[PATCH 0/2] Make C/POSIX and C.UTF-8 consistent.

[PATCH 0/2] Make C/POSIX and C.UTF-8 consistent.

[Carlos O'Donell]

We had a recent report from Michael Hudson-Doyle that he had seen a
problem with C.UTF-8 when running tests for rrdtool.  The report
prompted Florian to place this on the glibc 2.35 blocker for review.
Upon review I decided to haromize C.UTF-8 closer to C/POSIX and I
worked with Florian to fix the discrepancies between the C.UTF-8
locale and the builtin C/POSIX locale.  The work uncoverd a problem
in the parsing of LC_MONETARY by localedef which needed fixing in order
to make C/POSIX and C.UTF-8 consistent.  The first commit fixes the
mon_decimal_point handling in localedef parsing, while the second commit
fixes C.UTF-8 and adds a new test to check for consistency beween
C/POSIX and C.UTF-8.  The test is based on work that Florian Weimer did
to help me identify the inconsistencies between the locales.

Carlos O'Donell (2):
  localedef: Fix handling of empty mon_decimal_point
  localedata: Adjust C.UTF-8 to align with C/POSIX.

 locale/programs/ld-monetary.c       |   4 +-
 localedata/Makefile                 |  30 +-
 localedata/locales/C                |  22 +-
 localedata/tst-c-utf8-consistency.c | 539 ++++++++++++++++++++++++++++
 4 files changed, 579 insertions(+), 16 deletions(-)
 create mode 100644 localedata/tst-c-utf8-consistency.c

-- 
2.31.1

[PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Carlos O'Donell]

The handling of mon_decimal_point is incorrect when it comes to
handling the empty "" value.  The existing parser in monetary_read()
will correctly handle setting the non-wide-character value and the
wide-character value e.g. STR_ELEM_WC(mon_decimal_point) if they are
set in the locale definition.  However, in monetary_finish() we have
conflicting TEST_ELEM() which sets a default value (if the locale
definition doesn't include one), and subsequent code which looks for
mon_decimal_point to be NULL to issue a specific error message and set
the defaults. The latter is unused because TEST_ELEM() always sets a
default.  The simplest solution is to remove the TEST_ELEM() check,
and allow the existing check to look to see if mon_decimal_point is
NULL and set an appropriate default.  The final fix is to move the
setting of mon_decimal_point_wc so it occurs only when
mon_decimal_point is being set to a default, keeping both values
consistent. There is no way to tell the difference between
mon_decimal_point_wc having been set to the empty string and not
having been defined at all, for that distinction we must use
mon_decimal_point being NULL or "", and so we must logically set
the default together with mon_decimal_point.

Lastly, there are more fixes similar to this that could be made to
ld-monetary.c, but we avoid that in order to fix just the code
required for mon_decimal_point, which impacts the ability for C.UTF-8
to set mon_decimal_point to "", since without this fix we end up with
an inconsistent setting of mon_decimal_point set to "", but
mon_decimal_point_wc set to "." which is incorrect.

Tested on x86_64 and i686 without regression.
---
 locale/programs/ld-monetary.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/locale/programs/ld-monetary.c b/locale/programs/ld-monetary.c
index 277b9ff042..3b0412b405 100644
--- a/locale/programs/ld-monetary.c
+++ b/locale/programs/ld-monetary.c
@@ -207,7 +207,6 @@ No definition for %s category found"), "LC_MONETARY");
 
   TEST_ELEM (int_curr_symbol, "");
   TEST_ELEM (currency_symbol, "");
-  TEST_ELEM (mon_decimal_point, ".");
   TEST_ELEM (mon_thousands_sep, "");
   TEST_ELEM (positive_sign, "");
   TEST_ELEM (negative_sign, "");
@@ -257,6 +256,7 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
 	record_error (0, 0, _("%s: field `%s' not defined"),
 		      "LC_MONETARY", "mon_decimal_point");
       monetary->mon_decimal_point = ".";
+      monetary->mon_decimal_point_wc = L'.';
     }
   else if (monetary->mon_decimal_point[0] == '\0' && ! be_quiet && ! nothing)
     {
@@ -264,8 +264,6 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
 %s: value for field `%s' must not be an empty string"),
 		    "LC_MONETARY", "mon_decimal_point");
     }
-  if (monetary->mon_decimal_point_wc == L'\0')
-    monetary->mon_decimal_point_wc = L'.';
 
   if (monetary->mon_grouping_len == 0)
     {
-- 
2.31.1

[PATCH 2/2] localedata: Adjust C.UTF-8 to align with C/POSIX.

[Carlos O'Donell]

We have had one downstream report from Canonical [1] that
an rrdtool test was broken by the differences in LC_TIME
that we had in the non-builtin C locale (C.UTF-8). If one
application has an issue there are going to be others, and
so with this commit we review and fix all the issues that
cause the builtin C locale to be different from C.UTF-8,
which includes:
* mon_decimal_point should be empty e.g. ""
 - Depends on mon_decimal_point_wc fix.
* negative_sign should be empty e.g. ""
* week should be aligned with ISO 30112 default e.g. 7;19971130;4
* d_fmt corrected with escaped slashes e.g. "%m//%d//%y"
* yesstr and nostr should be empty e.g. ""
* country_ab2 and country_ab3 should be empty e.g. ""

We bump LC_IDENTIFICATION version and adjust the date to
indicate the change in the locale.

A new tst-c-utf8-consistency test is added to ensure
consistency between C/POSIX and C.UTF-8.

Tested on x86_64 and i686 without regression.

[1] https://sourceware.org/pipermail/libc-alpha/2022-January/135703.html

Co-authored-by: Florian Weimer <fweimer@redhat.com>
---
 localedata/Makefile                 |  30 +-
 localedata/locales/C                |  22 +-
 localedata/tst-c-utf8-consistency.c | 539 ++++++++++++++++++++++++++++
 3 files changed, 578 insertions(+), 13 deletions(-)
 create mode 100644 localedata/tst-c-utf8-consistency.c

diff --git a/localedata/Makefile b/localedata/Makefile
index 79db713925..9ae2e5c161 100644
--- a/localedata/Makefile
+++ b/localedata/Makefile
@@ -155,11 +155,31 @@ locale_test_suite := tst_iswalnum tst_iswalpha tst_iswcntrl            \
 		     tst_wcsxfrm tst_wctob tst_wctomb tst_wctrans      \
 		     tst_wctype tst_wcwidth
 
-tests = $(locale_test_suite) tst-digits tst-setlocale bug-iconv-trans \
-	tst-leaks tst-mbswcs1 tst-mbswcs2 tst-mbswcs3 tst-mbswcs4 tst-mbswcs5 \
-	tst-mbswcs6 tst-xlocale1 tst-xlocale2 bug-usesetlocale \
-	tst-strfmon1 tst-sscanf bug-setlocale1 tst-setlocale2 tst-setlocale3 \
-	tst-wctype tst-iconv-math-trans
+tests = \
+  $(locale_test_suite) \
+  bug-iconv-trans \
+  bug-setlocale1 \
+  bug-usesetlocale \
+  tst-c-utf8-consistency \
+  tst-digits \
+  tst-iconv-math-trans \
+  tst-leaks \
+  tst-mbswcs1 \
+  tst-mbswcs2 \
+  tst-mbswcs3 \
+  tst-mbswcs4 \
+  tst-mbswcs5 \
+  tst-mbswcs6 \
+  tst-setlocale \
+  tst-setlocale2 \
+  tst-setlocale3 \
+  tst-sscanf \
+  tst-strfmon1 \
+  tst-wctype \
+  tst-xlocale1 \
+  tst-xlocale2 \
+  # tests
+
 tests-static = bug-setlocale1-static
 tests += $(tests-static)
 ifeq (yes,$(build-shared))
diff --git a/localedata/locales/C b/localedata/locales/C
index ca801c79cf..fb647ccc4b 100644
--- a/localedata/locales/C
+++ b/localedata/locales/C
@@ -12,8 +12,8 @@ tel        ""
 fax        ""
 language   ""
 territory  ""
-revision   "2.0"
-date       "2020-06-28"
+revision   "2.1"
+date       "2022-01-30"
 category  "i18n:2012";LC_IDENTIFICATION
 category  "i18n:2012";LC_CTYPE
 category  "i18n:2012";LC_COLLATE
@@ -68,11 +68,11 @@ LC_MONETARY
 % glibc/locale/C-monetary.c.).
 int_curr_symbol     ""
 currency_symbol     ""
-mon_decimal_point   "."
+mon_decimal_point   ""
 mon_thousands_sep   ""
 mon_grouping        -1
 positive_sign       ""
-negative_sign       "-"
+negative_sign       ""
 int_frac_digits     -1
 frac_digits         -1
 p_cs_precedes       -1
@@ -121,7 +121,9 @@ mon         "January";"February";"March";"April";"May";"June";"July";/
 %
 % ISO 8601 conforming applications should use the values 7, 19971201 (a
 % Monday), and 4 (Thursday), respectively.
-week    7;19971201;4
+%
+% This field is consciously aligned with ISO 30112 and the C/POSIX locale.
+week    7;19971130;4
 first_weekday	1
 first_workday	2
 
