contrib/grep/lib/nl_langinfo.c

   1 /* nl_langinfo() replacement: query locale dependent information.
   2
   3    Copyright (C) 2007-2015 Free Software Foundation, Inc.
   4
   5    This program is free software: you can redistribute it and/or modify
   6    it under the terms of the GNU General Public License as published by
   7    the Free Software Foundation; either version 3 of the License, or
   8    (at your option) any later version.
   9
  10    This program is distributed in the hope that it will be useful,
  11    but WITHOUT ANY WARRANTY; without even the implied warranty of
  12    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13    GNU General Public License for more details.
  14
  15    You should have received a copy of the GNU General Public License
  16    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  17
  18 #include <config.h>
  19
  20 /* Specification.  */
  21 #include <langinfo.h>
  22
  23 #include <locale.h>
  24 #include <string.h>
  25 #if (defined _WIN32 || defined __WIN32__) && ! defined __CYGWIN__
  26 # define WIN32_LEAN_AND_MEAN  /* avoid including junk */
  27 # include <windows.h>
  28 # include <stdio.h>
  29 #endif
  30
  31 /* Return the codeset of the current locale, if this is easily deducible.
  32    Otherwise, return "".  */
  33 static char *
  34 ctype_codeset (void)
  35 {
  36   static char buf[2 + 10 + 1];
  37   size_t buflen = 0;
  38   char const *locale = setlocale (LC_CTYPE, NULL);
  39   char *codeset = buf;
  40   size_t codesetlen;
  41   codeset[0] = '\0';
  42
  43   if (locale && locale[0])
  44     {
  45       /* If the locale name contains an encoding after the dot, return it.  */
  46       char *dot = strchr (locale, '.');
  47
  48       if (dot)
  49         {
  50           /* Look for the possible @... trailer and remove it, if any.  */
  51           char *codeset_start = dot + 1;
  52           char const *modifier = strchr (codeset_start, '@');
  53
  54           if (! modifier)
  55             codeset = codeset_start;
  56           else
  57             {
  58               codesetlen = modifier - codeset_start;
  59               if (codesetlen < sizeof buf)
  60                 {
  61                   codeset = memcpy (buf, codeset_start, codesetlen);
  62                   codeset[codesetlen] = '\0';
  63                 }
  64             }
  65         }
  66     }
  67
  68 #if (defined _WIN32 || defined __WIN32__) && ! defined __CYGWIN__
  69   /* If setlocale is successful, it returns the number of the
  70      codepage, as a string.  Otherwise, fall back on Windows API
  71      GetACP, which returns the locale's codepage as a number (although
  72      this doesn't change according to what the 'setlocale' call specified).
  73      Either way, prepend "CP" to make it a valid codeset name.  */
  74   codesetlen = strlen (codeset);
  75   if (0 < codesetlen && codesetlen < sizeof buf - 2)
  76     memmove (buf + 2, codeset, codesetlen + 1);
  77   else
  78     sprintf (buf + 2, "%u", GetACP ());
  79   codeset = memcpy (buf, "CP", 2);
  80 #endif
  81   return codeset;
  82 }
  83
  84
  85 #if REPLACE_NL_LANGINFO
  86
  87 /* Override nl_langinfo with support for added nl_item values.  */
  88
  89 # undef nl_langinfo
  90
  91 char *
  92 rpl_nl_langinfo (nl_item item)
  93 {
  94   switch (item)
  95     {
  96 # if GNULIB_defined_CODESET
  97     case CODESET:
  98       return ctype_codeset ();
  99 # endif
 100 # if GNULIB_defined_T_FMT_AMPM
 101     case T_FMT_AMPM:
 102       return "%I:%M:%S %p";
