Thanks, I tried that out on Fedora 20 (pretending that its nl_langinfo
didn't exist) and had some trouble building and testing it, indicating
trouble for other POSIXish platforms. I fixed the problems I found. A
few things:
* nl_langinfo (FRAC_DIGITS) returns a pointer to a character, not a
string, and similarly for the other values representing tiny integers.
There's no need for null-termination.
* struct tm might not have the traditional Unix v7 layout.
* Some code can be shared between rpl_nl_langinfo and nl_langinfo by
hoisting it into a shared static function ctype_codeset.
* Often it's better to use char[][] constants instead of char *[]
constants when strings don't vary in size much, as this avoids
unnecessary relocations in shared libraries.
* I didn't follow why tmm had some members updated but not others. Does
Microsoft strftime care about otherwise-unused struct tm members? If
so, perhaps we should initialize all the members to a valid and
consistent value; if not, let's not bother with anything but the minimal.
Revised proposed patch attached.
From 4630fdeb658f96fce13ed2d4d9d89e9955aaa273 Mon Sep 17 00:00:00 2001
From: Eli Zaretskii <e...@gnu.org>
Date: Sat, 5 Jul 2014 14:42:47 -0700
Subject: [PATCH] nl_langinfo: CODESET on MS-Windows and more items from
localeconv
* lib/langinfo.in.h (DECIMAL_POINT, THOUSANDS_SEP, GROUPING)
(CURRENCY_SYMBOL, INT_CURR_SYMBOL, MON_DECIMAL_POINT)
(MON_THOUSANDS_SEP, MON_GROUPING, POSITIVE_SIGN, NEGATIVE_SIGN)
(FRAC_DIGITS, INT_FRAC_DIGITS, P_CS_PRECEDES, N_CS_PRECEDES)
(P_SEP_BY_SPACE, N_SEP_BY_SPACE, P_SIGN_POSN, N_SIGN_POSN): Define.
* lib/nl_langinfo.c: Include <locale.h> and <string.h> early.
Include <time.h> if !REPLACE_NL_LANGINFO.
(ctype_codeset): New function, taken from rpl_nl_langinfo.
(rpl_nl_langinfo): Use it.
(nl_langinfo) [!REPLACE_NL_LANGINFO]: Likewise.
Compute the values of RADIXCHAR, THOUSEP, GROUPING, CRNCYSTR,
INT_CURR_SYMBOL, MON_DECIMAL_POINT, MON_THOUSANDS_SEP,
MON_GROUPING, POSITIVE_SIGN, NEGATIVE_SIGN, FRAC_DIGITS,
INT_FRAC_DIGITS, P_CS_PRECEDES, N_CS_PRECEDES, P_SEP_BY_SPACE,
N_SEP_BY_SPACE, P_SIGN_POSN, and N_SIGN_POSN from the
corresponding values returned by 'localeconv'. Compute the values
of AM_STR, PM_STR, DAY_n, ABDAY_n, MON_n, and ABMON_n by calling
'strftime' with a suitable struct tm value.
---
ChangeLog | 23 ++++++
lib/langinfo.in.h | 18 +++++
lib/nl_langinfo.c | 215 +++++++++++++++++++++++++++++++++---------------------
3 files changed, 172 insertions(+), 84 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index eccf03e..0e800a1 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,26 @@
+2014-07-05 Eli Zaretskii <e...@gnu.org>
+ Paul Eggert <egg...@cs.ucla.edu>
+
+ nl_langinfo: CODESET on MS-Windows and more items from localeconv
+ * lib/langinfo.in.h (DECIMAL_POINT, THOUSANDS_SEP, GROUPING)
+ (CURRENCY_SYMBOL, INT_CURR_SYMBOL, MON_DECIMAL_POINT)
+ (MON_THOUSANDS_SEP, MON_GROUPING, POSITIVE_SIGN, NEGATIVE_SIGN)
+ (FRAC_DIGITS, INT_FRAC_DIGITS, P_CS_PRECEDES, N_CS_PRECEDES)
+ (P_SEP_BY_SPACE, N_SEP_BY_SPACE, P_SIGN_POSN, N_SIGN_POSN): Define.