@@ -129,7 +131,7 @@ first_workday	2
 d_t_fmt "%a %b %e %H:%M:%S %Y"
 
 % Appropriate date representation (%x)
-d_fmt   "%m/%d/%y"
+d_fmt   "%m//%d//%y"
 
 % Appropriate time representation (%X)
 t_fmt   "%H:%M:%S"
@@ -150,8 +152,8 @@ LC_MESSAGES
 %
 yesexpr "^[yY]"
 noexpr  "^[nN]"
-yesstr  "Yes"
-nostr   "No"
+yesstr  ""
+nostr   ""
 END LC_MESSAGES
 
 LC_PAPER
@@ -175,6 +177,10 @@ LC_ADDRESS
 % the LC_ADDRESS category.
 % (also used in the built in C/POSIX locale in glibc/locale/C-address.c)
 postal_fmt    "%a%N%f%N%d%N%b%N%s %h %e %r%N%C-%z %T%N%c%N"
+% The abbreviated 2 char and 3 char should be set to empty strings to
+% match the C/POSIX locale.
+country_ab2   ""
+country_ab3   ""
 END LC_ADDRESS
 
 LC_TELEPHONE
diff --git a/localedata/tst-c-utf8-consistency.c b/localedata/tst-c-utf8-consistency.c
new file mode 100644
index 0000000000..50feed3090
--- /dev/null
+++ b/localedata/tst-c-utf8-consistency.c
@@ -0,0 +1,539 @@
+/* Test that C/POSIX and C.UTF-8 are consistent.
+   Copyright (C) 2022 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, see
+   <https://www.gnu.org/licenses/>.  */
+
+#include <langinfo.h>
+#include <locale.h>
+#include <stdbool.h>
+#include <stdio.h>
+#include <support/check.h>
+
+/* Initialized by do_test using newlocale.  */
+static locale_t c_utf8;
+
+/* Set to true for second pass.  */
+static bool use_nl_langinfo_l;
+
+static void
+switch_to_c (void)
+{
+  if (setlocale (LC_ALL, "C") == NULL)
+    FAIL_EXIT1 ("setlocale (LC_ALL, \"C\")");
+}
+
+static void
+switch_to_c_utf8 (void)
+{
+  if (setlocale (LC_ALL, "C.UTF-8") == NULL)
+    FAIL_EXIT1 ("setlocale (LC_ALL, \"C.UTF-8\")");
+}
+
+static char *
+str (nl_item item)
+{
+  if (!use_nl_langinfo_l)
+    switch_to_c  ();
+  return nl_langinfo (item);
+}
+
+static char *
+str_utf8 (nl_item item)
+{
+  if (use_nl_langinfo_l)
+    return nl_langinfo_l (item, c_utf8);
+  else
+    {
+      switch_to_c_utf8  ();
+      return nl_langinfo (item);
+    }
+}
+
+static wchar_t *
+wstr (nl_item item)
+{
+  return (wchar_t *) str (item);
+}
+
+static wchar_t *
+wstr_utf8 (nl_item item)
+{
+  return (wchar_t *) str_utf8 (item);
+}
+
+static int
+byte (nl_item item)
+{
+  return (signed char) *str (item);
+}
+
+static int
+byte_utf8 (nl_item item)
+{
+  return (signed char) *str_utf8 (item);
+}
+
+static int
+word (nl_item item)
+{
+  union
+  {
+    char *ptr;
+    int word;
+  } u;
+  u.ptr = str (item);
+  return u.word;
+}
+
+static int
+word_utf8 (nl_item item)
+{
+  union
+  {
+    char *ptr;
+    int word;
+  } u;
+  u.ptr = str_utf8 (item);
+  return u.word;
+}
+
+static void
+one_pass (void)
+{
+  /* LC_TIME.  */
+  TEST_COMPARE_STRING (str (ABDAY_1), str_utf8 (ABDAY_1));
+  TEST_COMPARE_STRING (str (ABDAY_2), str_utf8 (ABDAY_2));
+  TEST_COMPARE_STRING (str (ABDAY_3), str_utf8 (ABDAY_3));
+  TEST_COMPARE_STRING (str (ABDAY_4), str_utf8 (ABDAY_4));
+  TEST_COMPARE_STRING (str (ABDAY_5), str_utf8 (ABDAY_5));
+  TEST_COMPARE_STRING (str (ABDAY_6), str_utf8 (ABDAY_6));
+  TEST_COMPARE_STRING (str (ABDAY_7), str_utf8 (ABDAY_7));
+
+  TEST_COMPARE_STRING (str (DAY_1), str_utf8 (DAY_1));
+  TEST_COMPARE_STRING (str (DAY_2), str_utf8 (DAY_2));
+  TEST_COMPARE_STRING (str (DAY_3), str_utf8 (DAY_3));
+  TEST_COMPARE_STRING (str (DAY_4), str_utf8 (DAY_4));
+  TEST_COMPARE_STRING (str (DAY_5), str_utf8 (DAY_5));
+  TEST_COMPARE_STRING (str (DAY_6), str_utf8 (DAY_6));
+  TEST_COMPARE_STRING (str (DAY_7), str_utf8 (DAY_7));
+
+  TEST_COMPARE_STRING (str (ABMON_1), str_utf8 (ABMON_1));
+  TEST_COMPARE_STRING (str (ABMON_2), str_utf8 (ABMON_2));
+  TEST_COMPARE_STRING (str (ABMON_3), str_utf8 (ABMON_3));
+  TEST_COMPARE_STRING (str (ABMON_4), str_utf8 (ABMON_4));
+  TEST_COMPARE_STRING (str (ABMON_5), str_utf8 (ABMON_5));
+  TEST_COMPARE_STRING (str (ABMON_6), str_utf8 (ABMON_6));
+  TEST_COMPARE_STRING (str (ABMON_7), str_utf8 (ABMON_7));
+  TEST_COMPARE_STRING (str (ABMON_8), str_utf8 (ABMON_8));
+  TEST_COMPARE_STRING (str (ABMON_9), str_utf8 (ABMON_9));
+  TEST_COMPARE_STRING (str (ABMON_10), str_utf8 (ABMON_10));
+  TEST_COMPARE_STRING (str (ABMON_11), str_utf8 (ABMON_11));
+  TEST_COMPARE_STRING (str (ABMON_12), str_utf8 (ABMON_12));
+
+  TEST_COMPARE_STRING (str (MON_1), str_utf8 (MON_1));
+  TEST_COMPARE_STRING (str (MON_2), str_utf8 (MON_2));
+  TEST_COMPARE_STRING (str (MON_3), str_utf8 (MON_3));
+  TEST_COMPARE_STRING (str (MON_4), str_utf8 (MON_4));
+  TEST_COMPARE_STRING (str (MON_5), str_utf8 (MON_5));
+  TEST_COMPARE_STRING (str (MON_6), str_utf8 (MON_6));
+  TEST_COMPARE_STRING (str (MON_7), str_utf8 (MON_7));
+  TEST_COMPARE_STRING (str (MON_8), str_utf8 (MON_8));
+  TEST_COMPARE_STRING (str (MON_9), str_utf8 (MON_9));
+  TEST_COMPARE_STRING (str (MON_10), str_utf8 (MON_10));
+  TEST_COMPARE_STRING (str (MON_11), str_utf8 (MON_11));
+  TEST_COMPARE_STRING (str (MON_12), str_utf8 (MON_12));
+
+  TEST_COMPARE_STRING (str (AM_STR), str_utf8 (AM_STR));
+  TEST_COMPARE_STRING (str (PM_STR), str_utf8 (PM_STR));
+
+  TEST_COMPARE_STRING (str (D_T_FMT), str_utf8 (D_T_FMT));
+  TEST_COMPARE_STRING (str (D_FMT), str_utf8 (D_FMT));
+  TEST_COMPARE_STRING (str (T_FMT), str_utf8 (T_FMT));
+  TEST_COMPARE_STRING (str (T_FMT_AMPM),
+                       str_utf8 (T_FMT_AMPM));
+
+  TEST_COMPARE_STRING (str (ERA), str_utf8 (ERA));
+  TEST_COMPARE_STRING (str (ERA_YEAR), str_utf8 (ERA_YEAR));
+  TEST_COMPARE_STRING (str (ERA_D_FMT), str_utf8 (ERA_D_FMT));
+  TEST_COMPARE_STRING (str (ALT_DIGITS), str_utf8 (ALT_DIGITS));
+  TEST_COMPARE_STRING (str (ERA_D_T_FMT), str_utf8 (ERA_D_T_FMT));
+  TEST_COMPARE_STRING (str (ERA_T_FMT), str_utf8 (ERA_T_FMT));
+  TEST_COMPARE (word (_NL_TIME_ERA_NUM_ENTRIES),
+                word_utf8 (_NL_TIME_ERA_NUM_ENTRIES));
+  /* No array elements, so nothing to compare for _NL_TIME_ERA_ENTRIES.  */
+  TEST_COMPARE (word (_NL_TIME_ERA_NUM_ENTRIES), 0);
