Fix for %c in multibyte locale + new tests.

8 files changed, 162 insertions, 14 deletions

diff --git a/ChangeLog b/ChangeLog
index bc49b4c0..c6a32d6d 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,8 @@
+2013-07-04         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* builtin.c (format_tree): Fixes for %c with multibyte characters
+	and field width > 1. Bugs reported by Nethox <nethox@gmail.com>.
+
 2013-07-02         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* profile.c (pp_string): Add a call to chksize and fix another.
diff --git a/builtin.c b/builtin.c
index ba1d8dcb..b8e24cb3 100644
--- a/builtin.c
+++ b/builtin.c
@@ -1097,6 +1097,7 @@ out0:
 			 * used to work? 6/2003.)
 			 */
 			cp = arg->stptr;
+			prec = 1;
 #if MBS_SUPPORT
 			/*
 			 * First character can be multiple bytes if
@@ -1108,17 +1109,14 @@ out0:
 
 				memset(& state, 0, sizeof(state));
 				count = mbrlen(cp, arg->stlen, & state);
-				if (count == 0
-				    || count == (size_t)-1
-				    || count == (size_t)-2)
-					goto out2;
-				prec = count;
-				goto pr_tail;
+				if (count > 0) {
+					prec = count;
+					/* may need to increase fw so that padding happens, see pr_tail code */
+					if (fw > 0)
+						fw += count - 1;
+				}
 			}
-out2:
-			;
 #endif
-			prec = 1;
 			goto pr_tail;
 		case 's':
 			need_format = false;
@@ -1421,9 +1419,14 @@ mpf1:
 			copy_count = prec;
 			if (fw == 0 && ! have_prec)
 				;
-			else if (gawk_mb_cur_max > 1 && (cs1 == 's' || cs1 == 'c')) {
-				assert(cp == arg->stptr || cp == cpbuf);
-				copy_count = mbc_byte_count(arg->stptr, prec);
+			else if (gawk_mb_cur_max > 1) {
+				if (cs1 == 's') {
+					assert(cp == arg->stptr || cp == cpbuf);
+					copy_count = mbc_byte_count(arg->stptr, prec);
+				}
+				/* prec was set by code for %c */
+				/* else
+					copy_count = prec; */
 			}
 			bchunk(cp, copy_count);
 			while (fw > prec) {
diff --git a/test/ChangeLog b/test/ChangeLog
index dd71a1b5..87ad809f 100644
--- a/test/ChangeLog
+++ b/test/ChangeLog
@@ -1,3 +1,9 @@
+2013-07-04         Arnold D. Robbins     <arnold@skeeve.com>
+
+	* Makefile.am (mbprintf4): New test.
+	* mbprintf4.awk, mbprintf4.in, mbprintf4.ok: New files.
+	Test cases from Nethox <nethox@gmail.com>.
+
 2013-06-27         Arnold D. Robbins     <arnold@skeeve.com>
 
 	* Makefile.am (dfamb1): New test.
diff --git a/test/Makefile.am b/test/Makefile.am
index fe6d3a59..e7c9cefd 100644
--- a/test/Makefile.am
+++ b/test/Makefile.am
@@ -493,6 +493,9 @@ EXTRA_DIST = \
 	mbprintf3.awk \
 	mbprintf3.in \
 	mbprintf3.ok \
+	mbprintf4.awk \
+	mbprintf4.in \
+	mbprintf4.ok \
 	mbstr1.awk \
 	mbstr1.ok \
 	membug1.awk \
@@ -991,7 +994,7 @@ MPFR_TESTS = mpfrnr mpfrrnd mpfrieee mpfrexprange mpfrsort mpfrbigint
 
 LOCALE_CHARSET_TESTS = \
 	asort asorti fmttest fnarydel fnparydl jarebug lc_num1 mbfw1 \
-	mbprintf1 mbprintf2 mbprintf3 rebt8b2 rtlenmb sort1 sprintfc
+	mbprintf1 mbprintf2 mbprintf3 mbprintf4 rebt8b2 rtlenmb sort1 sprintfc
 
 SHLIB_TESTS = \
 	fnmatch filefuncs fork fork2 fts functab4 inplace1 inplace2 inplace3 \
@@ -1564,6 +1567,12 @@ mbprintf3::
 	$(AWK) -f $(srcdir)/$@.awk $(srcdir)/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >> _$@
 	@-$(CMP) $(srcdir)/$@.ok _$@ && rm -f _$@
 
