Another Win32 patch to gettext/intl: Get locale value from system

[Tor Lillqvist]

Hi again,

The following addition to dcigettext.c makes libintl look up the
system locale, translate it to the Unix style xx_XX, and use it if
none of the environment variables are set, instead of using just "C".

Cheers,
--tml

Add this function:

#ifdef _WIN32
/* Lifted from GLib: */

/**
 * g_win32_getlocale:
 *
 * The setlocale in the Microsoft C library uses locale names of the
 * form "English_United States.1252" etc. We want the Unixish standard
 * form "en", "zh_TW" etc. This function gets the current thread
 * locale from Windows and returns it as a string of the above form
 * for use in forming file names etc. The return value is a static buffer.
 *
 * Returns: locale name
 */

static char *
g_win32_getlocale (void)
{
  LCID lcid;
  char *ev;
  int primary, sub;
  char *l = NULL, *sl = NULL;
  static char bfr[20];

  lcid = GetThreadLocale ();
  primary = PRIMARYLANGID (LANGIDFROMLCID (lcid));
  sub = SUBLANGID (LANGIDFROMLCID (lcid));
  switch (primary)
    {
    case LANG_AFRIKAANS: l = "af"; break;
    case LANG_ALBANIAN: l = "sq"; break;
    case LANG_ARABIC:
      l = "ar";
      switch (sub)
        {
        case SUBLANG_ARABIC_SAUDI_ARABIA: sl = "SA"; break;
        case SUBLANG_ARABIC_IRAQ: sl = "IQ"; break;
        case SUBLANG_ARABIC_EGYPT: sl = "EG"; break;
        case SUBLANG_ARABIC_LIBYA: sl = "LY"; break;
        case SUBLANG_ARABIC_ALGERIA: sl = "DZ"; break;
        case SUBLANG_ARABIC_MOROCCO: sl = "MA"; break;
        case SUBLANG_ARABIC_TUNISIA: sl = "TN"; break;
        case SUBLANG_ARABIC_OMAN: sl = "OM"; break;
        case SUBLANG_ARABIC_YEMEN: sl = "YE"; break;
        case SUBLANG_ARABIC_SYRIA: sl = "SY"; break;
        case SUBLANG_ARABIC_JORDAN: sl = "JO"; break;
        case SUBLANG_ARABIC_LEBANON: sl = "LB"; break;
        case SUBLANG_ARABIC_KUWAIT: sl = "KW"; break;
        case SUBLANG_ARABIC_UAE: sl = "AE"; break;
        case SUBLANG_ARABIC_BAHRAIN: sl = "BH"; break;
        case SUBLANG_ARABIC_QATAR: sl = "QA"; break;
        }
      break;
    case LANG_ARMENIAN: l = "hy"; break;
    case LANG_ASSAMESE: l = "as"; break;
    case LANG_AZERI: l = "az"; break;
    case LANG_BASQUE: l = "eu"; break;
    case LANG_BELARUSIAN: l = "be"; break;
    case LANG_BENGALI: l = "bn"; break;
    case LANG_BULGARIAN: l = "bg"; break;
    case LANG_CATALAN: l = "ca"; break;
    case LANG_CHINESE:
      l = "zh";
      switch (sub)
        {
        case SUBLANG_CHINESE_TRADITIONAL: sl = "TW"; break;
        case SUBLANG_CHINESE_SIMPLIFIED: sl = "CH"; break;
        case SUBLANG_CHINESE_HONGKONG: sl = "HK"; break;
        case SUBLANG_CHINESE_SINGAPORE: sl = "SG"; break;
        case SUBLANG_CHINESE_MACAU: sl = "MO"; break;
        }
      break;
    case LANG_CROATIAN:         /* LANG_CROATIAN == LANG_SERBIAN
                                 * What used to be called Serbo-Croatian
                                 * should really now be two separate
                                 * languages because of political reasons.
                                 * (Says tml, who knows nothing about Serbian
                                 * or Croatian.)
                                 * (I can feel those flames coming already.)
                                 */
      switch (sub)
        {
        case SUBLANG_SERBIAN_LATIN: l = "sp"; break;