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_1), wstr_utf8 (_NL_WABDAY_1));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_2), wstr_utf8 (_NL_WABDAY_2));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_3), wstr_utf8 (_NL_WABDAY_3));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_4), wstr_utf8 (_NL_WABDAY_4));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_5), wstr_utf8 (_NL_WABDAY_5));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_6), wstr_utf8 (_NL_WABDAY_6));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABDAY_7), wstr_utf8 (_NL_WABDAY_7));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_1), wstr_utf8 (_NL_WDAY_1));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_2), wstr_utf8 (_NL_WDAY_2));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_3), wstr_utf8 (_NL_WDAY_3));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_4), wstr_utf8 (_NL_WDAY_4));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_5), wstr_utf8 (_NL_WDAY_5));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_6), wstr_utf8 (_NL_WDAY_6));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WDAY_7), wstr_utf8 (_NL_WDAY_7));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_1), wstr_utf8 (_NL_WABMON_1));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_2), wstr_utf8 (_NL_WABMON_2));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_3), wstr_utf8 (_NL_WABMON_3));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_4), wstr_utf8 (_NL_WABMON_4));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_5), wstr_utf8 (_NL_WABMON_5));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_6), wstr_utf8 (_NL_WABMON_6));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_7), wstr_utf8 (_NL_WABMON_7));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_8), wstr_utf8 (_NL_WABMON_8));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_9), wstr_utf8 (_NL_WABMON_9));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_10), wstr_utf8 (_NL_WABMON_10));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_11), wstr_utf8 (_NL_WABMON_11));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABMON_12), wstr_utf8 (_NL_WABMON_12));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_1), wstr_utf8 (_NL_WMON_1));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_2), wstr_utf8 (_NL_WMON_2));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_3), wstr_utf8 (_NL_WMON_3));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_4), wstr_utf8 (_NL_WMON_4));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_5), wstr_utf8 (_NL_WMON_5));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_6), wstr_utf8 (_NL_WMON_6));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_7), wstr_utf8 (_NL_WMON_7));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_8), wstr_utf8 (_NL_WMON_8));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_9), wstr_utf8 (_NL_WMON_9));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_10), wstr_utf8 (_NL_WMON_10));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_11), wstr_utf8 (_NL_WMON_11));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WMON_12), wstr_utf8 (_NL_WMON_12));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WAM_STR), wstr_utf8 (_NL_WAM_STR));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WPM_STR), wstr_utf8 (_NL_WPM_STR));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WD_T_FMT), wstr_utf8 (_NL_WD_T_FMT));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WD_FMT), wstr_utf8 (_NL_WD_FMT));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WT_FMT), wstr_utf8 (_NL_WT_FMT));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WT_FMT_AMPM),
+                            wstr_utf8 (_NL_WT_FMT_AMPM));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WERA_YEAR), wstr_utf8 (_NL_WERA_YEAR));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WERA_D_FMT), wstr_utf8 (_NL_WERA_D_FMT));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALT_DIGITS),
+                            wstr_utf8 (_NL_WALT_DIGITS));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WERA_D_T_FMT),
+                            wstr_utf8 (_NL_WERA_D_T_FMT));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WERA_T_FMT), wstr_utf8 (_NL_WERA_T_FMT));
+
+  /* This is somewhat inconsistent, but see locale/categories.def.  */
+  TEST_COMPARE (byte (_NL_TIME_WEEK_NDAYS), byte_utf8 (_NL_TIME_WEEK_NDAYS));
+  TEST_COMPARE (word (_NL_TIME_WEEK_1STDAY),
+                word_utf8 (_NL_TIME_WEEK_1STDAY));
+  TEST_COMPARE (byte (_NL_TIME_WEEK_1STWEEK),
+                byte_utf8 (_NL_TIME_WEEK_1STWEEK));
+  TEST_COMPARE (byte (_NL_TIME_FIRST_WEEKDAY),
+                byte_utf8 (_NL_TIME_FIRST_WEEKDAY));
+  TEST_COMPARE (byte (_NL_TIME_FIRST_WORKDAY),
+                byte_utf8 (_NL_TIME_FIRST_WORKDAY));
+  TEST_COMPARE (byte (_NL_TIME_CAL_DIRECTION),
+                byte_utf8 (_NL_TIME_CAL_DIRECTION));
+  TEST_COMPARE_STRING (str (_NL_TIME_TIMEZONE), str_utf8 (_NL_TIME_TIMEZONE));
+
+  TEST_COMPARE_STRING (str (_DATE_FMT), str_utf8 (_DATE_FMT));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_W_DATE_FMT), wstr_utf8 (_NL_W_DATE_FMT));
+
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_TIME_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_TIME_CODESET), "UTF-8");
+
+  TEST_COMPARE_STRING (str (ALTMON_1), str_utf8 (ALTMON_1));
+  TEST_COMPARE_STRING (str (ALTMON_2), str_utf8 (ALTMON_2));
+  TEST_COMPARE_STRING (str (ALTMON_3), str_utf8 (ALTMON_3));
+  TEST_COMPARE_STRING (str (ALTMON_4), str_utf8 (ALTMON_4));
+  TEST_COMPARE_STRING (str (ALTMON_5), str_utf8 (ALTMON_5));
+  TEST_COMPARE_STRING (str (ALTMON_6), str_utf8 (ALTMON_6));
+  TEST_COMPARE_STRING (str (ALTMON_7), str_utf8 (ALTMON_7));
+  TEST_COMPARE_STRING (str (ALTMON_8), str_utf8 (ALTMON_8));
+  TEST_COMPARE_STRING (str (ALTMON_9), str_utf8 (ALTMON_9));
+  TEST_COMPARE_STRING (str (ALTMON_10), str_utf8 (ALTMON_10));
+  TEST_COMPARE_STRING (str (ALTMON_11), str_utf8 (ALTMON_11));
+  TEST_COMPARE_STRING (str (ALTMON_12), str_utf8 (ALTMON_12));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_1), wstr_utf8 (_NL_WALTMON_1));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_2), wstr_utf8 (_NL_WALTMON_2));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_3), wstr_utf8 (_NL_WALTMON_3));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_4), wstr_utf8 (_NL_WALTMON_4));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_5), wstr_utf8 (_NL_WALTMON_5));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_6), wstr_utf8 (_NL_WALTMON_6));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_7), wstr_utf8 (_NL_WALTMON_7));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_8), wstr_utf8 (_NL_WALTMON_8));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_9), wstr_utf8 (_NL_WALTMON_9));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_10), wstr_utf8 (_NL_WALTMON_10));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_11), wstr_utf8 (_NL_WALTMON_11));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WALTMON_12), wstr_utf8 (_NL_WALTMON_12));
+