 103 # endif
 104 # if GNULIB_defined_ERA
 105     case ERA:
 106       /* The format is not standardized.  In glibc it is a sequence of strings
 107          of the form "direction:offset:start_date:end_date:era_name:era_format"
 108          with an empty string at the end.  */
 109       return "";
 110     case ERA_D_FMT:
 111       /* The %Ex conversion in strftime behaves like %x if the locale does not
 112          have an alternative time format.  */
 113       item = D_FMT;
 114       break;
 115     case ERA_D_T_FMT:
 116       /* The %Ec conversion in strftime behaves like %c if the locale does not
 117          have an alternative time format.  */
 118       item = D_T_FMT;
 119       break;
 120     case ERA_T_FMT:
 121       /* The %EX conversion in strftime behaves like %X if the locale does not
 122          have an alternative time format.  */
 123       item = T_FMT;
 124       break;
 125     case ALT_DIGITS:
 126       /* The format is not standardized.  In glibc it is a sequence of 10
 127          strings, appended in memory.  */
 128       return "\0\0\0\0\0\0\0\0\0\0";
 129 # endif
 130 # if GNULIB_defined_YESEXPR || !FUNC_NL_LANGINFO_YESEXPR_WORKS
 131     case YESEXPR:
 132       return "^[yY]";
 133     case NOEXPR:
 134       return "^[nN]";
 135 # endif
 136     default:
 137       break;
 138     }
 139   return nl_langinfo (item);
 140 }
 141
 142 #else
 143
 144 /* Provide nl_langinfo from scratch, either for native MS-Windows, or
 145    for old Unix platforms without locales, such as Linux libc5 or
 146    BeOS.  */
 147
 148 # include <time.h>
 149
 150 char *
 151 nl_langinfo (nl_item item)
 152 {
 153   static char nlbuf[100];
 154   struct tm tmm = { 0 };
 155
 156   switch (item)
 157     {
 158     /* nl_langinfo items of the LC_CTYPE category */
 159     case CODESET:
 160       {
 161         char *codeset = ctype_codeset ();
 162         if (*codeset)
 163           return codeset;
 164       }
 165 # ifdef __BEOS__
 166       return "UTF-8";
 167 # else
 168       return "ISO-8859-1";
 169 # endif
 170     /* nl_langinfo items of the LC_NUMERIC category */
 171     case RADIXCHAR:
 172       return localeconv () ->decimal_point;
 173     case THOUSEP:
 174       return localeconv () ->thousands_sep;
 175     case GROUPING:
 176       return localeconv () ->grouping;
 177     /* nl_langinfo items of the LC_TIME category.
 178        TODO: Really use the locale.  */
 179     case D_T_FMT:
 180     case ERA_D_T_FMT:
 181       return "%a %b %e %H:%M:%S %Y";
 182     case D_FMT:
 183     case ERA_D_FMT:
 184       return "%m/%d/%y";
 185     case T_FMT:
 186     case ERA_T_FMT:
 187       return "%H:%M:%S";
 188     case T_FMT_AMPM:
 189       return "%I:%M:%S %p";
 190     case AM_STR:
 191       if (!strftime (nlbuf, sizeof nlbuf, "%p", &tmm))
 192         return "AM";
 193       return nlbuf;
 194     case PM_STR:
 195       tmm.tm_hour = 12;
 196       if (!strftime (nlbuf, sizeof nlbuf, "%p", &tmm))
 197         return "PM";
 198       return nlbuf;
 199     case DAY_1:
 200     case DAY_2:
 201     case DAY_3:
 202     case DAY_4:
 203     case DAY_5:
 204     case DAY_6:
 205     case DAY_7:
 206       {
 207         static char const days[][sizeof "Wednesday"] = {
 208           "Sunday", "Monday", "Tuesday", "Wednesday", "Thursday",
 209           "Friday", "Saturday"
 210         };
 211         tmm.tm_wday = item - DAY_1;
 212         if (!strftime (nlbuf, sizeof nlbuf, "%A", &tmm))
 213           return (char *) days[item - DAY_1];