        case SUBLANG_SERBIAN_CYRILLIC: l = "sr"; break;
        default: l = "hr";      /* ??? */
        }
      break;
    case LANG_CZECH: l = "cs"; break;
    case LANG_DANISH: l = "da"; break;
    case LANG_DUTCH:
      l = "nl";
      switch (sub)
        {
        case SUBLANG_DUTCH_BELGIAN: sl = "BE"; break;
        }
      break;
    case LANG_ENGLISH:
      l = "en";
      switch (sub)
        {
        /* SUBLANG_ENGLISH_US == SUBLANG_DEFAULT. Heh. I thought
         * English was the language spoken in England.
         * Oh well.
         */
        case SUBLANG_ENGLISH_UK: sl = "GB"; break;
        case SUBLANG_ENGLISH_AUS: sl = "AU"; break;
        case SUBLANG_ENGLISH_CAN: sl = "CA"; break;
        case SUBLANG_ENGLISH_NZ: sl = "NZ"; break;
        case SUBLANG_ENGLISH_EIRE: sl = "IE"; break;
        case SUBLANG_ENGLISH_SOUTH_AFRICA: sl = "ZA"; break;
        case SUBLANG_ENGLISH_JAMAICA: sl = "JM"; break;
        case SUBLANG_ENGLISH_CARIBBEAN: sl = "@caribbean"; break; /* ??? */
        case SUBLANG_ENGLISH_BELIZE: sl = "BZ"; break;
        case SUBLANG_ENGLISH_TRINIDAD: sl = "TT"; break;
        case SUBLANG_ENGLISH_ZIMBABWE: sl = "ZW"; break;
        case SUBLANG_ENGLISH_PHILIPPINES: sl = "PH"; break;
        }
      break;
    case LANG_ESTONIAN: l = "et"; break;
    case LANG_FAEROESE: l = "fo"; break;
    case LANG_FARSI: l = "fa"; break;
    case LANG_FINNISH: l = "fi"; break;
    case LANG_FRENCH:
      l = "fr";
      switch (sub)
        {
        case SUBLANG_FRENCH_BELGIAN: sl = "BE"; break;
        case SUBLANG_FRENCH_CANADIAN: sl = "CA"; break;
        case SUBLANG_FRENCH_SWISS: sl = "CH"; break;
        case SUBLANG_FRENCH_LUXEMBOURG: sl = "LU"; break;
        case SUBLANG_FRENCH_MONACO: sl = "MC"; break;
        }
      break;
    case LANG_GEORGIAN: l = "ka"; break;
    case LANG_GERMAN:
      l = "de";
      switch (sub)
        {
        case SUBLANG_GERMAN_SWISS: sl = "CH"; break;
        case SUBLANG_GERMAN_AUSTRIAN: sl = "AT"; break;
        case SUBLANG_GERMAN_LUXEMBOURG: sl = "LU"; break;
        case SUBLANG_GERMAN_LIECHTENSTEIN: sl = "LI"; break;
        }
      break;
    case LANG_GREEK: l = "el"; break;
    case LANG_GUJARATI: l = "gu"; break;
    case LANG_HEBREW: l = "he"; break;
    case LANG_HINDI: l = "hi"; break;
    case LANG_HUNGARIAN: l = "hu"; break;
    case LANG_ICELANDIC: l = "is"; break;
    case LANG_INDONESIAN: l = "id"; break;
    case LANG_ITALIAN:
      l = "it";
      switch (sub)
        {
        case SUBLANG_ITALIAN_SWISS: sl = "CH"; break;
        }
      break;
    case LANG_JAPANESE: l = "ja"; break;
    case LANG_KANNADA: l = "kn"; break;
    case LANG_KASHMIRI:
      l = "ks";
      switch (sub)
        {
        case SUBLANG_KASHMIRI_INDIA: sl = "IN"; break;
        }
      break;
    case LANG_KAZAK: l = "kk"; break;
    case LANG_KONKANI: l = "kok"; break; /* ??? */
    case LANG_KOREAN: l = "ko"; break;
    case LANG_LATVIAN: l = "lv"; break;
    case LANG_LITHUANIAN: l = "lt"; break;
    case LANG_MACEDONIAN: l = "mk"; break;
    case LANG_MALAY:
      l = "ms";
      switch (sub)
        {
        case SUBLANG_MALAY_BRUNEI_DARUSSALAM: sl = "BN"; break;
        }
      break;
    case LANG_MALAYALAM: l = "ml"; break;
    case LANG_MANIPURI: l = "mni"; break;
    case LANG_MARATHI: l = "mr"; break;
    case LANG_NEPALI:
      l = "ne";
      switch (sub)
        {
        case SUBLANG_NEPALI_INDIA: sl = "IN"; break;
