* getlocale.c (main): Rename local variable to avoid problems.

	Avoid string comparisons, rather test language and sublanguage codes
	wheere possible.  Add more code to handle Serbian language/territory
	state identical on all Windows versions.  Fix handling for "@latin"
	modifier in Belarusian locale.
	* utils.sgml (getlocale): Try to make wording and example clearer.

3 files changed, 83 insertions, 24 deletions

diff --git a/winsup/utils/ChangeLog b/winsup/utils/ChangeLog
index 28d8be499..01e7cd852 100644
--- a/winsup/utils/ChangeLog
+++ b/winsup/utils/ChangeLog
@@ -1,3 +1,12 @@
+2010-02-11  Corinna Vinschen  <corinna@vinschen.de>
+
+	* getlocale.c (main): Rename local variable to avoid problems.
+	Avoid string comparisons, rather test language and sublanguage codes
+	wheere possible.  Add more code to handle Serbian language/territory
+	state identical on all Windows versions.  Fix handling for "@latin"
+	modifier in Belarusian locale.
+	* utils.sgml (getlocale): Try to make wording and example clearer.
+
 2010-02-10  Corinna Vinschen  <corinna@vinschen.de>
 
 	* getlocale.c (usage): Change text slightly.
diff --git a/winsup/utils/getlocale.c b/winsup/utils/getlocale.c
index 4356701a6..9d6a3827a 100644
--- a/winsup/utils/getlocale.c
+++ b/winsup/utils/getlocale.c
@@ -84,6 +84,7 @@ int main (int argc, char **argv)
   int all = 0;
   const char *utf = "";
   char name[32];
+  DWORD cp;
 
   setlocale (LC_ALL, "");
   while ((opt = getopt_long (argc, argv, opts, longopts, NULL)) != EOF)
@@ -112,7 +113,7 @@ int main (int argc, char **argv)
       for (lang = 1; lang <= 0xff; ++lang)
 	{
 	  struct {
-	    wchar_t lang[256];
+	    wchar_t language[256];
 	    wchar_t country[256];
 	    char loc[32];
 	  } loc_list[32];
@@ -123,7 +124,7 @@ int main (int argc, char **argv)
 	      lcid = (sublang << 10) | lang;
 	      if (getlocale (lcid, name))
 		{
-		  wchar_t lang[256];
+		  wchar_t language[256];
 		  wchar_t country[256];
 		  int i;
 		  char *c, loc[32];
@@ -131,29 +132,37 @@ int main (int argc, char **argv)
 