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_1), str_utf8 (_NL_ABALTMON_1));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_2), str_utf8 (_NL_ABALTMON_2));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_3), str_utf8 (_NL_ABALTMON_3));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_4), str_utf8 (_NL_ABALTMON_4));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_5), str_utf8 (_NL_ABALTMON_5));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_6), str_utf8 (_NL_ABALTMON_6));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_7), str_utf8 (_NL_ABALTMON_7));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_8), str_utf8 (_NL_ABALTMON_8));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_9), str_utf8 (_NL_ABALTMON_9));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_10), str_utf8 (_NL_ABALTMON_10));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_11), str_utf8 (_NL_ABALTMON_11));
+  TEST_COMPARE_STRING (str (_NL_ABALTMON_12), str_utf8 (_NL_ABALTMON_12));
+
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_1),
+                            wstr_utf8 (_NL_WABALTMON_1));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_2),
+                            wstr_utf8 (_NL_WABALTMON_2));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_3),
+                            wstr_utf8 (_NL_WABALTMON_3));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_4),
+                            wstr_utf8 (_NL_WABALTMON_4));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_5),
+                            wstr_utf8 (_NL_WABALTMON_5));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_6),
+                            wstr_utf8 (_NL_WABALTMON_6));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_7),
+                            wstr_utf8 (_NL_WABALTMON_7));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_8),
+                            wstr_utf8 (_NL_WABALTMON_8));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_9),
+                            wstr_utf8 (_NL_WABALTMON_9));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_10),
+                            wstr_utf8 (_NL_WABALTMON_10));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_11),
+                            wstr_utf8 (_NL_WABALTMON_11));
+  TEST_COMPARE_STRING_WIDE (wstr (_NL_WABALTMON_12),
+                            wstr_utf8 (_NL_WABALTMON_12));
+
+  /* LC_COLLATE.  Mostly untested, only expected differences.  */
+  TEST_COMPARE_STRING (str (_NL_COLLATE_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_COLLATE_CODESET), "UTF-8");
+
+  /* LC_CTYPE.  Mostly untested, only expected differences.  */
+  TEST_COMPARE_STRING (str (CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (CODESET), "UTF-8");
+
+  /* LC_MONETARY.  */
+  TEST_COMPARE_STRING (str (INT_CURR_SYMBOL), str_utf8 (INT_CURR_SYMBOL));
+  TEST_COMPARE_STRING (str (CURRENCY_SYMBOL), str_utf8 (CURRENCY_SYMBOL));
+  TEST_COMPARE_STRING (str (MON_DECIMAL_POINT), str_utf8 (MON_DECIMAL_POINT));
+  TEST_COMPARE_STRING (str (MON_THOUSANDS_SEP), str_utf8 (MON_THOUSANDS_SEP));
+  TEST_COMPARE_STRING (str (MON_GROUPING), str_utf8 (MON_GROUPING));
+  TEST_COMPARE_STRING (str (POSITIVE_SIGN), str_utf8 (POSITIVE_SIGN));
+  TEST_COMPARE_STRING (str (NEGATIVE_SIGN), str_utf8 (NEGATIVE_SIGN));
+  TEST_COMPARE (byte (INT_FRAC_DIGITS), byte_utf8 (INT_FRAC_DIGITS));
+  TEST_COMPARE (byte (FRAC_DIGITS), byte_utf8 (FRAC_DIGITS));
+  TEST_COMPARE (byte (P_CS_PRECEDES), byte_utf8 (P_CS_PRECEDES));
+  TEST_COMPARE (byte (P_SEP_BY_SPACE), byte_utf8 (P_SEP_BY_SPACE));
+  TEST_COMPARE (byte (N_CS_PRECEDES), byte_utf8 (N_CS_PRECEDES));
+  TEST_COMPARE (byte (N_SEP_BY_SPACE), byte_utf8 (N_SEP_BY_SPACE));
+  TEST_COMPARE (byte (P_SIGN_POSN), byte_utf8 (P_SIGN_POSN));
+  TEST_COMPARE (byte (N_SIGN_POSN), byte_utf8 (N_SIGN_POSN));
+  TEST_COMPARE_STRING (str (CRNCYSTR), str_utf8 (CRNCYSTR));
+  TEST_COMPARE (byte (INT_P_CS_PRECEDES), byte_utf8 (INT_P_CS_PRECEDES));
+  TEST_COMPARE (byte (INT_P_SEP_BY_SPACE), byte_utf8 (INT_P_SEP_BY_SPACE));
+  TEST_COMPARE (byte (INT_N_CS_PRECEDES), byte_utf8 (INT_N_CS_PRECEDES));
+  TEST_COMPARE (byte (INT_N_SEP_BY_SPACE), byte_utf8 (INT_N_SEP_BY_SPACE));
+  TEST_COMPARE (byte (INT_P_SIGN_POSN), byte_utf8 (INT_P_SIGN_POSN));
+  TEST_COMPARE (byte (INT_N_SIGN_POSN), byte_utf8 (INT_N_SIGN_POSN));
+  TEST_COMPARE_STRING (str (_NL_MONETARY_DUO_INT_CURR_SYMBOL),
+                       str_utf8 (_NL_MONETARY_DUO_INT_CURR_SYMBOL));
+  TEST_COMPARE_STRING (str (_NL_MONETARY_DUO_CURRENCY_SYMBOL),
+                       str_utf8 (_NL_MONETARY_DUO_CURRENCY_SYMBOL));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_FRAC_DIGITS),
+                byte_utf8 (_NL_MONETARY_DUO_INT_FRAC_DIGITS));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_FRAC_DIGITS),
+                byte_utf8 (_NL_MONETARY_DUO_FRAC_DIGITS));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_P_CS_PRECEDES),
+                byte_utf8 (_NL_MONETARY_DUO_P_CS_PRECEDES));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_P_SEP_BY_SPACE),
+                byte_utf8 (_NL_MONETARY_DUO_P_SEP_BY_SPACE));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_N_CS_PRECEDES),
+                byte_utf8 (_NL_MONETARY_DUO_N_CS_PRECEDES));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_N_SEP_BY_SPACE),
+                byte_utf8 (_NL_MONETARY_DUO_N_SEP_BY_SPACE));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_P_CS_PRECEDES),
+                byte_utf8 (_NL_MONETARY_DUO_INT_P_CS_PRECEDES));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_P_SEP_BY_SPACE),
+                byte_utf8 (_NL_MONETARY_DUO_INT_P_SEP_BY_SPACE));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_N_CS_PRECEDES),
+                byte_utf8 (_NL_MONETARY_DUO_INT_N_CS_PRECEDES));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_N_SEP_BY_SPACE),
+                byte_utf8 (_NL_MONETARY_DUO_INT_N_SEP_BY_SPACE));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_P_SIGN_POSN),
+                byte_utf8 (_NL_MONETARY_DUO_INT_P_SIGN_POSN));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_N_SIGN_POSN),
+                byte_utf8 (_NL_MONETARY_DUO_INT_N_SIGN_POSN));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_P_SIGN_POSN),
+                byte_utf8 (_NL_MONETARY_DUO_P_SIGN_POSN));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_N_SIGN_POSN),
+                byte_utf8 (_NL_MONETARY_DUO_N_SIGN_POSN));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_P_SIGN_POSN),
+                byte_utf8 (_NL_MONETARY_DUO_INT_P_SIGN_POSN));
+  TEST_COMPARE (byte (_NL_MONETARY_DUO_INT_N_SIGN_POSN),