 214         return nlbuf;
 215       }
 216     case ABDAY_1:
 217     case ABDAY_2:
 218     case ABDAY_3:
 219     case ABDAY_4:
 220     case ABDAY_5:
 221     case ABDAY_6:
 222     case ABDAY_7:
 223       {
 224         static char const abdays[][sizeof "Sun"] = {
 225           "Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat"
 226         };
 227         tmm.tm_wday = item - ABDAY_1;
 228         if (!strftime (nlbuf, sizeof nlbuf, "%a", &tmm))
 229           return (char *) abdays[item - ABDAY_1];
 230         return nlbuf;
 231       }
 232     case MON_1:
 233     case MON_2:
 234     case MON_3:
 235     case MON_4:
 236     case MON_5:
 237     case MON_6:
 238     case MON_7:
 239     case MON_8:
 240     case MON_9:
 241     case MON_10:
 242     case MON_11:
 243     case MON_12:
 244       {
 245         static char const months[][sizeof "September"] = {
 246           "January", "February", "March", "April", "May", "June", "July",
 247           "September", "October", "November", "December"
 248         };
 249         tmm.tm_mon = item - MON_1;
 250         if (!strftime (nlbuf, sizeof nlbuf, "%B", &tmm))
 251           return (char *) months[item - MON_1];
 252         return nlbuf;
 253       }
 254     case ABMON_1:
 255     case ABMON_2:
 256     case ABMON_3:
 257     case ABMON_4:
 258     case ABMON_5:
 259     case ABMON_6:
 260     case ABMON_7:
 261     case ABMON_8:
 262     case ABMON_9:
 263     case ABMON_10:
 264     case ABMON_11:
 265     case ABMON_12:
 266       {
 267         static char const abmonths[][sizeof "Jan"] = {
 268           "Jan", "Feb", "Mar", "Apr", "May", "Jun", "Jul",
 269           "Sep", "Oct", "Nov", "Dec"
 270         };
 271         tmm.tm_mon = item - ABMON_1;
 272         if (!strftime (nlbuf, sizeof nlbuf, "%b", &tmm))
 273           return (char *) abmonths[item - ABMON_1];
 274         return nlbuf;
 275       }
 276     case ERA:
 277       return "";
 278     case ALT_DIGITS:
 279       return "\0\0\0\0\0\0\0\0\0\0";
 280     /* nl_langinfo items of the LC_MONETARY category.  */
 281     case CRNCYSTR:
 282       return localeconv () ->currency_symbol;
 283     case INT_CURR_SYMBOL:
 284       return localeconv () ->int_curr_symbol;
 285     case MON_DECIMAL_POINT:
 286       return localeconv () ->mon_decimal_point;
 287     case MON_THOUSANDS_SEP:
 288       return localeconv () ->mon_thousands_sep;
 289     case MON_GROUPING:
 290       return localeconv () ->mon_grouping;
 291     case POSITIVE_SIGN:
 292       return localeconv () ->positive_sign;
 293     case NEGATIVE_SIGN:
 294       return localeconv () ->negative_sign;
 295     case FRAC_DIGITS:
 296       return & localeconv () ->frac_digits;
 297     case INT_FRAC_DIGITS:
 298       return & localeconv () ->int_frac_digits;
 299     case P_CS_PRECEDES:
 300       return & localeconv () ->p_cs_precedes;
 301     case N_CS_PRECEDES:
 302       return & localeconv () ->n_cs_precedes;
 303     case P_SEP_BY_SPACE:
 304       return & localeconv () ->p_sep_by_space;
 305     case N_SEP_BY_SPACE:
 306       return & localeconv () ->n_sep_by_space;
 307     case P_SIGN_POSN:
 308       return & localeconv () ->p_sign_posn;
 309     case N_SIGN_POSN:
 310       return & localeconv () ->n_sign_posn;
 311     /* nl_langinfo items of the LC_MESSAGES category
 312        TODO: Really use the locale. */
 313     case YESEXPR:
 314       return "^[yY]";
 315     case NOEXPR:
 316       return "^[nN]";
 317     default:
 318       return "";
 319     }
 320 }
 321
 322 #endif