 		  /* Go figure.  Even the English name of a language or
 		     locale might contain native characters. */
-		  GetLocaleInfoW (lcid, LOCALE_SENGLANGUAGE, lang, 256);
+		  GetLocaleInfoW (lcid, LOCALE_SENGLANGUAGE, language, 256);
 		  GetLocaleInfoW (lcid, LOCALE_SENGCOUNTRY, country, 256);
 		  /* Avoid dups */
 		  for (i = 0; i < lcnt; ++ i)
-		    if (!wcscmp (loc_list[i].lang, lang)
+		    if (!wcscmp (loc_list[i].language, language)
 			&& !wcscmp (loc_list[i].country, country))
 		      break;
 		  if (i < lcnt)
 		    continue;
 		  if (lcnt < 32)
 		    {
-		      wcscpy (loc_list[lcnt].lang, lang);
+		      wcscpy (loc_list[lcnt].language, language);
 		      wcscpy (loc_list[lcnt].country, country);
 		    }
+		  c = stpcpy (loc, name);
+		  /* Convert old sr_SP silently to sr_CS on old systems.
+		     Make sure sr_CS country is in recent shape. */
+		  if (lang == LANG_SERBIAN
+		      && (sublang == SUBLANG_SERBIAN_LATIN
+			  || sublang == SUBLANG_SERBIAN_CYRILLIC))
+		    {
+		      c = stpcpy (loc, "sr_CS");
+		      wcscpy (country, L"Serbia and Montenegro (Former)");
+		    }
 		  /* Now check certain conditions to figure out if that
 		     locale requires a modifier. */
-		  c = stpcpy (loc, name);
-		  if (wcsstr (lang, L"(Latin)")
-		      && (!strncmp (loc, "sr_", 3)
-			  || !strcmp (loc, "be_BY")))
+		  if (lang == LANG_SERBIAN && !strncmp (loc, "sr_", 3)
+		      && wcsstr (language, L"(Latin)"))
 		    stpcpy (c, "@latin");
-		  else if (wcsstr (lang, L"(Cyrillic)")
-			   && !strcmp (loc, "uz_UZ"))
+		  else if (lang == LANG_UZBEK
+			   && sublang == SUBLANG_UZBEK_CYRILLIC)
 		    stpcpy (c, "@cyrillic");
 		  /* Avoid more dups */
 		  for (i = 0; i < lcnt; ++ i)
@@ -167,21 +176,55 @@ int main (int argc, char **argv)
 		  if (lcnt < 32)
 		    strcpy (loc_list[lcnt++].loc, loc);
 		  /* Print */
-		  printf ("%-16s %ls (%ls)\n", loc, lang, country);
+		  printf ("%-16s %ls (%ls)\n", loc, language, country);
 		  /* Check for locales which sport a modifier for
 		     changing the codeset and other stuff. */
-		  if (!strcmp (loc, "tt_RU"))
+		  if (lang == LANG_BELARUSIAN
+		      && sublang == SUBLANG_BELARUSIAN_BELARUS)
+		    stpcpy (c, "@latin");
+		  else if (lang == LANG_TATAR
+			   && sublang == SUBLANG_TATAR_RUSSIA)
 		    stpcpy (c, "@iqtelif");
-		  else if (GetLocaleInfoW (lcid, LOCALE_SINTLSYMBOL, wbuf, 9)
+		  else if (GetLocaleInfoW (lcid,
+					   LOCALE_IDEFAULTANSICODEPAGE
+					   | LOCALE_RETURN_NUMBER,
+					   (PWCHAR) &cp, sizeof cp)
+			   && cp == 1252 /* Latin1*/
+			   && GetLocaleInfoW (lcid, LOCALE_SINTLSYMBOL, wbuf, 9)
 			   && !wcsncmp (wbuf, L"EUR", 3))
 		    stpcpy (c, "@euro");
-		  else if (!strncmp (loc, "ja_", 3)
-			   || !strncmp (loc, "ko_", 3)
-			   || !strncmp (loc, "zh_", 3))
+		  else if (lang == LANG_JAPANESE
+			   || lang == LANG_KOREAN
+			   || lang == LANG_CHINESE)
 		    stpcpy (c, "@cjknarrow");
 		  else
 		    continue;
-		  printf ("%-16s %ls (%ls)\n", loc, lang, country);
+		  printf ("%-16s %ls (%ls)\n", loc, language, country);
+		}
+	    }
+	  /* Check Serbian language for the available territories.  Up to
+	     Server 2003 we only had sr_SP (silently converted to sr_CS
+	     above), in Vista we had only sr_CS.  First starting with W7 we
+	     have the actual sr_RS and sr_ME.  However, all of them are
+	     supported on all systems in Cygwin.  So we fake them here, if
+	     they are missing. */
+	  if (lang == LANG_SERBIAN)
+	    {
+	      int sr_CS_idx = -1;
+	      int sr_RS_idx = -1;
+	      int i;
+
+	      for (i = 0; i < lcnt; ++ i)
+		if (!strcmp (loc_list[i].loc, "sr_CS"))
+		  sr_CS_idx = i;
+		else if (!strcmp (loc_list[i].loc, "sr_RS"))
+		  sr_RS_idx = i;
+	      if (sr_CS_idx > 0 && sr_RS_idx == -1)
+	      	{
+		  puts ("sr_RS@latin      Serbian (Latin) (Serbia)");
+		  puts ("sr_RS            Serbian (Latin) (Serbia)");
+		  puts ("sr_ME@latin      Serbian (Latin) (Montenegro)");
+		  puts ("sr_ME            Serbian (Latin) (Montenegro)");
 		}
 	    }
 	}
diff --git a/winsup/utils/utils.sgml b/winsup/utils/utils.sgml
index 0c6499c41..79673c274 100644
--- a/winsup/utils/utils.sgml
+++ b/winsup/utils/utils.sgml
@@ -520,7 +520,8 @@ Options:
 current user's Windows default locale to stdout.  The <literal>-s</literal>
 option prints the systems default locale instead.  With the
 <literal>-u</literal> option <command>getlocale</command> appends a ".UTF-8".
-This can be used in scripts to set the Cygwin locale, for instance:
+This can be used in scripts to set the Cygwin locale to the Windows user
+or system default, for instance, for a US-based user:
 </para>
 
 <screen>
@@ -530,21 +531,27 @@ en_US.UTF-8
 </screen>
 
 <para>The <literal>-a</literal> option is helpful to learn which locales
-are supported by your Windows machine.  It prints the locales and their
-actual meaning, like this:</para>
+are supported by your Windows machine.  It prints all available locales,
+their meaning, and the allowed modifiers.  Example:</para>
 
 <screen>
 bash$ getlocale -a
 ar_SA            Arabic (Saudi Arabia)
 ar_IQ            Arabic (Iraq)
+...
 zh_TW            Chinese (Traditional) (Taiwan)
-zh_CN            Chinese (Simplified) (People's Republic of China)
-zh_HK            Chinese (Traditional) (Hong Kong S.A.R.)
-da_DK            Danish (Denmark)
+zh_TW@cjknarrow  Chinese (Traditional) (Taiwan)
+...
+de_AT            German (Austria)
+de_AT@euro       German (Austria)
+...
 en_US            English (United States)
 en_GB            English (United Kingdom)
 en_AU            English (Australia)
 ...
+sr_RS            Serbian (Cyrillic) (Serbia)
+sr_RS@latin      Serbian (Latin) (Serbia)
+...
 </screen>
 
 </sect2>