+                byte_utf8 (_NL_MONETARY_DUO_INT_N_SIGN_POSN));
+  TEST_COMPARE (word (_NL_MONETARY_UNO_VALID_FROM),
+                word_utf8 (_NL_MONETARY_UNO_VALID_FROM));
+  TEST_COMPARE (word (_NL_MONETARY_UNO_VALID_TO),
+                word_utf8 (_NL_MONETARY_UNO_VALID_TO));
+  TEST_COMPARE (word (_NL_MONETARY_DUO_VALID_FROM),
+                word_utf8 (_NL_MONETARY_DUO_VALID_FROM));
+  TEST_COMPARE (word (_NL_MONETARY_DUO_VALID_TO),
+                word_utf8 (_NL_MONETARY_DUO_VALID_TO));
+  /* _NL_MONETARY_CONVERSION_RATE cannot be tested (word array).  */
+  TEST_COMPARE (word (_NL_MONETARY_DECIMAL_POINT_WC),
+                word_utf8 (_NL_MONETARY_DECIMAL_POINT_WC));
+  TEST_COMPARE (word (_NL_MONETARY_THOUSANDS_SEP_WC),
+                word_utf8 (_NL_MONETARY_THOUSANDS_SEP_WC));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_MONETARY_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_MONETARY_CODESET), "UTF-8");
+
+  /* LC_NUMERIC.  */
+
+  TEST_COMPARE_STRING (str (DECIMAL_POINT), str_utf8 (DECIMAL_POINT));
+  TEST_COMPARE_STRING (str (RADIXCHAR), str_utf8 (RADIXCHAR));
+  TEST_COMPARE_STRING (str (THOUSANDS_SEP), str_utf8 (THOUSANDS_SEP));
+  TEST_COMPARE_STRING (str (THOUSEP), str_utf8 (THOUSEP));
+  TEST_COMPARE_STRING (str (GROUPING), str_utf8 (GROUPING));
+  TEST_COMPARE (word (_NL_NUMERIC_DECIMAL_POINT_WC),
+                word_utf8 (_NL_NUMERIC_DECIMAL_POINT_WC));
+  TEST_COMPARE (word (_NL_NUMERIC_THOUSANDS_SEP_WC),
+                word_utf8 (_NL_NUMERIC_THOUSANDS_SEP_WC));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_NUMERIC_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_NUMERIC_CODESET), "UTF-8");
+
+  /* LC_MESSAGES.  */
+
+  TEST_COMPARE_STRING (str (YESEXPR), str_utf8 (YESEXPR));
+  TEST_COMPARE_STRING (str (NOEXPR), str_utf8 (NOEXPR));
+  TEST_COMPARE_STRING (str (YESSTR), str_utf8 (YESSTR));
+  TEST_COMPARE_STRING (str (NOSTR), str_utf8 (NOSTR));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_MESSAGES_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_MESSAGES_CODESET), "UTF-8");
+
+  /* LC_PAPER.  */
+
+  TEST_COMPARE (word (_NL_PAPER_HEIGHT), word_utf8 (_NL_PAPER_HEIGHT));
+  TEST_COMPARE (word (_NL_PAPER_WIDTH), word_utf8 (_NL_PAPER_WIDTH));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_PAPER_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_PAPER_CODESET), "UTF-8");
+
+  /* LC_NAME.  */
+
+  TEST_COMPARE_STRING (str (_NL_NAME_NAME_FMT),
+                       str_utf8 (_NL_NAME_NAME_FMT));
+  TEST_COMPARE_STRING (str (_NL_NAME_NAME_GEN),
+                       str_utf8 (_NL_NAME_NAME_GEN));
+  TEST_COMPARE_STRING (str (_NL_NAME_NAME_MR),
+                       str_utf8 (_NL_NAME_NAME_MR));
+  TEST_COMPARE_STRING (str (_NL_NAME_NAME_MRS),
+                       str_utf8 (_NL_NAME_NAME_MRS));
+  TEST_COMPARE_STRING (str (_NL_NAME_NAME_MISS),
+                       str_utf8 (_NL_NAME_NAME_MISS));
+  TEST_COMPARE_STRING (str (_NL_NAME_NAME_MS),
+                       str_utf8 (_NL_NAME_NAME_MS));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_NAME_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_NAME_CODESET), "UTF-8");
+
+  /* LC_ADDRESS.  */
+
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_POSTAL_FMT),
+                       str_utf8 (_NL_ADDRESS_POSTAL_FMT));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_COUNTRY_NAME),
+                       str_utf8 (_NL_ADDRESS_COUNTRY_NAME));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_COUNTRY_POST),
+                       str_utf8 (_NL_ADDRESS_COUNTRY_POST));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_COUNTRY_AB2),
+                       str_utf8 (_NL_ADDRESS_COUNTRY_AB2));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_COUNTRY_AB3),
+                       str_utf8 (_NL_ADDRESS_COUNTRY_AB3));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_COUNTRY_CAR),
+                       str_utf8 (_NL_ADDRESS_COUNTRY_CAR));
+  TEST_COMPARE (word (_NL_ADDRESS_COUNTRY_NUM),
+                word_utf8 (_NL_ADDRESS_COUNTRY_NUM));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_COUNTRY_ISBN),
+                       str_utf8 (_NL_ADDRESS_COUNTRY_ISBN));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_LANG_NAME),
+                       str_utf8 (_NL_ADDRESS_LANG_NAME));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_LANG_AB),
+                       str_utf8 (_NL_ADDRESS_LANG_AB));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_LANG_TERM),
+                       str_utf8 (_NL_ADDRESS_LANG_TERM));
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_LANG_LIB),
+                       str_utf8 (_NL_ADDRESS_LANG_LIB));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_ADDRESS_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_ADDRESS_CODESET), "UTF-8");
+
+  /* LC_TELEPHONE.  */
+
+  TEST_COMPARE_STRING (str (_NL_TELEPHONE_TEL_INT_FMT),
+                       str_utf8 (_NL_TELEPHONE_TEL_INT_FMT));
+  TEST_COMPARE_STRING (str (_NL_TELEPHONE_TEL_DOM_FMT),
+                       str_utf8 (_NL_TELEPHONE_TEL_DOM_FMT));
+  TEST_COMPARE_STRING (str (_NL_TELEPHONE_INT_SELECT),
+                       str_utf8 (_NL_TELEPHONE_INT_SELECT));
+  TEST_COMPARE_STRING (str (_NL_TELEPHONE_INT_PREFIX),
+                       str_utf8 (_NL_TELEPHONE_INT_PREFIX));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_TELEPHONE_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_TELEPHONE_CODESET), "UTF-8");
+
+  /* LC_MEASUREMENT.  */
+
+  TEST_COMPARE (byte (_NL_MEASUREMENT_MEASUREMENT),
+                byte_utf8 (_NL_MEASUREMENT_MEASUREMENT));
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_MEASUREMENT_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_MEASUREMENT_CODESET), "UTF-8");
+
+  /* LC_IDENTIFICATION is skipped since C.UTF-8 is distinct from C.  */
+
+  /* _NL_IDENTIFICATION_CATEGORY cannot be tested because it is a
+     string array.  */
+  /* Expected difference.  */
+  TEST_COMPARE_STRING (str (_NL_IDENTIFICATION_CODESET), "ANSI_X3.4-1968");
+  TEST_COMPARE_STRING (str_utf8 (_NL_IDENTIFICATION_CODESET), "UTF-8");
+}
+
+static int
+do_test (void)
+{
+  puts ("info: using setlocale and nl_langinfo");
+  one_pass ();
+
+  puts ("info: using nl_langinfo_l");
+
+  c_utf8 = newlocale (LC_ALL_MASK, "C.UTF-8", (locale_t) 0);
+  TEST_VERIFY_EXIT (c_utf8 != (locale_t) 0);
+
+  switch_to_c ();
+  use_nl_langinfo_l = true;
+  one_pass ();
+
+  freelocale (c_utf8);
+
+  return 0;
+}
+
+#include <support/test-driver.c>
-- 
2.31.1