+mbprintf4::
+	@echo $@
+	@GAWKLOCALE=en_US.UTF-8 ; export GAWKLOCALE ; \
+	$(AWK) -f $(srcdir)/$@.awk $(srcdir)/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >> _$@
+	@-$(CMP) $(srcdir)/$@.ok _$@ && rm -f _$@
+
 mbfw1::
 	@echo $@
 	@GAWKLOCALE=en_US.UTF-8 ; export GAWKLOCALE ; \
diff --git a/test/Makefile.in b/test/Makefile.in
index 06946c0c..147eeaf3 100644
--- a/test/Makefile.in
+++ b/test/Makefile.in
@@ -711,6 +711,9 @@ EXTRA_DIST = \
 	mbprintf3.awk \
 	mbprintf3.in \
 	mbprintf3.ok \
+	mbprintf4.awk \
+	mbprintf4.in \
+	mbprintf4.ok \
 	mbstr1.awk \
 	mbstr1.ok \
 	membug1.awk \
@@ -1204,7 +1207,7 @@ MACHINE_TESTS = double1 double2 fmtspcl intformat
 MPFR_TESTS = mpfrnr mpfrrnd mpfrieee mpfrexprange mpfrsort mpfrbigint
 LOCALE_CHARSET_TESTS = \
 	asort asorti fmttest fnarydel fnparydl jarebug lc_num1 mbfw1 \
-	mbprintf1 mbprintf2 mbprintf3 rebt8b2 rtlenmb sort1 sprintfc
+	mbprintf1 mbprintf2 mbprintf3 mbprintf4 rebt8b2 rtlenmb sort1 sprintfc
 
 SHLIB_TESTS = \
 	fnmatch filefuncs fork fork2 fts functab4 inplace1 inplace2 inplace3 \
@@ -1960,6 +1963,12 @@ mbprintf3::
 	$(AWK) -f $(srcdir)/$@.awk $(srcdir)/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >> _$@
 	@-$(CMP) $(srcdir)/$@.ok _$@ && rm -f _$@
 
+mbprintf4::
+	@echo $@
+	@GAWKLOCALE=en_US.UTF-8 ; export GAWKLOCALE ; \
+	$(AWK) -f $(srcdir)/$@.awk $(srcdir)/$@.in >_$@ 2>&1 || echo EXIT CODE: $$? >> _$@
+	@-$(CMP) $(srcdir)/$@.ok _$@ && rm -f _$@
+
 mbfw1::
 	@echo $@
 	@GAWKLOCALE=en_US.UTF-8 ; export GAWKLOCALE ; \
diff --git a/test/mbprintf4.awk b/test/mbprintf4.awk
new file mode 100644
index 00000000..a4b2a218
--- /dev/null
+++ b/test/mbprintf4.awk
@@ -0,0 +1,32 @@
+# printf with multi-byte text encoding, %c and %s, width and precision, and left-alignment.
+{
+	print "printf %c " $0
+	printf "|%c|\n", $0
+	printf "|%1c|\n", $0
+	printf "|%3c|\n", $0
+	# precision is ignored by %c.
+	printf "|%3.1c|\n", $0
+	printf "|%3.5c|\n", $0
+	print "printf %-c " $0
+	printf "|%-c|\n", $0
+	printf "|%-1c|\n", $0
+	printf "|%-3c|\n", $0
+	# precision is ignored by %c.
+	printf "|%-3.1c|\n", $0
+	printf "|%-3.5c|\n", $0
+	printf ORS
+
+	print "printf %s " $0
+	printf "|%s|\n", $0
+	printf "|%1s|\n", $0
+	printf "|%3s|\n", $0
+	printf "|%3.1s|\n", $0
+	printf "|%3.5s|\n", $0
+	print "printf %-s " $0
+	printf "|%-s|\n", $0
+	printf "|%-1s|\n", $0
+	printf "|%-3s|\n", $0
+	printf "|%-3.1s|\n", $0
+	printf "|%-3.5s|\n", $0
+	printf ORS ORS
+}
diff --git a/test/mbprintf4.in b/test/mbprintf4.in
new file mode 100644
index 00000000..c93d40de
--- /dev/null
+++ b/test/mbprintf4.in
@@ -0,0 +1,3 @@
+ú
+último
+áé
diff --git a/test/mbprintf4.ok b/test/mbprintf4.ok
new file mode 100644
index 00000000..9b9dd4e2
--- /dev/null
+++ b/test/mbprintf4.ok
@@ -0,0 +1,81 @@
+printf %c ú
+|ú|
+|ú|
+|  ú|
+|  ú|
+|  ú|
+printf %-c ú
+|ú|
+|ú|
+|ú  |
+|ú  |
+|ú  |
+
+printf %s ú
+|ú|
+|ú|
+|  ú|
+|  ú|
+|  ú|
+printf %-s ú
+|ú|
+|ú|
+|ú  |
+|ú  |
+|ú  |
+
+
+printf %c último
+|ú|
+|ú|
+|  ú|
+|  ú|
+|  ú|
+printf %-c último
+|ú|
+|ú|
+|ú  |
+|ú  |
+|ú  |
+
+printf %s último
+|último|
+|último|
+|último|
+|  ú|
+|últim|
+printf %-s último
+|último|
+|último|
+|último|
+|ú  |
+|últim|
+
+
+printf %c áé
+|á|
+|á|
+|  á|
+|  á|
+|  á|
+printf %-c áé
+|á|
+|á|
+|á  |
+|á  |
+|á  |
+
+printf %s áé
+|áé|
+|áé|
+| áé|
+|  á|
+| áé|
+printf %-s áé
+|áé|
+|áé|
+|áé |
+|á  |
+|áé |
+
+