchar: no (no UTF8_LOCALE)" 1>&2
+	echo "wchar: no (no UTF8_LOCALE)" 1>&3
+fi
 
 # --- nanosleep ---
 if [ -n "${LD_NANOSLEEP}" ]; then
@@ -264,6 +290,7 @@ echo
 echo "#define MAN_CONF_FILE \"/etc/${MANM_MANCONF}\""
 echo "#define MANPATH_DEFAULT \"${MANPATH_DEFAULT}\""
 [ -n "${OSNAME}" ] && echo "#define OSNAME \"${OSNAME}\""
+[ -n "${UTF8_LOCALE}" ] && echo "#define UTF8_LOCALE \"${UTF8_LOCALE}\""
 [ -n "${HOMEBREWDIR}" ] && echo "#define HOMEBREWDIR \"${HOMEBREWDIR}\""
 [ ${HAVE_EFTYPE} -eq 0 ] && echo "#define EFTYPE EINVAL"
 
Index: TODO
===================================================================
RCS file: /home/cvs/mdocml/mdocml/TODO,v
retrieving revision 1.219
retrieving revision 1.220
diff -LTODO -LTODO -u -p -r1.219 -r1.220
--- TODO
+++ TODO
@@ -529,16 +529,6 @@ are mere guesses, and some may be wrong.
   loc **  exist **  algo **  size *  imp **
 
 ************************************************************************
-* portability
-************************************************************************
-
-- systems having UTF-8 but not en_US.UTF-8
-  call locale(1) from ./configure, select a UTF-8-locale,
-  and use that for test-wchar.c and term_ascii.c
-  to Markus Waldeck  Sat, 18 Jul 2015 01:55:37 +0200
-  loc *  exist *  algo *  size *  imp *
-
-************************************************************************
 * warning issues
 ************************************************************************
 
Index: configure.local.example
===================================================================
RCS file: /home/cvs/mdocml/mdocml/configure.local.example,v
retrieving revision 1.17
retrieving revision 1.18
diff -Lconfigure.local.example -Lconfigure.local.example -u -p -r1.17 -r1.18
--- configure.local.example
+++ configure.local.example
@@ -48,6 +48,13 @@ HAVE_WCHAR=1
 
 HAVE_WCHAR=0
 
+# For -Tutf8 mode, mandoc needs to set an arbitrary locale having
+# a UTF-8 character set.  If autodetection of a suitable locale
+# fails or selects an undesirable locale, you can manually choose
+# the locale for -Tutf8 mode:
+
+UTF8_LOCALE=en_US.UTF-8
+
 # When man(1) or apropos(1) is called without -m and -M options,
 # MANPATH is not set in the environment, and man.conf(5) is not
 # available, manuals are searched for in the following directory
--
 To unsubscribe send an email to source+unsubscribe@mdocml.bsd.lv