Re: [PATCH 2/2] localedata: Adjust C.UTF-8 to align with C/POSIX.

[Andreas Schwab]

On Jan 31 2022, Carlos O'Donell via Libc-alpha wrote:

> +%
> +% This field is consciously aligned with ISO 30112 and the C/POSIX locale.
> +week    7;19971130;4

The copy of ISO 30112 that I could find says 7;19971201;4.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Florian Weimer]

* Carlos O'Donell:

> diff --git a/locale/programs/ld-monetary.c b/locale/programs/ld-monetary.c
> index 277b9ff042..3b0412b405 100644
> --- a/locale/programs/ld-monetary.c
> +++ b/locale/programs/ld-monetary.c
> @@ -207,7 +207,6 @@ No definition for %s category found"), "LC_MONETARY");
>  
>    TEST_ELEM (int_curr_symbol, "");
>    TEST_ELEM (currency_symbol, "");
> -  TEST_ELEM (mon_decimal_point, ".");
>    TEST_ELEM (mon_thousands_sep, "");
>    TEST_ELEM (positive_sign, "");
>    TEST_ELEM (negative_sign, "");
> @@ -257,6 +256,7 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>  	record_error (0, 0, _("%s: field `%s' not defined"),
>  		      "LC_MONETARY", "mon_decimal_point");
>        monetary->mon_decimal_point = ".";
> +      monetary->mon_decimal_point_wc = L'.';
>      }
>    else if (monetary->mon_decimal_point[0] == '\0' && ! be_quiet && ! nothing)
>      {
> @@ -264,8 +264,6 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>  %s: value for field `%s' must not be an empty string"),
>  		    "LC_MONETARY", "mon_decimal_point");
>      }
> -  if (monetary->mon_decimal_point_wc == L'\0')
> -    monetary->mon_decimal_point_wc = L'.';
>  
>    if (monetary->mon_grouping_len == 0)
>      {

There's an existing comment

  /* The decimal point must not be empty.  This is not said explicitly
     in POSIX but ANSI C (ISO/IEC 9899) says in 4.4.2.1 it has to be
     != "".  */

that says that empty strings/null characters are invalid.
The comment was clearly copied from locale/programs/ld-numeric.c.

*However* we have got this code in stdio-common/printf_fp.c:

      decimal = _nl_lookup (loc, LC_MONETARY, MON_DECIMAL_POINT);
      if (*decimal == '\0')
	decimal = _nl_lookup (loc, LC_NUMERIC, DECIMAL_POINT);
      decimalwc = _nl_lookup_word (loc, LC_MONETARY,
				    _NL_MONETARY_DECIMAL_POINT_WC);
      if (decimalwc == L'\0')
	decimalwc = _nl_lookup_word (loc, LC_NUMERIC,
				      _NL_NUMERIC_DECIMAL_POINT_WC);

So we use LC_NUMERIC as the fallback, and our strfmon implementation is
okay with it.  But our localeconv implementation lacks this fallback,
which looks like a bug because the built-in C locale uses an empty
string/a null character.

Still I think simplifying the locale data is the right direction here.

Thanks,
Florian

Re: [PATCH 2/2] localedata: Adjust C.UTF-8 to align with C/POSIX.

[Carlos O'Donell]

On 1/31/22 03:47, Andreas Schwab wrote:
> On Jan 31 2022, Carlos O'Donell via Libc-alpha wrote:
> 
>> +%
>> +% This field is consciously aligned with ISO 30112 and the C/POSIX locale.
>> +week    7;19971130;4
> 
> The copy of ISO 30112 that I could find says 7;19971201;4.
> 

First and foremost we try to align with C/POSIX builtin locale.

Aligning with C/POSIX ensures that application developers can seamlessly
change from "C" to "C.UTF-8" without breaking existing tests.

e.g. locale/C-time.c
135     { .string = "\7" },
136     { .word = 19971130 },
137     { .string = "\4" },

I'm using ISO 30112 WD12 (2018-02-12), which I have access to as part
of my SC22 involvement. This version would go on to become the final
published standard 2020-09, though I don't yet have a copy of this
version.

Section 4.7 "LC_TIME" under week:
~~~
If the keyword is not specified the values are taken as 7,
19971130 (a Sunday), and 7 (Saturday), respectively.
ISO 8601 conforming applications should use the values
7, 19971201 (a Monday), and 4 (Thursday),
respectively. 
~~~

This matches the ISO 30112 WD10 [2014] that was used when creating
the defaults in ld-time.c:

482   /* Set up defaults based on ISO 30112 WD10 [2014].  */
483   if (time->week_ndays == 0)
484     time->week_ndays = 7;
485 
486   if (time->week_1stday == 0)
487     time->week_1stday = 19971130;
488 
489   if (time->week_1stweek == 0)
490     time->week_1stweek = 7;

This also matches the legacy withdrawn ISO/IEC 14652:2002.

I could change the comment to read:

% This field is consciously aligned with the builtin C/POSIX locale.

Would a new comment resolve your review?

-- 
Cheers,
Carlos.

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Andreas Schwab]

On Jan 31 2022, Florian Weimer via Libc-alpha wrote:

> There's an existing comment
>
>   /* The decimal point must not be empty.  This is not said explicitly
>      in POSIX but ANSI C (ISO/IEC 9899) says in 4.4.2.1 it has to be
>      != "".  */
>
> that says that empty strings/null characters are invalid.

This is only about decimal_point, mon_decimal_point can be empty.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Florian Weimer]

* Andreas Schwab:

> On Jan 31 2022, Florian Weimer via Libc-alpha wrote:
>
>> There's an existing comment
>>
>>   /* The decimal point must not be empty.  This is not said explicitly
>>      in POSIX but ANSI C (ISO/IEC 9899) says in 4.4.2.1 it has to be
>>      != "".  */
>>
>> that says that empty strings/null characters are invalid.
>
> This is only about decimal_point, mon_decimal_point can be empty.

Hmm, I'll take your word for it.

So the comment should definitely go, and the Carlos' change is the right
way to do it?

Thanks,
Florian

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Andreas Schwab]

On Jan 31 2022, Florian Weimer wrote:

> * Andreas Schwab:
>
>> On Jan 31 2022, Florian Weimer via Libc-alpha wrote:
>>
>>> There's an existing comment
>>>
>>>   /* The decimal point must not be empty.  This is not said explicitly
>>>      in POSIX but ANSI C (ISO/IEC 9899) says in 4.4.2.1 it has to be
>>>      != "".  */
>>>
>>> that says that empty strings/null characters are invalid.
>>
>> This is only about decimal_point, mon_decimal_point can be empty.
>
> Hmm, I'll take your word for it.

See 7.11.2.1, paragraph 3 and 10.

-- 
Andreas Schwab, schwab@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
"And now for something completely different."

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Florian Weimer]

* Andreas Schwab:

> On Jan 31 2022, Florian Weimer wrote:
>
>> * Andreas Schwab:
>>
>>> On Jan 31 2022, Florian Weimer via Libc-alpha wrote:
>>>
>>>> There's an existing comment
>>>>
>>>>   /* The decimal point must not be empty.  This is not said explicitly
>>>>      in POSIX but ANSI C (ISO/IEC 9899) says in 4.4.2.1 it has to be
>>>>      != "".  */
>>>>
>>>> that says that empty strings/null characters are invalid.
>>>
>>> This is only about decimal_point, mon_decimal_point can be empty.
>>
>> Hmm, I'll take your word for it.
>
> See 7.11.2.1, paragraph 3 and 10.