+ * lib/nl_langinfo.c: Include <locale.h> and <string.h> early.
+ Include <time.h> if !REPLACE_NL_LANGINFO.
+ (ctype_codeset): New function, taken from rpl_nl_langinfo.
+ (rpl_nl_langinfo): Use it.
+ (nl_langinfo) [!REPLACE_NL_LANGINFO]: Likewise.
+ Compute the values of RADIXCHAR, THOUSEP, GROUPING, CRNCYSTR,
+ INT_CURR_SYMBOL, MON_DECIMAL_POINT, MON_THOUSANDS_SEP,
+ MON_GROUPING, POSITIVE_SIGN, NEGATIVE_SIGN, FRAC_DIGITS,
+ INT_FRAC_DIGITS, P_CS_PRECEDES, N_CS_PRECEDES, P_SEP_BY_SPACE,
+ N_SEP_BY_SPACE, P_SIGN_POSN, and N_SIGN_POSN from the
+ corresponding values returned by 'localeconv'. Compute the values
+ of AM_STR, PM_STR, DAY_n, ABDAY_n, MON_n, and ABMON_n by calling
+ 'strftime' with a suitable struct tm value.
+
2014-07-05 Paul Eggert <egg...@cs.ucla.edu>
Bruno Haible has stepped down as maintainer.
diff --git a/lib/langinfo.in.h b/lib/langinfo.in.h
index 17f4b9b..d118d5d 100644
--- a/lib/langinfo.in.h
+++ b/lib/langinfo.in.h
@@ -49,7 +49,10 @@ typedef int nl_item;
# define CODESET 10000
/* nl_langinfo items of the LC_NUMERIC category */
# define RADIXCHAR 10001
+# define DECIMAL_POINT RADIXCHAR
# define THOUSEP 10002
+# define THOUSANDS_SEP THOUSEP
+# define GROUPING 10114
/* nl_langinfo items of the LC_TIME category */
# define D_T_FMT 10003
# define D_FMT 10004
@@ -102,6 +105,21 @@ typedef int nl_item;
# define ALT_DIGITS 10051
/* nl_langinfo items of the LC_MONETARY category */
# define CRNCYSTR 10052
+# define CURRENCY_SYMBOL CRNCYSTR
+# define INT_CURR_SYMBOL 10100
+# define MON_DECIMAL_POINT 10101
+# define MON_THOUSANDS_SEP 10102
+# define MON_GROUPING 10103
+# define POSITIVE_SIGN 10104
+# define NEGATIVE_SIGN 10105
+# define FRAC_DIGITS 10106
+# define INT_FRAC_DIGITS 10107
+# define P_CS_PRECEDES 10108
+# define N_CS_PRECEDES 10109
+# define P_SEP_BY_SPACE 10110
+# define N_SEP_BY_SPACE 10111
+# define P_SIGN_POSN 10112
+# define N_SIGN_POSN 10113
/* nl_langinfo items of the LC_MESSAGES category */
# define YESEXPR 10053
# define NOEXPR 10054
diff --git a/lib/nl_langinfo.c b/lib/nl_langinfo.c
index 287abfd..e461186 100644
--- a/lib/nl_langinfo.c
+++ b/lib/nl_langinfo.c
@@ -20,13 +20,47 @@
/* Specification. */
#include <langinfo.h>
+#include <locale.h>
+#include <string.h>
+
+/* Return the codeset of the current locale, if this is easily deducible.