        }
      break;
    case LANG_NORWEGIAN:
      l = "no";
      switch (sub)
        {
        /* SUBLANG_NORWEGIAN_BOKMAL == SUBLANG_DEFAULT */
        case SUBLANG_NORWEGIAN_NYNORSK: l = "nn"; break;
        }
      break;
    case LANG_ORIYA: l = "or"; break;
    case LANG_POLISH: l = "pl"; break;
    case LANG_PORTUGUESE:
      l = "pt";
      switch (sub)
        {
        /* Hmm. SUBLANG_PORTUGUESE_BRAZILIAN == SUBLANG_DEFAULT */
        case SUBLANG_PORTUGUESE_BRAZILIAN: sl = "BR"; break;
        }
      break;
    case LANG_PUNJABI: l = "pa"; break;
    case LANG_ROMANIAN: l = "ro"; break;
    case LANG_RUSSIAN: l = "ru"; break;
    case LANG_SANSKRIT: l = "sa"; break;
    case LANG_SINDHI: l = "sd"; break;
    case LANG_SLOVAK: l = "sk"; break;
    case LANG_SLOVENIAN: l = "sl"; break;
    case LANG_SPANISH:
      l = "es";
      switch (sub)
        {
        case SUBLANG_SPANISH_MEXICAN: sl = "MX"; break;
        case SUBLANG_SPANISH_MODERN: sl = "@modern"; break;     /* ??? */
        case SUBLANG_SPANISH_GUATEMALA: sl = "GT"; break;
        case SUBLANG_SPANISH_COSTA_RICA: sl = "CR"; break;
        case SUBLANG_SPANISH_PANAMA: sl = "PA"; break;
        case SUBLANG_SPANISH_DOMINICAN_REPUBLIC: sl = "DO"; break;
        case SUBLANG_SPANISH_VENEZUELA: sl = "VE"; break;
        case SUBLANG_SPANISH_COLOMBIA: sl = "CO"; break;
        case SUBLANG_SPANISH_PERU: sl = "PE"; break;
        case SUBLANG_SPANISH_ARGENTINA: sl = "AR"; break;
        case SUBLANG_SPANISH_ECUADOR: sl = "EC"; break;
        case SUBLANG_SPANISH_CHILE: sl = "CL"; break;
        case SUBLANG_SPANISH_URUGUAY: sl = "UY"; break;
        case SUBLANG_SPANISH_PARAGUAY: sl = "PY"; break;
        case SUBLANG_SPANISH_BOLIVIA: sl = "BO"; break;
        case SUBLANG_SPANISH_EL_SALVADOR: sl = "SV"; break;
        case SUBLANG_SPANISH_HONDURAS: sl = "HN"; break;
        case SUBLANG_SPANISH_NICARAGUA: sl = "NI"; break;
        case SUBLANG_SPANISH_PUERTO_RICO: sl = "PR"; break;
        }
      break;
    case LANG_SWAHILI: l = "sw"; break;
    case LANG_SWEDISH:
      l = "sv";
      switch (sub)
        {
        case SUBLANG_SWEDISH_FINLAND: sl = "FI"; break;
        }
      break;
    case LANG_TAMIL: l = "ta"; break;
    case LANG_TATAR: l = "tt"; break;
    case LANG_TELUGU: l = "te"; break;
    case LANG_THAI: l = "th"; break;
    case LANG_TURKISH: l = "tr"; break;
    case LANG_UKRAINIAN: l = "uk"; break;
    case LANG_URDU:
      l = "ur";
      switch (sub)
        {
        case SUBLANG_URDU_PAKISTAN: sl = "PK"; break;
        case SUBLANG_URDU_INDIA: sl = "IN"; break;
        }
      break;
    case LANG_UZBEK:
      l = "uz";
      switch (sub)
        {
        case SUBLANG_UZBEK_CYRILLIC: sl = "@cyrillic"; break;
        }
      break;
    case LANG_VIETNAMESE: l = "vi"; break;
    default: l = "C"; break;
    }
  strcpy (bfr, l);
  if (sl != NULL)
    {
      if (sl[0] != '@')
        strcat (bfr, "_");
      strcat (bfr, sl);
    }

  return bfr;
}

#endif

And call it in guess_category_value():

          /* Last possibility is the LANG environment variable.  */
          retval = getenv ("LANG");
          if (retval == NULL || retval[0] == '\0')
#ifdef _WIN32
            return g_win32_getlocale ();
#else
           /* We use C as the default domain.  POSIX says this is
              implementation defined.  */
           return "C";
#endif