That is fairly conclusive indeed (numbers match C11).

Are you okay with Carlos' patch with a comment update?

Thanks,
Florian

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Florian Weimer]

* Carlos O'Donell:

> The handling of mon_decimal_point is incorrect when it comes to
> handling the empty "" value.  The existing parser in monetary_read()
> will correctly handle setting the non-wide-character value and the
> wide-character value e.g. STR_ELEM_WC(mon_decimal_point) if they are
> set in the locale definition.  However, in monetary_finish() we have
> conflicting TEST_ELEM() which sets a default value (if the locale
> definition doesn't include one), and subsequent code which looks for
> mon_decimal_point to be NULL to issue a specific error message and set
> the defaults. The latter is unused because TEST_ELEM() always sets a
> default.  The simplest solution is to remove the TEST_ELEM() check,
> and allow the existing check to look to see if mon_decimal_point is
> NULL and set an appropriate default.  The final fix is to move the
> setting of mon_decimal_point_wc so it occurs only when
> mon_decimal_point is being set to a default, keeping both values
> consistent. There is no way to tell the difference between
> mon_decimal_point_wc having been set to the empty string and not
> having been defined at all, for that distinction we must use
> mon_decimal_point being NULL or "", and so we must logically set
> the default together with mon_decimal_point.
>
> Lastly, there are more fixes similar to this that could be made to
> ld-monetary.c, but we avoid that in order to fix just the code
> required for mon_decimal_point, which impacts the ability for C.UTF-8
> to set mon_decimal_point to "", since without this fix we end up with
> an inconsistent setting of mon_decimal_point set to "", but
> mon_decimal_point_wc set to "." which is incorrect.
>
> Tested on x86_64 and i686 without regression.
> ---
>  locale/programs/ld-monetary.c | 4 +---
>  1 file changed, 1 insertion(+), 3 deletions(-)
>
> diff --git a/locale/programs/ld-monetary.c b/locale/programs/ld-monetary.c
> index 277b9ff042..3b0412b405 100644
> --- a/locale/programs/ld-monetary.c
> +++ b/locale/programs/ld-monetary.c
> @@ -207,7 +207,6 @@ No definition for %s category found"), "LC_MONETARY");
>  
>    TEST_ELEM (int_curr_symbol, "");
>    TEST_ELEM (currency_symbol, "");
> -  TEST_ELEM (mon_decimal_point, ".");
>    TEST_ELEM (mon_thousands_sep, "");
>    TEST_ELEM (positive_sign, "");
>    TEST_ELEM (negative_sign, "");
> @@ -257,6 +256,7 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>  	record_error (0, 0, _("%s: field `%s' not defined"),
>  		      "LC_MONETARY", "mon_decimal_point");
>        monetary->mon_decimal_point = ".";
> +      monetary->mon_decimal_point_wc = L'.';
>      }
>    else if (monetary->mon_decimal_point[0] == '\0' && ! be_quiet && ! nothing)
>      {
> @@ -264,8 +264,6 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>  %s: value for field `%s' must not be an empty string"),
>  		    "LC_MONETARY", "mon_decimal_point");
>      }
> -  if (monetary->mon_decimal_point_wc == L'\0')
> -    monetary->mon_decimal_point_wc = L'.';
>  
>    if (monetary->mon_grouping_len == 0)
>      {

I have verified that this does not change the localedef output for the
existing locales created by install-locale-files.

I think we need further cleanups in the comments and checks (which were
coped from LC_NUMERIC, but should not apply to LC_MONETARY).  But I
think we can release with this version.

Reviewed-by: Florian Weimer <fweimer@redhat.com>

Thanks,
Florian

Re: [PATCH 2/2] localedata: Adjust C.UTF-8 to align with C/POSIX.

[Florian Weimer]

* Carlos O'Donell:

> We have had one downstream report from Canonical [1] that
> an rrdtool test was broken by the differences in LC_TIME
> that we had in the non-builtin C locale (C.UTF-8). If one
> application has an issue there are going to be others, and
> so with this commit we review and fix all the issues that
> cause the builtin C locale to be different from C.UTF-8,
> which includes:
> * mon_decimal_point should be empty e.g. ""
>  - Depends on mon_decimal_point_wc fix.
> * negative_sign should be empty e.g. ""
> * week should be aligned with ISO 30112 default e.g. 7;19971130;4
> * d_fmt corrected with escaped slashes e.g. "%m//%d//%y"
> * yesstr and nostr should be empty e.g. ""
> * country_ab2 and country_ab3 should be empty e.g. ""
>
> We bump LC_IDENTIFICATION version and adjust the date to
> indicate the change in the locale.
>
> A new tst-c-utf8-consistency test is added to ensure
> consistency between C/POSIX and C.UTF-8.
>
> Tested on x86_64 and i686 without regression.
>
> [1] https://sourceware.org/pipermail/libc-alpha/2022-January/135703.html
>
> Co-authored-by: Florian Weimer <fweimer@redhat.com>
> ---
>  localedata/Makefile                 |  30 +-
>  localedata/locales/C                |  22 +-
>  localedata/tst-c-utf8-consistency.c | 539 ++++++++++++++++++++++++++++
>  3 files changed, 578 insertions(+), 13 deletions(-)
>  create mode 100644 localedata/tst-c-utf8-consistency.c

This looks broadly okay to me.  Dropping the ISO standard reference
seems prudent if we can't check its contents.

Reviewed-by: Florian Weimer <fweimer@redhat.com>

Thanks,
Florian

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Carlos O'Donell]

On 2/1/22 06:47, Florian Weimer wrote:
> * Carlos O'Donell:
> 
>> The handling of mon_decimal_point is incorrect when it comes to
>> handling the empty "" value.  The existing parser in monetary_read()
>> will correctly handle setting the non-wide-character value and the
>> wide-character value e.g. STR_ELEM_WC(mon_decimal_point) if they are
>> set in the locale definition.  However, in monetary_finish() we have
>> conflicting TEST_ELEM() which sets a default value (if the locale
>> definition doesn't include one), and subsequent code which looks for
>> mon_decimal_point to be NULL to issue a specific error message and set
>> the defaults. The latter is unused because TEST_ELEM() always sets a
>> default.  The simplest solution is to remove the TEST_ELEM() check,
>> and allow the existing check to look to see if mon_decimal_point is
>> NULL and set an appropriate default.  The final fix is to move the
>> setting of mon_decimal_point_wc so it occurs only when
>> mon_decimal_point is being set to a default, keeping both values
>> consistent. There is no way to tell the difference between
>> mon_decimal_point_wc having been set to the empty string and not
>> having been defined at all, for that distinction we must use
>> mon_decimal_point being NULL or "", and so we must logically set
>> the default together with mon_decimal_point.
>>
>> Lastly, there are more fixes similar to this that could be made to
>> ld-monetary.c, but we avoid that in order to fix just the code
>> required for mon_decimal_point, which impacts the ability for C.UTF-8
>> to set mon_decimal_point to "", since without this fix we end up with
>> an inconsistent setting of mon_decimal_point set to "", but
>> mon_decimal_point_wc set to "." which is incorrect.
>>
>> Tested on x86_64 and i686 without regression.
>> ---
>>  locale/programs/ld-monetary.c | 4 +---
>>  1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/locale/programs/ld-monetary.c b/locale/programs/ld-monetary.c
>> index 277b9ff042..3b0412b405 100644
>> --- a/locale/programs/ld-monetary.c
>> +++ b/locale/programs/ld-monetary.c
>> @@ -207,7 +207,6 @@ No definition for %s category found"), "LC_MONETARY");
>>  
>>    TEST_ELEM (int_curr_symbol, "");
>>    TEST_ELEM (currency_symbol, "");
>> -  TEST_ELEM (mon_decimal_point, ".");
>>    TEST_ELEM (mon_thousands_sep, "");
>>    TEST_ELEM (positive_sign, "");
>>    TEST_ELEM (negative_sign, "");
>> @@ -257,6 +256,7 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>>  	record_error (0, 0, _("%s: field `%s' not defined"),
>>  		      "LC_MONETARY", "mon_decimal_point");
>>        monetary->mon_decimal_point = ".";
>> +      monetary->mon_decimal_point_wc = L'.';
>>      }
>>    else if (monetary->mon_decimal_point[0] == '\0' && ! be_quiet && ! nothing)
>>      {
>> @@ -264,8 +264,6 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>>  %s: value for field `%s' must not be an empty string"),
>>  		    "LC_MONETARY", "mon_decimal_point");
>>      }
>> -  if (monetary->mon_decimal_point_wc == L'\0')
>> -    monetary->mon_decimal_point_wc = L'.';
>>  
>>    if (monetary->mon_grouping_len == 0)
>>      {
> 
> I have verified that this does not change the localedef output for the
> existing locales created by install-locale-files.
> 
> I think we need further cleanups in the comments and checks (which were
> coped from LC_NUMERIC, but should not apply to LC_MONETARY).  But I
> think we can release with this version.