+ Otherwise, return "". */
+static char *
+ctype_codeset (void)
+{
+ static char buf[2 + 10 + 1];
+ size_t buflen = 0;
+ char const *locale = setlocale (LC_CTYPE, NULL);
+
+ if (locale && locale[0])
+ {
+ /* If the locale name contains an encoding after the dot, return it. */
+ char *dot = strchr (locale, '.');
+
+ if (dot)
+ {
+ /* Look for the possible @... trailer and remove it, if any. */
+ char const *modifier = strchr (++dot, '@');
+
+ if (! modifier)
+ return dot;
+ if (modifier - dot < sizeof buf)
+ {
+ buflen = modifier - dot;
+ memcpy (buf, dot, buflen);
+ }
+ }
+ }
+
+ buf[buflen] = '\0';
+ return buf;
+}
+
+
#if REPLACE_NL_LANGINFO
/* Override nl_langinfo with support for added nl_item values. */
-# include <locale.h>
-# include <string.h>
-
# undef nl_langinfo
char *
@@ -36,36 +70,7 @@ rpl_nl_langinfo (nl_item item)
{
# if GNULIB_defined_CODESET
case CODESET:
- {
- const char *locale;
- static char buf[2 + 10 + 1];
-
- locale = setlocale (LC_CTYPE, NULL);
- if (locale != NULL && locale[0] != '\0')
- {
- /* If the locale name contains an encoding after the dot, return
- it. */
- const char *dot = strchr (locale, '.');
-
- if (dot != NULL)
- {
- const char *modifier;
-
- dot++;
- /* Look for the possible @... trailer and remove it, if any.
*/
- modifier = strchr (dot, '@');
- if (modifier == NULL)
- return dot;
- if (modifier - dot < sizeof (buf))
- {
- memcpy (buf, dot, modifier - dot);
- buf [modifier - dot] = '\0';
- return buf;
- }
- }
- }
- return "";
- }
+ return ctype_codeset ();
# endif
# if GNULIB_defined_T_FMT_AMPM
case T_FMT_AMPM:
@@ -128,24 +133,31 @@ rpl_nl_langinfo (nl_item item)
# endif
-# include <locale.h>
+# include <time.h>
char *
nl_langinfo (nl_item item)
{
+ static char nlbuf[100];
+ struct tm tmm = { 0 };
+
switch (item)
{
/* nl_langinfo items of the LC_CTYPE category */
case CODESET:
-# if (defined _WIN32 || defined __WIN32__) && ! defined __CYGWIN__
{
- static char buf[2 + 10 + 1];
-
- /* The Windows API has a function returning the locale's codepage as
- a number. */
- sprintf (buf, "CP%u", GetACP ());
- return buf;
+ char *codeset = ctype_codeset ();
+ if (*codeset)
+ return codeset;
}
+# if (defined _WIN32 || defined __WIN32__) && ! defined __CYGWIN__
+ /* The Windows API has a function returning the locale's
+ codepage as a number, but the value doesn't change according
+ to what the 'setlocale' call specified. So use it as a
+ last resort, in case the string returned by 'setlocale'
+ doesn't specify the codepage. */
+ sprintf (nlbuf, "CP%u", GetACP ());
+ return nlbuf;
# elif defined __BEOS__
return "UTF-8";
# else
@@ -156,6 +168,8 @@ nl_langinfo (nl_item item)
return localeconv () ->decimal_point;
case THOUSEP:
return localeconv () ->thousands_sep;
+ case GROUPING:
+ return localeconv () ->grouping;
/* nl_langinfo items of the LC_TIME category.
TODO: Really use the locale. */
case D_T_FMT:
@@ -170,93 +184,126 @@ nl_langinfo (nl_item item)
case T_FMT_AMPM:
return "%I:%M:%S %p";
case AM_STR:
- return "AM";
+ if (!strftime (nlbuf, sizeof nlbuf, "%p", &tmm))
+ return "AM";
+ return nlbuf;
case PM_STR:
- return "PM";
+ tmm.tm_hour = 12;
+ if (!strftime (nlbuf, sizeof nlbuf, "%p", &tmm))
+ return "PM";
+ return nlbuf;
case DAY_1:
- return "Sunday";
case DAY_2:
- return "Monday";
case DAY_3:
- return "Tuesday";
case DAY_4:
- return "Wednesday";
case DAY_5:
- return "Thursday";
case DAY_6:
- return "Friday";
case DAY_7:
- return "Saturday";
+ {
+ static char const days[][sizeof "Wednesday"] = {
+ "Sunday", "Monday", "Tuesday", "Wednesday", "Thursday",
+ "Friday", "Saturday"
+ };
+ tmm.tm_wday = item - DAY_1;
+ if (!strftime (nlbuf, sizeof nlbuf, "%A", &tmm))
+ return (char *) days[item - DAY_1];
+ return nlbuf;
+ }
case ABDAY_1:
- return "Sun";
case ABDAY_2:
- return "Mon";
case ABDAY_3:
- return "Tue";
case ABDAY_4:
- return "Wed";
case ABDAY_5:
- return "Thu";
case ABDAY_6:
- return "Fri";
case ABDAY_7:
- return "Sat";
+ {
+ static char const abdays[][sizeof "Sun"] = {
+ "Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat"
+ };
+ tmm.tm_wday = item - ABDAY_1;
+ if (!strftime (nlbuf, sizeof nlbuf, "%a", &tmm))
+ return (char *) abdays[item - ABDAY_1];
+ return nlbuf;
+ }
case MON_1:
- return "January";
case MON_2:
- return "February";
case MON_3:
- return "March";
case MON_4:
- return "April";
case MON_5:
- return "May";
case MON_6:
- return "June";
case MON_7:
- return "July";
case MON_8:
- return "August";
case MON_9:
- return "September";
case MON_10:
- return "October";
case MON_11:
- return "November";
case MON_12:
- return "December";
+ {
+ static char const months[][sizeof "September"] = {
+ "January", "February", "March", "April", "May", "June", "July",
+ "September", "October", "November", "December"
+ };
+ tmm.tm_mon = item - MON_1;
+ if (!strftime (nlbuf, sizeof nlbuf, "%B", &tmm))
+ return (char *) months[item - MON_1];
+ return nlbuf;
+ }
case ABMON_1:
- return "Jan";
case ABMON_2:
- return "Feb";
case ABMON_3:
- return "Mar";
case ABMON_4:
- return "Apr";
case ABMON_5:
- return "May";
case ABMON_6:
- return "Jun";
case ABMON_7:
- return "Jul";
case ABMON_8:
- return "Aug";
case ABMON_9:
- return "Sep";
case ABMON_10:
- return "Oct";
case ABMON_11:
- return "Nov";
case ABMON_12:
- return "Dec";
+ {
+ static char const abmonths[][sizeof "Jan"] = {
+ "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul",
+ "Sep", "Oct", "Nov", "Dec"
+ };
+ tmm.tm_mon = item - ABMON_1;
+ if (!strftime (nlbuf, sizeof nlbuf, "%b", &tmm))
+ return (char *) abmonths[item - ABMON_1];
+ return nlbuf;
+ }
case ERA:
return "";
case ALT_DIGITS:
return "\0\0\0\0\0\0\0\0\0\0";
- /* nl_langinfo items of the LC_MONETARY category
- TODO: Really use the locale. */
+ /* nl_langinfo items of the LC_MONETARY category. */
case CRNCYSTR:
- return "-";
+ return localeconv () ->currency_symbol;
+ case INT_CURR_SYMBOL:
+ return localeconv () ->int_curr_symbol;
+ case MON_DECIMAL_POINT:
+ return localeconv () ->mon_decimal_point;
+ case MON_THOUSANDS_SEP:
+ return localeconv () ->mon_thousands_sep;
+ case MON_GROUPING:
+ return localeconv () ->mon_grouping;
+ case POSITIVE_SIGN:
+ return localeconv () ->positive_sign;
+ case NEGATIVE_SIGN:
+ return localeconv () ->negative_sign;
+ case FRAC_DIGITS:
+ return & localeconv () ->frac_digits;
+ case INT_FRAC_DIGITS:
+ return & localeconv () ->int_frac_digits;
+ case P_CS_PRECEDES:
+ return & localeconv () ->p_cs_precedes;
+ case N_CS_PRECEDES:
+ return & localeconv () ->n_cs_precedes;
+ case P_SEP_BY_SPACE:
+ return & localeconv () ->p_sep_by_space;
+ case N_SEP_BY_SPACE:
+ return & localeconv () ->n_sep_by_space;
+ case P_SIGN_POSN:
+ return & localeconv () ->p_sign_posn;
+ case N_SIGN_POSN:
+ return & localeconv () ->n_sign_posn;
/* nl_langinfo items of the LC_MESSAGES category
TODO: Really use the locale. */
case YESEXPR:
--
1.9.3