I filed this bug to track that:
Bug 28845 - ld-monetary.c should be updated to match ISO C and other standards.
https://sourceware.org/bugzilla/show_bug.cgi?id=28845

Thanks for the review!


> Reviewed-by: Florian Weimer <fweimer@redhat.com>
> 
> Thanks,
> Florian
> 


-- 
Cheers,
Carlos.

Re: [PATCH 2/2] localedata: Adjust C.UTF-8 to align with C/POSIX.

[Carlos O'Donell]

On 2/1/22 07:05, Florian Weimer wrote:
> * Carlos O'Donell:
> 
>> We have had one downstream report from Canonical [1] that
>> an rrdtool test was broken by the differences in LC_TIME
>> that we had in the non-builtin C locale (C.UTF-8). If one
>> application has an issue there are going to be others, and
>> so with this commit we review and fix all the issues that
>> cause the builtin C locale to be different from C.UTF-8,
>> which includes:
>> * mon_decimal_point should be empty e.g. ""
>>  - Depends on mon_decimal_point_wc fix.
>> * negative_sign should be empty e.g. ""
>> * week should be aligned with ISO 30112 default e.g. 7;19971130;4
>> * d_fmt corrected with escaped slashes e.g. "%m//%d//%y"
>> * yesstr and nostr should be empty e.g. ""
>> * country_ab2 and country_ab3 should be empty e.g. ""
>>
>> We bump LC_IDENTIFICATION version and adjust the date to
>> indicate the change in the locale.
>>
>> A new tst-c-utf8-consistency test is added to ensure
>> consistency between C/POSIX and C.UTF-8.
>>
>> Tested on x86_64 and i686 without regression.
>>
>> [1] https://sourceware.org/pipermail/libc-alpha/2022-January/135703.html
>>
>> Co-authored-by: Florian Weimer <fweimer@redhat.com>
>> ---
>>  localedata/Makefile                 |  30 +-
>>  localedata/locales/C                |  22 +-
>>  localedata/tst-c-utf8-consistency.c | 539 ++++++++++++++++++++++++++++
>>  3 files changed, 578 insertions(+), 13 deletions(-)
>>  create mode 100644 localedata/tst-c-utf8-consistency.c
> 
> This looks broadly okay to me.  Dropping the ISO standard reference
> seems prudent if we can't check its contents.
> 
> Reviewed-by: Florian Weimer <fweimer@redhat.com>

Thanks. I'll drop the ISO references from the commit message.

-- 
Cheers,
Carlos.

Re: [PATCH 1/2] localedef: Fix handling of empty mon_decimal_point

[Carlos O'Donell]

On 2/1/22 11:00, Carlos O'Donell wrote:
> On 2/1/22 06:47, Florian Weimer wrote:
>> * Carlos O'Donell:
>>
>>> The handling of mon_decimal_point is incorrect when it comes to
>>> handling the empty "" value.  The existing parser in monetary_read()
>>> will correctly handle setting the non-wide-character value and the
>>> wide-character value e.g. STR_ELEM_WC(mon_decimal_point) if they are
>>> set in the locale definition.  However, in monetary_finish() we have
>>> conflicting TEST_ELEM() which sets a default value (if the locale
>>> definition doesn't include one), and subsequent code which looks for
>>> mon_decimal_point to be NULL to issue a specific error message and set
>>> the defaults. The latter is unused because TEST_ELEM() always sets a
>>> default.  The simplest solution is to remove the TEST_ELEM() check,
>>> and allow the existing check to look to see if mon_decimal_point is
>>> NULL and set an appropriate default.  The final fix is to move the
>>> setting of mon_decimal_point_wc so it occurs only when
>>> mon_decimal_point is being set to a default, keeping both values
>>> consistent. There is no way to tell the difference between
>>> mon_decimal_point_wc having been set to the empty string and not
>>> having been defined at all, for that distinction we must use
>>> mon_decimal_point being NULL or "", and so we must logically set
>>> the default together with mon_decimal_point.
>>>
>>> Lastly, there are more fixes similar to this that could be made to
>>> ld-monetary.c, but we avoid that in order to fix just the code
>>> required for mon_decimal_point, which impacts the ability for C.UTF-8
>>> to set mon_decimal_point to "", since without this fix we end up with
>>> an inconsistent setting of mon_decimal_point set to "", but
>>> mon_decimal_point_wc set to "." which is incorrect.
>>>
>>> Tested on x86_64 and i686 without regression.
>>> ---
>>>  locale/programs/ld-monetary.c | 4 +---
>>>  1 file changed, 1 insertion(+), 3 deletions(-)
>>>
>>> diff --git a/locale/programs/ld-monetary.c b/locale/programs/ld-monetary.c
>>> index 277b9ff042..3b0412b405 100644
>>> --- a/locale/programs/ld-monetary.c
>>> +++ b/locale/programs/ld-monetary.c
>>> @@ -207,7 +207,6 @@ No definition for %s category found"), "LC_MONETARY");
>>>  
>>>    TEST_ELEM (int_curr_symbol, "");
>>>    TEST_ELEM (currency_symbol, "");
>>> -  TEST_ELEM (mon_decimal_point, ".");
>>>    TEST_ELEM (mon_thousands_sep, "");
>>>    TEST_ELEM (positive_sign, "");
>>>    TEST_ELEM (negative_sign, "");
>>> @@ -257,6 +256,7 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>>>  	record_error (0, 0, _("%s: field `%s' not defined"),
>>>  		      "LC_MONETARY", "mon_decimal_point");
>>>        monetary->mon_decimal_point = ".";
>>> +      monetary->mon_decimal_point_wc = L'.';
>>>      }
>>>    else if (monetary->mon_decimal_point[0] == '\0' && ! be_quiet && ! nothing)
>>>      {
>>> @@ -264,8 +264,6 @@ not correspond to a valid name in ISO 4217 [--no-warnings=intcurrsym]"),
>>>  %s: value for field `%s' must not be an empty string"),
>>>  		    "LC_MONETARY", "mon_decimal_point");
>>>      }
>>> -  if (monetary->mon_decimal_point_wc == L'\0')
>>> -    monetary->mon_decimal_point_wc = L'.';
>>>  
>>>    if (monetary->mon_grouping_len == 0)
>>>      {
>>
>> I have verified that this does not change the localedef output for the
>> existing locales created by install-locale-files.
>>
>> I think we need further cleanups in the comments and checks (which were
>> coped from LC_NUMERIC, but should not apply to LC_MONETARY).  But I
>> think we can release with this version.
> 
> I filed this bug to track that:
> Bug 28845 - ld-monetary.c should be updated to match ISO C and other standards.
> https://sourceware.org/bugzilla/show_bug.cgi?id=28845

And I filed one more bug to track the original bug, which I'll close after push:
https://sourceware.org/bugzilla/show_bug.cgi?id=28847


-- 
Cheers,
Carlos.